Issuu

Engineering Applications of Artificial Intelligence 67 (2018) 157â€“172

Contents lists available at ScienceDirect

Engineering Applications of Artificial Intelligence journal homepage: www.elsevier.com/locate/engappai

A novel trapezoidal intuitionistic fuzzy information axiom approach: An application to multicriteria landfill site selection Cengiz Kahraman a, *, Selcuk Cebi b , Sezi Cevik Onar a , Basar Oztaysi a a b

Istanbul Technical University, Department of Industrial Engineering, 34367 Macka, Besiktas, Istanbul, Turkey Yildiz Technical University, Department of Industrial Engineering, 34349 Yildiz, Besiktas, Istanbul, Turkey

a r t i c l e

i n f o

Keywords: Information axiom Axiomatic design Intuitionistic fuzzy sets Landfill site selection Decision making

a b s t r a c t Information axiom is used to determine the best option among the potential alternatives. In case of uncertain parameters, information axiom has been extended to its ordinary fuzzy versions in the literature. In this paper, we developed a trapezoidal intuitionistic fuzzy information axiom model in order to handle the hesitancy of experts. The developed model aggregates the judgments of more than one expert and proposes a new defuzzification method for trapezoidal intuitionistic fuzzy sets. Our model was applied to a multicriteria landfill site selection problem and a sensitivity analysis was conducted to check the robustness of the given decisions. ÂŠ 2017 Elsevier Ltd. All rights reserved.

1. Introduction Information axiom (IA) is the second axiom of the axiomatic design which is used to determine design characteristics (Suh, 1990). IA measures the information content of any design based on satisfaction degree between design and system ranges. It is frequently used as a decision making tool to select the best design that satisfies independence axiom. The main advantage of information axiom method is to enable the decision makers to define the desired characteristics for the considered criteria. In addition, each alternative has its own system characteristics. There is a need for a method which can measure the overlapping levels of these two sets of characteristics. Especially, when the characteristics cannot be defined precisely, fuzzy information axiom provides an excellent tool for measuring uncertain overlapping levels. In the literature, IA is extended to its fuzzy versions in order to capture impreciseness and vagueness in design problems. Hence, fuzzy information axiom (FIA) has been drawing attention in the literature with its wide variety of applications (Kulak et al., 2010). It guides decision makers to define decision requirements via functional requirements (FR) in order to minimize information content of the design. Ordinary fuzzy information axiom models have been used in the solution of product service system development problem (Chen et al., 2015), public transportation investment problem (Kaya et al., 2012), decision support system development problem (Cebi and Kahraman, 2010), ship design problem (Cebi et al., 2010, 2012), transportation company selection problem (Kulak and Kahraman, 2005a), and advanced manufacturing system selection problem (Kulak and Kahraman, 2005b).

A landfill site is defined as a site for the disposal of waste materials by using various methods. Landfill sites are required to dispose of waste materials that it is impossible or hard to be reused or recycled. Although governments spend time and money for developing new technologies in order to reduce the amount of waste materials, there is still being huge quantity of waste materials coming from both household and commercial sector. Therefore, determining the best site for landfill facility is a critical issue in urban planning process since it includes vital impact on the economy, ecology, and the environmental health of the region. The main purpose in the landfill site selection process is to determine the best location that minimizes hazards to the environment and public health (Uyan, 2014). The determining of an optimum location for landfill among potential alternative locations is a multiple criteria decision-making problem including both quantitative and qualitative criteria. Since the measurements of qualitative criteria include imprecise or vagueness, the conventional approaches to facility location problem tend to be less effective in dealing with these impreciseness or vagueness (Kahraman et al., 2003). Ordinary fuzzy set theory proposed by Zadeh (1965) has been used in order to cope with vagueness or uncertainty in human thinking process. However, it is hard to define an exact membership function by using ordinary fuzzy sets when there is some hesitancy of experts. Therefore, the membership and non-membership functions of a fuzzy set are simultaneously defined in an intuitionistic fuzzy set (IFS). IFS requires sum of degrees of membership and non-membership values of

* Corresponding author.

E-mail address: kahramanc@itu.edu.tr (C. Kahraman). https://doi.org/10.1016/j.engappai.2017.09.009 Received 8 May 2017; Received in revised form 4 September 2017; Accepted 7 September 2017 Available online 17 October 2017 0952-1976/ÂŠ 2017 Elsevier Ltd. All rights reserved.

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157â€“172

Fig. 1. Usage percentages of the evaluation criteria.

2. Literature review: Multicriteria landfill site selection

any element to be at most equal to 1. IFSs are the excellent extension of fuzzy sets that can consider the expertsâ€™ hesitancy and vagueness together. IFSs have been later extended to Pythagorean fuzzy sets in order to provide a larger range for membership degrees (Yager, 2013). Intuitionistic and Pythagorean fuzzy judgments assigned by different experts can be aggregated by various aggregation functions (Garg, 2016, 2017a, b; Garg and Arora, 2017; Garg et al., 2017). The contribution of the proposed multicriteria information axiom method is the captured uncertainty and hesitancy in decision-making process through intuitionistic fuzzy sets. The incorporation of expertsâ€™ hesitancies into the decision model is the main research area of the recently developed fuzzy set extensions such as intuitionistic fuzzy sets, type-2 fuzzy sets, hesitant fuzzy sets, Pythagorean fuzzy sets, etc. Information axiom measures the information content which indicates the best design with the highest probability of success. It compares the design range and system range and selects the best alternative as the one having the most overlapped ranges. The proposed method is different from the other approaches since decision-making is based on the comparisons of these ranges. In the literature, there are several ordinary FIA works that try to capture the vagueness and impreciseness in the definition of system and design requirements. However, these works cannot address the hesitancy of experts. The motivation of our study is to develop intuitionistic trapezoidal FIA in order to consider this hesitancy. It is the first time that FIA is extended to intuitionistic trapezoidal fuzzy sets in this paper. For this purpose, new definitions and formulas for FIA will be presented. In addition, an original area based defuzzification method has been proposed to the literature. In the scope of this paper, landfill site selection problem will be solved by using the proposed trapezoidal intuitionistic fuzzy information axiom. The rest of this paper is organized as follows. Section 2 presents a literature review for multicriteria landfill site selection. Section 3 gives the evaluation criteria for landfill site selection problem. Section 4 presents the basics of fuzzy information axiom method. Section 5 provides intuitionistic sets and Section 6 proposes intuitionistic fuzzy information axiom. Section 7 presents the solution of a landfill site selection problem and a sensitivity analysis. Finally, the concluding remarks are given in Section 8.

The waste management process is a vital problem for urban planning and it includes various steps which are collection, transport, processing, recycling or disposing, and monitoring. The effectiveness of this process is based on the selection of the most suitable landfill site. An inappropriate decision on landfill site selection has negative impacts on environmental, economical, and social factors (Moghaddas and Namaghi, 2011; Nazari et al. 2012). The solution of this problem is hard, complex, and time-consuming process because of the evaluation of various conflicting criteria. Therefore, during the last decade, the problem has attracted a lot of research works and academic studies. The list of the most common used criteria in the literature is given in Table 1. According to the table, the most preferred criteria are Distance to surface water resources (DSW), Distance to roads (DR), Distance to residential areas (DRA), Land use (LU), and Slope of the land (SL), respectively. The usage percentages of the evaluation criteria are presented in Fig. 1. Table 2 presents the methods used for the evaluation of landfill site selection problem. In the literature, different models and methods are used in order to determine appropriate site for landfill. An integrated method combining analytic hierarchy process (AHP) and geographic information system (GIS) is the most preferred method in the literature. Additionally, some other site selection methods such as genetic algorithms (Lee et al., 2010), regression based methods (Bilginol et al., 2015), spatial multi-criteria analysis (van Haaren and Fthenakis, 2011), agent based models (Sirikijpanichkul et al., 2007), and fuzzy based models (Nazari et al., 2012) can be used for landfill site selection. Fig. 2 presents the usage percentages of evaluation methods in the literature. 3. Evaluation criteria for landfill site selection In the literature, there are lots of criteria that have been considered for the solution problem. In the scope of this study, the criteria affecting the selection of a landfill site are classified into three main groups. These groups are Environmental factors, Social factors and Economical factors. Definition of these criteria and their design ranges are as follows: 158

C. Kahraman et al.

Table 1 Evaluation criteria for landfill site selection. Environmental factors

159

GWD

✔ ✔ ✔ ✔ ✔ ✔

✔

PS ✔

✔ ✔ ✔ ✔

✔

✔ ✔

✔ ✔ ✔ ✔

✔

✔ ✔ ✔ ✔ ✔

✔ ✔

✔ ✔ ✔ ✔

✔ ✔

✔ ✔ ✔ ✔ ✔ ✔

✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔

✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔

✔ ✔ ✔ ✔ ✔ ✔

DFA

✔

✔ ✔ ✔

✔

✔ ✔

✔ ✔ ✔ ✔

✔

✔ ✔ ✔

✔ ✔

✔ ✔ ✔ ✔

✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔

✔

DTA

PDA

✔

DRW

✔

✔ ✔

✔ ✔ ✔

✔

✔ ✔ ✔ ✔ ✔ ✔

✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔

✔

✔ ✔ ✔

✔

✔ ✔

✔

✔ ✔

✔ ✔ ✔

✔ ✔

✔

✔ ✔ ✔ ✔ ✔

✔ ✔ ✔ ✔ ✔ ✔

✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔

✔

✔ ✔

✔ ✔ ✔

✔ ✔

✔

✔ ✔

Economic factors WD

✔

✔ ✔

✔

DIA

✔

DRA ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔

✔ ✔

✔ ✔ ✔ ✔

✔

✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔

✔

✔ ✔ ✔ ✔ ✔ ✔ ✔

Social factors LU

✔

✔ ✔ ✔ ✔ ✔

✔

✔ ✔

✔ ✔ ✔

✔

✔ ✔ ✔

✔

✔ ✔

✔

✔ ✔

✔ ✔ ✔ ✔ ✔ ✔

✔ ✔

✔

Distance to surface water resources (DSW), Ground water depth (GWD)-Permeability of the soil (PS), Land use (LU), Lithology (L), Hydrology (H), Geology (G), Topography (T), Distance to fault (DF), Landslide (LS), Distance to flood plain area (DFA), Distance to residential areas (DRA), Distance to industrial areas (DIA), Wind directions (WD), Distance to historical/tourist areas (DTA), Distance to protected area (DPA), Distance to railway (DRW), Distance to airports (DA), Slope of the land (SL), Distance to roads (DR), Distance to infrastructure (power lines, water pipe lines) (DI), Land prices (LP), Initial costs (Setup Costs) (IC)

Engineering Applications of Artificial Intelligence 67 (2018) 157–172

Rahmat et al., 2016 Bahrani et al., 2016 Torabi-Kaveh et al., 2016 El Maguiri et al., 2016 Djokanovic et al., 2016 Chabuk et al., 2016 Hanine et al., 2016 Eskandari et al., 2015 El Baba et al., 2015 Beskese et al., 2015 Arkoc, 2014 Afzali et al., 2014 Yal and Akgun, 2014 Uyan, 2014 Shahabi et al., 2014 Koushik et al., 2014 Kumar and Mohammad, 2013 Isalou et al., 2013 Alavi et al., 2013 Donevska et al., 2012 Vasiljević et al., 2012 Gorsevski et al., 2012 Yildirim, 2012 Moghaddas and Namaghi, 2011 Sener et al., 2010 Nas et al., 2010 Ersoy and Bulut, 2009 Guiqin et al., 2009 Ramjeawon and Beerachee, 2008 Sener et al., 2006 Melo et al., 2006

DSW

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157–172

Table 2 Tools and Methods used for landfill site selection problem.

Bahrani et al., 2016 Djokanovic et al., 2016 Rahmat et al., 2016 Torabi-Kaveh et al., 2016 El Maguiri et al., 2016 Chubuk et al., 2016 Hanine et al., 2016 Eskandari et al., 2015 El Baba et al., 2015 Arkoc, 2014 Uyan, 2014 Shahabi et al., 2014 Kumar and Mohammad, 2013 Isalou et al., 2013 Alavi et al., 2013 Donevska et al., 2012 Vasiljević et al., 2012 Gorsevski et al., 2012 Yildirim, 2012 Moghaddas and Namaghi, 2011 Sener et al., 2010 Nas et al., 2010 Ersoy and Bulut, 2009 Guiqin et al., 2009 Ramjeawon and Beerachee, 2008 Sener et al., 2010 Melo et al., 2006

GIS

Other techniques

Fuzzy sets

Application location

✔ ✔ ✔ ✔ ✔ ✔

AHP AHP AHP AHP

✔

Shabestar, Iran Pancěvo, Serbia Behbahan, Iran Iranshahr, Iran Mohammedia, Morocco Al-Hillah Qadhaa Casablanca, Morocco Marvdasht, Iran Gaza Strip Corlu Konya Saqqez Delhi Kahak Mahshahr Polog Region, Macedonia Serbian Polog Region, Macedonia Trabzon Khorasan Razavi Beysehir Konya Trabzon Beijing Mauritius Ankara Cachoeiro de Itapemirim

✔ ✔ ✔ ✔ ✔

✔

AHP AHP SAW AHP

✔

AHP Weighted linear combination Boolean logic AHP

✔ ✔

✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔

Delphi AHP AHP OWA AHP

✔ ✔

AHP AHP AHP AHP SAW AHP

✔

Table 3 Distances to surface water resources used in the literature.

Uyan, 2014 Yal and Akgun, 2013 Isalou et al., 2013 Afzali et al., 2014 Alavi et al., 2013 Vasiljević et al., 2012 Gorsevski et al., 2012 Nas et al., 2010 Ersoy and Bulut, 2009

Unit

Min

Max

m m m m m m m m m

500 100 600 100 300 500 100 300 200

2000 5000 3000

5 5

# of category

2500 2000 1000 500 1000

6 3

Unit

Min

Max

m m m m m m

5 200 300 10 2 300

70 500 3500

5 5

5 500

3 3

3 4

Table 4 Ground water depth used in the literature.

Fig. 2. Usage percentages of evaluation methods in the literature.

Arkoc, 2014 Uyan, 2014 Isalou et al., 2013 Afzali et al., 2014 Vasiljević et al., 2012 Ersoy and Bulut, 2009

3.1. Environmental factors The first group factors are defined as Distance to surface water resources (DSW), Ground water depth (GWD), Permeability of the overlaying layer (POL), Land cover (LNC), Land use (LU), and Natural disaster potential (NDP). These factors are benefit criteria such that the biggest score represents the best or most suitable value.

# of category

3.1.3. Permeability of the soil (PS) Landfills cause contamination of groundwater by seepage to the soil. Therefore, this criterion is considered as the contamination potential of ground waters is required to be minimized (Isalou et al., 2013). In some research papers, this criterion and geology is considered as the same. In the literature, this criterion is taken by the type of soil structure. However, Arkoc (2014) classified permeability into five categories between 10−7 m/s and 10−2 m/s.

3.1.1. Distance to surface water resources (DSW) Since landfills create noxious gases and leachate, a landfill should be located far away from surface water resources. Table 3 presents the distances between landfills and surface waters. The minimum value means the worst case such that landfill site cannot be located to an area closer than minimum value. Maximum value means the best case that the site takes the best performance score.

3.1.4. Land cover This criterion represents either the natural characteristic or the usage. Arkoc (2014) classified this criterion into eight classes changing from Class no. 1 to Class no. 8 based on the Turkish regulations. The most suitable agricultural area was symbolized with Class no. 1 while the least suitable area for land was represented by Class no. 8. The agricultural areas in Class no. 1 is not permitted to set any landfill sites

3.1.2. Ground water depth (GWD) This criterion represents that a land-fill site should be located in an area where ground water is sufficiently deep so that the ground-water quality is not affected (Kumar and Hassan, 2013). The distances to the groundwater used in the literature are given in Table 4. 160

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157–172

Table 5 Magnitudes of distances used in the literature. Unit Afzali et al., 2014 Yal and Akgun, 2013 Gorsevski et al., 2012 Ersoy and Bulut, 2009

m m m m

Table 7 Distances to protected areas. Min

1000 100 500 60

Max 5000 1000 500

# of category

Unit Uyan, 2014 Afzali et al., 2014 Alavi et al., 2013 Vasiljević et al., 2012

5 3

Table 6 Distances to residential areas used in the literature. Unit Afzali et al., 2014 Uyan, 2014 Yal and Akgun, 2013 Isalou et al., 2013 Alavi et al., 2013 Vasiljević et al., 2012 Nas et al., 2010 Ersoy and Bulut, 2009

m m m m m m m m

Min 2000 1000 500 300 2000 500 1000 1000

m m m m

Min

Max

500 500 2000 500

1250

# of category

10000 2000

5 4

Table 8 Distances to transportation centers. Max 4000 2000 2000 15000 2500 10000 10000

# of category

Unit Uyan, 2014 Vasiljević et al., 2012 Ersoy and Bulut, 2009 Wang et al., 2009

5 5 5 5 11 4

m m m m

Min

Max

3000 500 3000 3000

5000 3000

4 4

# of category

12000

Table 9 Slope of the land used in the literature. Unit Arkoc, 2014 Uyan, 2014 Yal and Akgun, 2013 Isalou et al., 2013 Alavi et al., 2013 Vasiljević et al., 2012 Gorsevski et al., 2012 Moghaddas and Namaghi, 2011 Nas et al., 2010 Ersoy and Bulut, 2009 Senet et al., 2010

(Arkoc, 2014). However, some researchers (Yal and Akgun, 2013, Afzali et al. 2014) considered the land use factor into two categories: suitable and unsuitable. The areas such as moors, forests, irrigated fields, gardens and pastures are considered as unsuitable whereas the dry fields and the abandoned land areas are assumed to be suitable. 3.1.5. Land use Land use represents how people use a landscape. Alavi et al. (2013) divided this factor into residential, agricultural, industrial, and unused lands whereas Gorsevski et al. (2012) classified it as water, forest, barren, and agriculture. Isalou et al. (2013) preferred agriculture and garden criteria as land use factors.

Degree % % % % % % % % % %

Min 10 10 10 10 10 2 11 5 15 5 5

Max

# of category

60 20 40 50 45 20 30 30

5 3 5 3 5 3

20 15

4 2 4 4

3.2.4. Distance to protected areas (DPA) The closer the distance to protected areas the lower the suitability score. In Table 7, the distances to protected areas used in the literature are given.

3.1.6. Natural disaster potential (NDP) This criterion represents potential of disaster such as distance to fault (DF), distance to flood plain area (DFA), landslide (LS) or erosion potential. Among these disasters, the most considered one in the literature is distance to fault factor. The magnitudes of distances used in the literature are given in Table 5. Yal and Akgun (2013) divided landslide into the following categories as high, moderate, and low or no landslide.

3.2.5. Distance to transportation centers (Railways, Airports, Stations) (DTC) Landfill sites should be selected far from transportation centers. For instance, landfill sites close to airports may cause bird hazards (Alavi et al., 2013). Distances to transportation centers in the literature are given in Table 8.

3.2. Social factors

3.3. Economical factors

In this section, social factors consist of Distance to residential areas (DRA), Distance to industrial areas (DIA), Wind directions (WD), Distance to protected area (DPA), and Distance to transportation centers (Railways, Airports, Stations) (DTC).

Economical factors are divided into two categories as Setup Factors and Operational Factors. Setup factors consists of initial costs (IC), slope of the land (SL), and land prices (LP) while operational factors include distance to roads (DR), distance to infrastructure (DI), accessibility, distance to waste production centers (WPC), and landfill capacity.

3.2.1. Distance to residential areas (DRA) Placing the landfill site close to a residential area has a negative effect since it adversely affects public health, land value, and future developments of cities (Nas et al., 2010). The distances used for this criterion in the literature are given in Table 6.

3.3.1. Setup factors Initial costs (IC): This criterion includes the assessment of the number of landowners involved in land acquisition (Ramjeawon and Beerachee, 2008). Slope of the land (SL): Since the slope of the land may lead to higher excavation costs, it is considered as a cost factor by researchers. In other words, steep slopes increase setup costs. Land slopes used in the literature are given in Table 9. Land prices (LP): Land price is another cost factor that may lead to higher setup costs. Wang et al. (2009) consider this criterion in five categories in terms of suitability of the land for landfill site.

3.2.2. Wind directions (WD) This criterion avoids site selection downwind otherwise there would be a possibility to transfer wastes toward the community (Isalou et al., 2013). 3.2.3. Distance to historical/tourist areas (DTA) A landfill site should not be placed on a site close to historic/cultural sites since prevailing wind directions may negatively affect tourism activities. A buffer zone for this criterion is defined between 1000 and 4000 m with five categories by Uyan (2014) while it is up to 2000 m in Ramjeawon and Beerachee (2008).

3.3.2. Operational factors Distance to roads (DR): Proximity to roads makes the sites less attractive because of additional costs. Therefore, the suitability ranking decreases as you move away from the main roads (Yal and Akgun, 2014; 161

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157â€“172

Table 10 Distances to roads used in the literature. Unit Uyan, 2014 Afzali et al., (2014) Isalou et al., 2013 Gorsevski et al., 2012 Nas et al., 2010 Ersoy and Bulut, 2009

m m m m m m

Min

Max

250 100 150 2000 200 100

1000 1000 3500 5000 10000 1000

# of category 5

12 3

Uyan, 2014). Table 10 presents the distances to main roads in the literature. Distance to infrastructure (DI): This criterion includes structures such as power lines, water pipe lines, telecommunications, etc. Availability of this structure near landfill site is a requirement (Ramjeawon and Beerachee, 2008; VasiljeviÄ‡ et al., 2012). However, Uyan (2014) considers it as a cost criterion. In his study, landfill site is selected with a certain distance from infrastructure systems in order to prevent damage to systems. Sener et al. (2006) consider this criterion by dividing it into two categories: suitable and unsuitable areas. Accessibility: This criterion presents traffic movements on the access roads leading to the site (Ramjeawon and Beerachee, 2008). Heavy traffic increases the transportation costs. Distance to waste production centers (WPC): This criterion is an important factor in terms of operational costs since landfill sites near to the waste production centers lead to decreasing transportation costs (Wang et al., 2009). Wang et al. (2009) define distances in five categories in a range between 500 and 2000 m. Landfill capacity: The site should serve for at least 15 years (Ramjeawon and Beerachee, 2008) and it should provide opportunities for expansion.

Fig. 3. Relationships among design range, common area, and system range (Suh, 1990).

Fig. 4. System range, design range, and common range definitions with triangular fuzzy numbers.

4. Information axiom Information Axiom (IA) is the second axiom of Axiomatic Design (AD) methodology which is proposed by Suh (1990, 2001). The method consists of two axioms. The first one is known as Independence Axiom. Based on these axioms, the objective of the AD is to compose a scientific and a systematic basis to design processes by providing a thinking process to create a new design and/or to improve the existing design (Suh, 2005). The first axiom Independence Axiom requires maintaining independence of functional requirements (FRs) which are related to main characteristics of a design. The second axiom requires maintaining minimum information content of a design. The relation between independence axiom and information axiom is that information axiom is used to select the best alternative having the minimum information content among the alternatives which satisfy independence axiom (Suh, 1990). The information axiom is symbolized by the information content (I) that is related to the probability of satisfying the design goals representing by FRs. Eq. (1) is used to the information content of a design (Suh, 1990, 2001) đ??źđ?‘– = log2

1 đ?‘?đ?‘–

If there are more than one FR, information content is calculated as follows (Suh, 1990);

âˆŤđ?‘‘đ?‘&#x;

( ) đ?‘?đ?‘ đ??š đ?‘…đ?‘– đ?‘‘đ??š đ?‘…đ?‘–

(4)

đ?‘–=1

đ??źđ?‘ đ?‘Śđ?‘ đ?‘Ąđ?‘’đ?‘š = âˆ’

đ?‘š âˆ‘

log2 đ?‘?đ?‘– =

đ?‘–=1

đ?‘š âˆ‘

( ) log2 1âˆ•đ?‘?đ?‘– .

(5)

đ?‘–=1

IA has been extended under fuzzy environment by Kulak and Kahraman (2005a, b) and Kahraman and Cebi (2009) proposed new definitions and formulations for the method in order to extent the applications of the method under fuzziness. These definitions and formulations have been used for the solution of various types of axiomatic design problems under fuzzy environment. They used fuzzy numbers to depict the design goal and properties of the alternatives. Eq. (6) is used to calculate the information content under fuzzy environment. Fig. 4 illustrates the information content calculation procedure with triangular fuzzy numbers. Both system and design ranges consist of triangular fuzzy numbers. So, information content is calculated by Eq. (7) for the given design in Fig. 4.

(1)

where đ?‘?đ?‘– is the probability of success for a given FR with DP. The probability of success (đ?‘?đ?‘– ) is calculated by the intersection of a system range and a design range defined by designers. The intersection area between system range and design range (dr) is named common area (đ??´đ?‘?đ?‘&#x; ) or common range. Fig. 3 shows how to calculate the probability of a design goal for a single FR. Design range gives the upper and lower values of the design goal and the system probability density function (system pdf) indicates the distribution of the system. Then, the đ?‘?đ?‘– is calculated as follows; đ?‘?đ?‘– =

(3)

đ??źđ?‘ đ?‘Śđ?‘ đ?‘Ąđ?‘’đ?‘š = âˆ’log2 đ?‘?(đ?‘š) ( đ?‘š ) âˆ? đ?‘?đ?‘– đ??źđ?‘ đ?‘Śđ?‘ đ?‘Ąđ?‘’đ?‘š = âˆ’log2

đ??´đ?‘&#x;đ?‘’đ?‘Ž đ?‘œđ?‘“ đ?‘†đ?‘Śđ?‘ đ?‘Ąđ?‘’đ?‘š đ?‘…đ?‘Žđ?‘›đ?‘”đ?‘’ đ??śđ?‘œđ?‘šđ?‘šđ?‘œđ?‘› đ??´đ?‘&#x;đ?‘’đ?‘Ž đ?›ž âˆ’đ?›ź đ??ź = log2 . đ?œ‡1 (đ?›ž âˆ’ đ?›˝) đ??ź = log2

(6) (7)

In Fig. 5, both system and design ranges are given with trapezoidal fuzzy numbers. So, the information content value for the given figure is calculated by Eq. (8).

(2)

( ) where đ?‘?đ?‘ đ??š đ?‘…đ?‘– is the density function.

đ??ź = log2 162

đ?›żâˆ’đ?›ź+đ?›ž âˆ’đ?œ? đ?›żâˆ’đ?›˝+đ?›ž âˆ’đ?œ

(8)

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157â€“172

Fig. 5. System range, design range, and common range definitions with trapezoidal fuzzy numbers.

Fig. 6. A trapezoidal intuitionistic fuzzy number (TrIFN).

In the general case of m FRs, the total information content for the entire system is calculated by Eq. (9). đ??źđ?‘– =

đ?‘š âˆ‘

(9)

đ??źđ?‘—

đ?‘—=1

where i refers to ith alternative.

5. Intuitionistic fuzzy sets Intuitionistic fuzzy sets have been developed by Atanassov (1986). They have been widely used in the literature (Chang et al., 2008; Yager, 2009; Cevik Onar et al., 2015; Kahraman et al., 2015a, b, 2016a, b). This extension of fuzzy sets (Atanassov, 1986, 1999) considers the membership value as well as the non-membership value to describe any x in X such that the sum of membership and non-membership is at most equal to 1.

Fig. 7. Intersection of two TrIFNs.

Definition 1. Let đ?‘‹ â‰ âˆ… be a given set. An intuitionistic fuzzy set in X is an object A given by đ??´Ěƒ =

{â&#x;¨

â&#x;Š } đ?‘Ľ, đ?œ‡đ??´Ěƒ (đ?‘Ľ) , đ?‘Łđ??´Ěƒ (đ?‘Ľ) ; đ?‘Ľđ?œ–đ?‘‹ ,

(10)

where đ?œ‡đ??´Ěƒ âˆś đ?‘‹ â†’ [0, 1] and đ?‘Łđ??´Ěƒ âˆś đ?‘‹ â†’ [0, 1] satisfy the condition 0 â‰¤ đ?œ‡đ??´Ěƒ (đ?‘Ľ) + đ?‘Łđ??´Ěƒ (đ?‘Ľ) â‰¤ 1,

(11)

( ) for every đ?‘Ľđ?œ–đ?‘‹. Hesitancy is equal to â€˜â€˜1- đ?œ‡đ??´Ěƒ (đ?‘Ľ) + đ?‘Łđ??´Ěƒ (đ?‘Ľ) â€™â€™. Fig. 8. Transpose of non-membership of a TrIFN.

Definition 2. Trapezoidal Intuitionistic Fuzzy Numbers (TrIFNs): A TrIFN đ??´Ěƒ is an intuitionistic fuzzy subset in R with membership function and non-membership function as follows âŽ§ đ?‘Ľ âˆ’ đ?‘Ž1 , âŽŞ đ?‘Ž2 âˆ’ đ?‘Ž1 âŽŞ âŽŞ1, đ?œ‡đ??´Ěƒ (đ?‘Ľ) = âŽ¨ đ?‘Ž âˆ’ đ?‘Ľ , âŽŞ 4 âŽŞ đ?‘Ž4 âˆ’ đ?‘Ž3 âŽŞ0, âŽŠ

where đ?‘Žâ€˛1 â‰¤ đ?‘Ž1 â‰¤ đ?‘Ž2 â‰¤ đ?‘Ž3 â‰¤ đ?‘Ž4 â‰¤ đ?‘Žâ€˛4 , 0 â‰¤ đ?œ‡đ??´Ěƒ (đ?‘Ľ) + đ?‘Łđ??´Ěƒ (đ?‘Ľ) â‰¤ 1 and TrIFN is ( ) denoted byđ??´Ěƒ đ?‘‡ đ?‘&#x;đ??źđ??š đ?‘ = đ?‘Ž1 , đ?‘Ž2 , đ?‘Ž3 , đ?‘Ž4 ; đ?‘Žâ€˛1 , đ?‘Ž2 , đ?‘Ž3 , đ?‘Žâ€˛4 (see Fig. 6).

for đ?‘Ž1 â‰¤ đ?‘Ľ â‰¤ đ?‘Ž2 for đ?‘Ž2 â‰¤ đ?‘Ľ â‰¤ đ?‘Ž3

đ?œ‡đ??´Ěƒ (đ?‘Ľ) , đ?‘Łđ??´Ěƒ (đ?‘Ľ)

(12)

for đ?‘Ž3 â‰¤ đ?‘Ľ â‰¤ đ?‘Ž4 Definition 3. Let đ??´Ěƒ and đ??ľĚƒ be two Atanassovâ€™s IFSs in the set X. The intersection of đ??´Ěƒ and đ??ľĚƒ is defined as in Eq. (14) (Bora and Neog, 2012). It is illustrated in Fig. 7. { } ( ( ) ( )) ( ( ) ( ))| đ??´Ěƒ âˆŠ đ??ľĚƒ = đ?‘Ľđ?‘– , min đ?œ‡đ??´Ěƒ đ?‘Ľđ?‘– , đ?œ‡đ??ľĚƒ đ?‘Ľđ?‘– , max đ?‘Łđ??´Ěƒ đ?‘Ľđ?‘– , đ?‘Łđ??ľĚƒ đ?‘Ľđ?‘– | đ?‘Ľđ?‘– âˆˆ đ?‘‹ . | (14) ( ) ( â€˛ â€˛ â€˛ â€˛ Let đ??´Ěƒ đ?‘‡ đ?‘&#x;đ??źđ??š đ?‘ = đ?‘Ž1 , đ?‘Ž2 , đ?‘Ž3 , đ?‘Ž4 ; đ?‘Ž1 , đ?‘Ž2 , đ?‘Ž3 , đ?‘Ž4 and đ??ľĚƒ đ?‘‡ đ?‘&#x;đ??źđ??š đ?‘ = đ?‘?1 , đ?‘?2 , ) đ?‘?3 , đ?‘?4 ; đ?‘?â€˛1 , đ?‘?â€˛2 , đ?‘?â€˛3 , đ?‘?â€˛4 . Assume that the magnitudes of these parameters are ( ) ( ) as in Fig. 7. In this figure, đ?œ‡đ??´Ěƒ đ?‘Ľđ?‘– and đ?œ‡đ??ľĚƒ đ?‘Ľđ?‘– represent the membership

otherwise

and đ?‘Ž âˆ’đ?‘Ľ âŽ§ 2 , âŽŞ đ?‘Ž2 âˆ’ đ?‘Žâ€˛1 âŽŞ âŽŞ0, đ?‘Łđ??´Ěƒ (đ?‘Ľ) = âŽ¨ đ?‘Ľ âˆ’ đ?‘Ž 3 âŽŞ â€˛ , âŽŞ đ?‘Ž4 âˆ’ đ?‘Ž3 âŽŞ âŽŠ1,

for đ?‘Žâ€˛1 â‰¤ đ?‘Ľ â‰¤ đ?‘Ž2 for đ?‘Ž2 â‰¤ đ?‘Ľ â‰¤ đ?‘Ž3 for đ?‘Ž3 â‰¤ đ?‘Ľ â‰¤ đ?‘Žâ€˛4

(13)

otherwise 163

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157â€“172

( ) Ěƒ respectively whereas đ?‘Łđ??´Ěƒ đ?‘Ľđ?‘– and functions of the fuzzy sets đ??´Ěƒ and đ??ľ, ( ) Ěƒ respectively. đ?‘Łđ??ľĚƒ đ?‘Ľđ?‘– are the non-membership functions of đ??´Ěƒ and đ??ľ, The intersection, đ??´Ěƒ đ?‘‡ đ?‘&#x;đ??źđ??š đ?‘ âˆŠ đ??ľĚƒ đ?‘‡ đ?‘&#x;đ??źđ??š đ?‘ is a Trapezoidal Intuitionistic ( )Fuzzy â€˛ â€˛ â€˛ â€˛ Number (TrIFN) denoted by đ??śĚƒđ?‘‡ đ?‘&#x;đ??źđ??š đ?‘ = đ?‘?1 , đ?‘?2 , đ?‘?3 , đ?‘?4 ; đ?‘?1 , đ?‘?2 , đ?‘?3 , đ?‘?4 . This intersection operation is illustrated in Fig. 7. The red colored line Ěƒ which represents the intersection of membership functions of đ??´Ěƒ and đ??ľ, is denoted by đ??śĚƒđ?‘‡ đ?‘&#x;đ??źđ??š đ?‘ . 5.1. Aggregation operators for TrIFNs Aggregation of preferences is an important problem in decision making problems. Hesitancy of different experts can be considered by aggregating intuitionistic fuzzy information. Zhang and Liu (2010) proposed an aggregation method for trapezoidal intuitionistic fuzzy numbers as follows; ([ ] [ đ??ż đ?‘€1 đ?‘€2 đ?‘ˆ ]) đ?‘€1 đ?‘€2 đ?‘ˆ Suppose đ??źđ?‘– = đ?‘Žđ??ż (đ?‘– = 1, 2, â€Ś , đ?‘›) đ?‘– , đ?‘Žđ?‘– , đ?‘Žđ?‘– , đ?‘Žđ?‘– , đ?‘?đ?‘– , đ?‘?đ?‘– , đ?‘?đ?‘– , đ?‘?đ?‘– is a set of TrIFNs, then the result is a TrIFN aggregated by Eq. (15), and ( ) đ?‘“đ?‘š đ??ź1 , đ??ź2 , â€Ś , đ??źđ?‘› [ đ?‘› đ?‘› âŽ› âŽž âˆ? âˆ? )đ?‘¤đ?‘– )đ?‘¤đ?‘– ( ( 1 âˆ’ đ?‘Žđ??ż , 1 âˆ’ 1 âˆ’ đ?‘Žđ?‘€1 , âŽœ 1âˆ’ âŽ&#x; đ?‘– đ?‘– âŽœ âŽ&#x; đ?‘–=1 đ?‘–=1 ] âŽœ âŽ&#x; đ?‘› đ?‘› âˆ? âˆ? (15) ( ) ( ) âŽœ âŽ&#x; đ?‘¤ đ?‘¤đ?‘– đ?‘– = âŽœ1 âˆ’ 1 âˆ’ đ?‘Žđ?‘€2 1âˆ’ 1 âˆ’ đ?‘Žđ?‘ˆ , âŽ&#x; đ?‘– đ?‘– âŽœ âŽ&#x; đ?‘–=1 đ?‘–=1 ]âŽ&#x; âŽœ[ đ?‘› đ?‘› đ?‘› đ?‘› âˆ? âˆ? âˆ? âˆ? ( ) ( ) ( ) ( ) đ?‘¤ đ?‘¤ đ?‘¤ đ?‘¤ âŽœ âŽ&#x; đ?‘– đ?‘– đ?‘– đ?‘?đ?‘–đ?‘€1 đ?‘– , đ?‘?đ??ż , đ?‘?đ?‘€2 , đ?‘?đ?‘ˆ , âŽ&#x; đ?‘– đ?‘– đ?‘– âŽœ âŽ? đ?‘–=1 âŽ đ?‘–=1 đ?‘–=1 đ?‘–=1 ( )đ?‘‡ where, đ?‘¤ = đ?‘¤1 , đ?‘¤2 , â€Ś , đ?‘¤đ?‘› is the weight vector of đ??źđ?‘– (đ?‘– = 1, 2, â€Ś , đ?‘›) , âˆ‘ đ?‘¤đ?‘– âˆˆ [0, 1] , đ?‘›đ?‘–=1 đ?‘¤đ?‘– = 1. Eq. (15) is transformed to Eq. (16) given in Box I, since the widespread of non-membership function must be larger than that of membership function and the midpoints of these functions must be equal to each other. These two conditions provide that the sum of membership and non-membership degrees for any x is at most equal to 1.

Ěƒ and design range (đ??ˇđ?‘…). Ěƒ Fig. 9. Intuitionistic fuzzy common area of system range (đ?‘†đ?‘…)

Ěƒ and design Fig. 10. Triangular intuitionistic fuzzy common area of system range (đ?‘†đ?‘…) Ěƒ range (đ??ˇđ?‘…).

the membership and non-membership functions in the defuzzification process. ( ) ( ) đ?œ‡2 Ă— đ?›˝3â€˛ + đ?›˝4â€˛ âˆ’ đ?›˝1â€˛ âˆ’ đ?›˝2â€˛ + đ?œ‡1 Ă— đ?›˝3 + đ?›˝4 âˆ’ đ?›˝1 âˆ’ đ?›˝2 đ??ˇđ??śđ??´ = (20) 4 The common area may also be a triangular IFN as in Fig. 10. The defuzzified value (DCA) of triangular intuitionistic fuzzy comĚƒ can be calcumon areas of system range (Ěƒ đ?‘†đ?‘…) and design range (đ??ˇđ?‘…) lated by using Eq. (21). ( ) ( ) đ?œ‡2 Ă— đ?›˝3â€˛ âˆ’ đ?›˝1â€˛ + đ?œ‡1 Ă— đ?›˝3 âˆ’ đ?›˝1 (21) đ??ˇđ??śđ??´ = 4 Let us assume that there is a decision matrix which consists of m alternatives and n criteria. To obtain the total information content đ??źđ?‘–đ?‘‡ of any design, Eq. (22) is used.

5.2. A new proposal for the defuzzification of TrIFNs A TrIFN can be represented by taking the inverse of a non-membership value as in Fig. 8 (Kahraman et al., 2017) (Kahraman et al., 2017) . The defuzzification of a TrIFN can be calculated as in Eq. (17): ( â€˛ ) ( ) đ?‘Ž + đ?‘Žâ€˛4 âˆ’ đ?‘Žâ€˛1 âˆ’ đ?‘Žâ€˛2 + đ?‘Ž3 + đ?‘Ž4 âˆ’ đ?‘Ž1 âˆ’ đ?‘Ž2 . (17) Defuzzified đ??´Ěƒ = 3 4 6. Proposed intuitionistic fuzzy information axiom

đ?‘› âˆ‘

In this paper, new definitions for intuitionistic fuzzy information axiom have been proposed to the literature. In Fig. 9, the intuitionistic Ěƒ is fuzzy common area of system range (Ěƒ đ?‘†đ?‘…) and design range (đ??ˇđ?‘…) illustrated. Defuzzified system range (DSR) is obtained by using Eq. (18): ( â€˛ ) ( ) đ?‘Ž3 + đ?‘Žâ€˛4 âˆ’ đ?‘Žâ€˛1 âˆ’ đ?‘Žâ€˛2 + đ?‘Ž3 + đ?‘Ž4 âˆ’ đ?‘Ž1 âˆ’ đ?‘Ž2 đ??ˇđ?‘†đ?‘… = . (18) 4

đ??źđ?‘–đ?‘‡ =

The common area can be obtained by the average of the common areas covered by membership functions and complement of nonmembership functions of both system range (Ěƒ đ?‘†đ?‘…) and design range Ěƒ This is the average of gray and blue diagonal areas. The defuzzi(đ??ˇđ?‘…). fied common area can be obtained by Eq. (19). ( ) ( ) Ěƒ + đ??´đ?‘&#x;đ?‘’đ?‘Ž đ?‘ đ?‘†đ?‘… Ěƒ âˆŠ đ?‘ đ??ˇđ?‘… Ěƒ đ??´đ?‘&#x;đ?‘’đ?‘Ž Ěƒ đ?‘†đ?‘… âˆŠ đ??ˇđ?‘… đ??ˇđ??śđ??´ = . (19) 2

The performance of the proposed IFIA model is examined through the sample calculations given in Table 11. Table 11 shows the calculations of eight alternative system ranges for a given intuitionistic fuzzy design range (0.6, 0.7, 0.8, 0.9; 0.5, 0.7, 0.8, 1). In order to find the information content of each alternative, DSR value is first calculated. For Alternative 1, the DSR is calculated using Eq. (18):

đ??źđ?‘–đ?‘—

đ?‘— = 1, 2, â€Ś , đ?‘› and đ?‘– = 1, 2, â€Ś , đ?‘š.

(22)

đ?‘—=1

If the criteria in the problem have different importance weights, the total weighted information content đ??źđ?‘–đ?‘¤đ?‘‡ is given by Eq. (23). đ??źđ?‘–đ?‘¤đ?‘‡ =

đ?‘› âˆ‘

đ??źđ?‘–đ?‘— đ?‘¤đ?‘—

đ?‘— = 1, 2, â€Ś , đ?‘›.

(23)

đ?‘—=1

(0.6 + 0.8 âˆ’ 0.5 âˆ’ 0.3) + (0.6 + 0.7 âˆ’ 0.5 âˆ’ 0.4) = 0.5. 4 In order to calculate the DCA value, the intersection of design range and alternative system range should be determined. The intersection

When the common area is a TrIFN as in Fig. 10, Eq. (20) calculates the defuzzified common area (DCA) by considering the heights of 164

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157â€“172

( ) âˆš ) âˆš( âŽ›âŽĄ âŽ¤ đ?‘› đ?‘› đ?‘› đ?‘› âˆš âŽœâŽ˘max 1 âˆ’ âˆ? (1 âˆ’ đ?‘Žđ??ż )đ?œ”đ?‘– , âˆ? (đ?‘?đ??ż )đ?œ”đ?‘– , âˆš 1 âˆ’ âˆ? (1 âˆ’ đ?‘Žđ?‘€1 )đ?œ”đ?‘– Ă— âˆ? (đ?‘?đ?‘€1 )đ?œ”đ?‘– ,âŽĽ đ?‘– đ?‘– đ?‘– đ?‘– âŽĽ ( ) âŽœâŽ˘ đ?‘–=1 đ?‘–=1 đ?‘–=1 đ?‘–=1 âŽĽ, đ?‘“đ?œ” đ??ź1 , đ??ź2 , â€Ś , đ??źđ?‘› = âŽœâŽ˘ âˆš ( ) ( ) âˆš đ?‘› đ?‘› đ?‘› đ?‘› âŽœâŽ˘ âˆš âˆ? âˆ? âˆ? ( )đ?œ”đ?‘– ( đ?‘€2 )đ?œ”đ?‘– ( ) âˆ? ( đ?‘ˆ )đ?œ”đ?‘– âŽĽâŽĽ đ?‘ˆ đ?œ”đ?‘– âŽœâŽ˘ âˆš 1 âˆ’ đ?‘€2 1 âˆ’ đ?‘Žđ?‘– Ă— đ?‘?đ?‘– , min 1 âˆ’ 1 âˆ’ đ?‘Žđ?‘– , đ?‘?đ?‘– âŽœâŽ˘ âŽĽ đ?‘–=1 đ?‘–=1 đ?‘–=1 đ?‘–=1 âŽŚ âŽ?âŽŁ ) ( ) âˆš âˆš( âŽž âŽ¤ âŽĄ đ?‘› đ?‘› đ?‘› đ?‘› âˆš âŽ˘min 1 âˆ’ âˆ? (1 âˆ’ đ?‘Žđ??ż )đ?œ”đ?‘– , âˆ? (đ?‘?đ??ż )đ?œ”đ?‘– , âˆš 1 âˆ’ âˆ? (1 âˆ’ đ?‘Žđ?‘€1 )đ?œ”đ?‘– Ă— âˆ? (đ?‘?đ?‘€1 )đ?œ”đ?‘– ,âŽĽâŽ&#x; đ?‘– đ?‘– đ?‘– đ?‘– âŽĽâŽ&#x; âŽ˘ đ?‘–=1 đ?‘–=1 đ?‘–=1 đ?‘–=1 âŽ˘ âˆš( ) ( ) âŽĽâŽĽâŽ&#x;âŽ&#x; đ?‘› đ?‘› đ?‘› đ?‘› âŽ˘âˆš âˆš âˆ?( âˆ?( âˆ?( )đ?œ” )đ?œ”đ?‘– )đ?œ”đ?‘– âˆ? ( đ?‘ˆ )đ?œ”đ?‘– âŽĽâŽ&#x; âŽ˘âˆš 1 âˆ’ 1 âˆ’ đ?‘Žđ?‘–đ?‘€2 đ?‘– Ă— đ?‘?đ?‘€2 , max 1 âˆ’ 1 âˆ’ đ?‘Žđ?‘ˆ , đ?‘?đ?‘– đ?‘– đ?‘– âŽĽâŽ&#x; âŽ˘ đ?‘–=1 đ?‘–=1 đ?‘–=1 đ?‘–=1 âŽŚâŽ âŽŁ ( )đ?‘‡ âˆ‘ where đ?œ” = đ?œ”1 , đ?œ”2 , â€Ś , đ?œ”đ?‘› is the weight vector of đ??źđ?‘– (đ?‘– = 1, 2, â€Ś , đ?‘›) , đ?‘¤đ?‘– âˆˆ [0, 1] , đ?‘›đ?‘–=1 đ?œ”đ?‘– = 1.

(16)

Box I.

Fig. 11. Waste management facilities of Istanbul Metropolitan Municipality (URL1).

Table 11 Performance of the proposed model. Design range: (0.6,0.7,0.8,0.9;0.5,0.7,0.8,1) Alternative system ranges

DSR

DCA

DSR/DCA

(0.4,0.5,0.6,0.7;0.3,0.5,0.6,0.8) (0.45,0.55,0.65,0.75;0.35,0.55,0.65,0.85) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.55,0.65,0.75,0.85;0.45,0.65,0.75,0.95) (0.6,0.7,0.8,0.9;0.5,0.7,0.8,1) (0.65,0.75,0.85,0.95;0.55,0.75,0.85,1) (0.7,0.8,0.9,1;0.6,0.8,0.9,1) (0.75,0.85,0.95,1;0.65,0.85,0.95,1) (0.8,0.9,1,1;0.7,0.9,1,1)

0.500 0.500 0.500 0.500 0.500 0.475 0.450 0.400 0.350

0.069 0.105 0.150 0.200 0.250 0.200 0.150 0.105 0.069

7.273 4.776 3.333 2.500 2.000 2.375 3.000 3.821 5.091

2.862 2.256 1.737 1.322 1.000 1.248 1.585 1.934 2.348

7. Application: Landfill site selection in Istanbul

of (0.6,0.7,0.8,0.9;0.5,0.7,0.8,1) and (0.4,0.5,0.6,0.7;0.3,0.5,0.6,0.8) is (0.6,0.65[0.5],0.65[0.5],0,7;0.5,0.65[0.75],0.65[0.75],0.8).

7.1. Problem definition đ??ˇđ??śđ??´ 0.75 Ă— (0.65 + 0.8 âˆ’ 0.5 âˆ’ 0.65) + 0.5 Ă— (0.65 + 0.7 âˆ’ 0.60 âˆ’ 0.65) 4 = 0.069.

Istanbul is the most crowded city in Turkey with the population of almost 15 million and it is the cultural, economic, and historic center of the country. Furthermore, it is the largest city in Europe and it is a kind of bridge connecting Asia and Europa. The geographic coordinates of Istanbul, Turkey are latitude 41â—Ś 00â€˛ 49â€˛â€˛ N and longitude 28â—Ś 56â€˛ 58â€˛â€˛ E with 39 m elevation above sea level. The city is administered by the Istanbul Metropolitan Municipality. In the city, solid wastes are

Ěƒ gets larger, the information As the common area between Ěƒ đ?‘†đ?‘… and đ??ˇđ?‘… content becomes smaller, which is wholly compatible with the classical IA and ordinary fuzzy IA models. 165

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157–172

Fig. 12. Locations of the potential alternatives.

collected by the district municipalities every day and wastes are brought to the solid waste transfer stations located at various points of Istanbul. Then, these solid wastes are transported to sanitary landfills. Fig. 11 presents the solid waste facilities of the city. An average of 16,500 tons of solid waste are disposed per day. For this purpose, Istanbul Metropolitan Municipality has three sanitary landfills including Kemerburgaz-Odayeri, Silivri-Seymen and Şile-Kömürcüoda. The capacities of these sites are 2.33×106 m2 , 2.64×106 m2 , and 2.27×106 m2 , respectively. In the near future, the capacities of these sanitary landfills will be fulfilled and new landfills are required to be determined. Hence, the main objective of this study is to determine a new sanitary landfill for İstanbul. The possible alternatives have been determined as Gümüşp1nar (A1), Dağyenice (A2), and Çiftalan (A3) since they have the suitable large areas for landfill. The locations of these alternatives are illustrated in Fig. 12.

Table 12 Criteria used for the evaluation. Evaluation criteria

Abbr.

Tang.

C1. Environmental factors C11 Distances to surface water resources C12 In terms of land use

DSW LU

✔

NDP

✔

C13 Natural disaster potential C2. Social factors C21 In terms of residential areas C22 In terms of wind directions C23 In terms of protected areas C3. Setup factors C31 In terms of initial costs C32 Land prices C4. Operational factors C41 Distance to roads C42 Distance to recycling plants C43 In terms of landfill capacity C44 Distance to waste transfer stations

7.2. Problem solution The criteria used in the evaluation are given in Table 12. The criteria are divided into four categories; ‘‘Environmental Factors’’, ‘‘Social Factors’’, ‘‘Setup Factors’’, and ‘‘Operational Factors’’. Environmental factors involves ‘‘Distances to surface water resources’’, ‘‘Land use’’, and ‘‘Natural disaster potential’’. ‘‘Residential areas’’, ‘‘Wind direction’’, and ‘‘Protected areas’’ are included under social factors. Setup factors consist of ‘‘Initial cost’’ and ‘‘Land price’’ while ‘‘Distance to roads’’, ‘‘Distance to recycling plants’’, ‘‘Landfill capacity’’ and ‘‘Distance to waste transfer stations’’ are taken into consideration under operational factors. The criterion ‘‘Distances to surface water resources’’ is obtained by measuring Euclidean distance of the location alternatives to surface water resources (Fig. 13). The landfill site should be far away from water resource to prevent surface water pollution. The criterion ‘‘Land use’’ involves land uses of the location alternatives such as agricultural areas, residential areas, tourism areas, industrial areas etc. Under this criterion, the current potential of the location are evaluated. Unused lands are considered as the most suitable while residential and tourism lands are considered as unsuitable. The Euclidean distance to fault line is used for the criterion ‘‘Natural disaster potential’’. The landfill site should be far away from fault line to prevent increases on natural disaster. The criterion ‘‘Residential areas’’ presents the effects of the locations to residential areas. The landfill site should be far away from residential areas to prevent negative effects of landfills on residential population such as dust, noise, odor etc. The criterion ‘‘Wind direction’’ involves the wind effects of the location alternatives on the residential areas since winds carry dust and odor of the site to residential area. In general, the dominant wind for Istanbul is the northeast and the second dominant wind direction is southwest (URL2). The relationship among location of the landfills, residential area and wind direction are considered under this criterion. The criterion ‘‘protected areas’’ provides

✔

✔ ✔ ✔

DRA WD DPA IC LP DR WPC LC DTC

Intangible

✔ ✔ ✔ ✔ ✔ ✔

the distance to natural, ecological and/or cultural values. The criterion ‘‘Initial costs’’ includes design and construction costs such as new road construction, infrastructure costs, improvements to existing facilities etc. The criterion ‘‘Land price’’ presents the cost of purchasing facility site. The criterion ‘‘Distance to roads’’ presents Euclidean distance to highway. Since distances to main roads increase operation costs it is an important criterion. Hence, farther distance is considered as the worst grade. The distance to recycling plants given in Fig. 13 is considered by the criterion ‘‘Distance to recycling plants’’. The recycling plants are presented in Fig. 11. When the criterion distance to roads is considered, the shorter distances to recycling plans are desired. The criterion ‘‘Landfill capacity’’ includes current capacity of the site and potential expansion of the site. The distance to the waste transfer stations which are the processing sites for the temporary deposition is considered under the criterion ‘‘Distance to waste transfer stations’’. As the other distances, this criterion also affects the operation costs. The locations of solid waste transfer stations are given in Fig. 11. There are eight waste transfer stations; Silivri, Şile, Halkalı, Aydınlı, Halkalı Yenibosna, Küçükbakkalköy, Baruthane, and Hekimbaşı. The distances of any potential alternative to the waste transfer station are obtained by the fuzzy arithmetic mean of the shortest distances of the two transfer locations. The data types are presented in Table 12. They are composed of trapezoidal fuzzy numbers for tangible criteria and linguistic subjective judgments for intangible criteria. Three experts assigned their linguistic evaluations for the intangible criteria. The criteria that involve distances are considered as by either tangible or intangible depending on the following rule. If there is only one area (such as residential area or surface water) under consideration, we 166

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157–172

Fig. 13. Distances and wind directions.

Table 13 Evaluations for the tangible criteria.

A1 A2 A3

C11 [km]

C13 [km]

C32 [TL]

C41 [km]

C42 [km]

C44 [km]

∼22 ∼7 ∼1

∼52 ∼40 ∼45

∼50 ∼120 ∼1100

∼1 ∼2 ∼3

∼70 ∼50 ∼25

∼50 ∼45 ∼35

Table 14 Linguistic evaluations of three experts for the intangible criteria.

A1 A2 A3

C12 E2

C21 E2

C22 E2

C23 E2

C31 E2

C43 E2

VG VG G

G VG F

G G F

F VG P

F G F

F G P

VG G F

G G P

G F P

G G VG

G F VG

G G VG

G F VG

VG G VG

F P G

G F VG

G P G

VG F VG

define its distance to the landfill location as a fuzzy number (Tangible). If there is more than one area with various population densities, the distances cannot be easily measured with respect to the landfill location. Under such circumstances, the appropriateness of the distance to the residential areas is defined by a linguistic term. The three experts mentioned in the study represent the point of views of their organizations. The expert 1 represents the point of view of the specialists from the local government. The expert 2 represents the point of view of academicians studying on the subject. Finally, the expert 3 represents the point of view of the non-profit organizations. The collected data are given in Tables 13 and 14, respectively. The linguistic scale given in Table 15 is adapted from Tuşe (2016) and it is used to transform linguistic variables to fuzzy numbers. Table 16 presents the corresponding trapezoidal intuitionistic fuzzy numbers (TrIFNs) for the evaluations given in Table 14. The preferences given in Table 16 are aggregated by using Eqs. (15) and (16). The TrIFNs for all aggregated preferences, evaluations and design ranges are presented in Table 17. An illustration of the calculations is given in Box II. To obtain information content values, Eq. (23) are used and the results of the calculations are presented in Table 18. The values of DCA and DSR are obtained by using Eqs. (18) and (20), respectively. Finally, the total information contents of the alternative locations are 1.02 for Gümüşpınar (A1), 1.343 Dağyenice (A2), and 1.679 for Çiftalan (A3). Hence, it is concluded that Gümüşpınar is the best location.

Table 15 Five level linguistic scale and TrIFNs. Linguistic term

Abbr.

Trapezoidal intuitionistic fuzzy numbers

Very poor Poor Fair Good Very good

VP P F G VG

(0,0.1,0.2,0.3;0,0.1,0.2,0.3) (0.1,0.2,0.3,0.4;0.05,0.2,0.3,0.5) (0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.7,0.8,0.9,1;0.7,0.8,0.9,1)

As an illustrative calculation, let us show the information content of A1 for FR11 criterion. The design range is (2,5,23,25;1,5,23,26) while the system range is (20,21,23,24;20,21,23,24). The common area is obtained as (20, 21, 23, 24; 20, 21, 23, 24) which is the same as the system range. As a result, the ratio of the common area to the system area becomes 1, and the information content is Log(1) = 0. In a similar way, the information contents of the alternatives are calculated for each criterion. In order to find the overall information content of an alternative, the weighted sum of the criteria information contents are summed up assuming that the weights of the criteria are equal. For example, total information content of Alternative 1 is calculated as:

Total I = 0.083 × (0 + 0.306 + 0 + ⋯ + 5.466) = 1.020.

167

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157â€“172 Table 16 TrIFNs for intangible criteria.

A1 A2 A3

C12

C21

C22

(0.7,0.8,0.9,1;0.7,0.8,0.9,1) (0.7,0.8,0.9,1;0.7,0.8,0.9,1) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9)

(0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7) (0.7,0.8,0.9,1;0.7,0.8,0.9,1) (0.1,0.2,0.3,0.4;0,0.2,0.3,0.5)

(0.7,0.8,0.9,1;0.7,0.8,0.9,1) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7)

C12

C21

C22

(0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.7,0.8,0.9,1;0.7,0.8,0.9,1) (0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7)

(0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7)

(0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.1,0.2,0.3,0.4;0,0.2,0.3,0.5)

C12

C21

C22

(0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7)

(0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.1,0.2,0.3,0.4;0,0.2,0.3,0.5)

(0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7) (0.1,0.2,0.3,0.4;0,0.2,0.3,0.5)

C23

C31

C43

(0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.7,0.8,0.9,1;0.7,0.8,0.9,1)

(0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7) (0.7,0.8,0.9,1;0.7,0.8,0.9,1)

C23

C31

C43

(0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7) (0.7,0.8,0.9,1;0.7,0.8,0.9,1)

(0.7,0.8,0.9,1;0.7,0.8,0.9,1) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.7,0.8,0.9,1;0.7,0.8,0.9,1)

(0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.1,0.2,0.3,0.4;0,0.2,0.3,0.5) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9)

C23

C31

C43

(0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.7,0.8,0.9,1;0.7,0.8,0.9,1)

(0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7) (0.1,0.2,0.3,0.4;0,0.2,0.3,0.5) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9)

(0.7,0.8,0.9,1;0.7,0.8,0.9,1) (0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7) (0.7,0.8,0.9,1;0.7,0.8,0.9,1)

Consider the evaluations for C31 by the three experts: (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9), (0.7,0.8,0.9,1;0.7,0.8,0.9,1), and (0.3,0.4,0.5,0.6; ([ ] [ đ??ż đ?‘€1 đ?‘€2 đ?‘ˆ ]) đ?‘€1 đ?‘€2 đ?‘ˆ 0.2,0.4,0.5,0.7). Let the aggregated value be Ag31 = đ?‘Žđ??ż , then the following results are obtained: đ?‘– , đ?‘Žđ?‘– , đ?‘Žđ?‘– , đ?‘Žđ?‘– , đ?‘?đ?‘– , đ?‘?đ?‘– , đ?‘?đ?‘– , đ?‘?đ?‘– ((( ) ( ) ( )) ( )) đ?‘Žđ??ż 1 âˆ’ 0.50.33 Ă— 1 âˆ’ 0.70.33 Ă— 1 âˆ’ 0.30.33 , 0.40.33 Ă— 0.70.33 Ă— 0.20.33 = 0.528 đ?‘– = max ((( ) ( ) ( )) ( )) đ??ż đ?‘?đ?‘– = min 1 âˆ’ 0.50.33 Ă— 1 âˆ’ 0.70.33 Ă— 1 âˆ’ 0.30.33 , 0.40.33 Ă— 0.70.33 Ă— 0.20.33 = 0.383 âˆš( (( ) ( ) ( )) ( )) đ?‘Žđ?‘€1 = đ?‘?đ?‘€1 = 1 âˆ’ 1 âˆ’ 0.60.33 Ă— 1 âˆ’ 0.80.33 Ă— 1 âˆ’ 0.40.33 Ă— 0.60.33 Ă— 0.80.33 Ă— 0.40.33 = 0.606 đ?‘– đ?‘– âˆš( (( ) ( ) ( )) ( )) 1 âˆ’ 1 âˆ’ 0.70.33 Ă— 1 âˆ’ 0.90.33 Ă— 1 âˆ’ 0.50.33 Ă— 0.70.33 Ă— 0.90.33 Ă— 0.50.33 = 0.716 đ?‘Žđ?‘€2 = đ?‘?đ?‘€2 = đ?‘– đ?‘– ((( ) ( ) ( )) ( )) đ?‘Žđ?‘ˆ 1 âˆ’ 0.80.33 Ă— 1 âˆ’ 10.33 Ă— 1 âˆ’ 0.60.33 , 0.90.33 Ă— 10.33 Ă— 0.70.33 = 0.857 đ?‘– = min ((( ) ( ) ( )) ( )) đ?‘?đ?‘ˆ 1 âˆ’ 0.50.33 Ă— 1 âˆ’ 0.70.33 Ă— 1 âˆ’ 0.30.33 , 0.40.33 Ă— 0.70.33 Ă— 0.20.33 = 1 đ?‘– = max Box II.

19 have been obtained. These results indicate that the ranking is still the same as the previous analyses.

7.3. Sensitivity analysis

Sensitivity analysis is conducted to reveal the robustness of the given decisions. The weights of the criteria are changed and the rankings of alternatives are observed. The results of sensitivity analysis are given in Fig. 14. From the sensitivity graphs in Fig. 14, we conclude that as the weights of DR11 and DR12 increase, A3 is always the selected alternative with an increasing superiority. This is also true for DR21 and DR22. As the weight of DR13 increases, a2 gets the superiority over A1. Increases in the weight of DR23 affect A3 negatively and cause A2 to be selected. Similar to DR13, important DRs for A2 are DR31, DR32, and DR43. Even an increase in the weight of DR41 affects all alternatives negatively, A3 is insensitively the best alternative. A1 becomes the best alternative when the weight of DR42 is larger than 0.25 or the weight of DR44 is larger than 0.20. All the calculations above are based on equal criteria weights. These calculations indicate that unless the criterion weight for DR44 is not larger than about 0.15, Alternative A1 is always the selected alternative. Additionally, we consulted an environmental expert for his evaluation on the importance degrees of the considered criteria. Based on the environmental expertâ€™s evaluation, the Information Contents in Table

7.4. Comparative analysis and discussion In this sub-section we compare our study with Kahraman and Cebi (2009)â€™s study to show the advantages. In Kahraman and Cebi (2009)â€™s study, ordinary fuzzy sets have been used to capture the vagueness in the design parameters, ignoring the possible hesitancies of experts about the values of design parameters. The present study handles both the membership and nonmembership degrees assigned by the experts that the sum of these degrees does not necessarily have to be equal to 1.0. The comparison with ordinary FIA presents the difference of the present work with Kahraman and Cebi (2009)â€™s study. Table 20 presents the results obtained from ordinary fuzzy information axiom approach. Based on these results, A1 is selected as the best alternative and the rank of the alternatives are same in the both approaches. However, there are differences among the numerical values obtained from ordinary FIA and trapezoidal intuitionistic FIA approaches. The comparative analysis by using equal weights for FRs shows that the information contents of 16 evaluations are the same in both 168

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157â€“172

Table 17 Design ranges and the aggregated preferences for the alternatives. Criteria

Evaluations

CR11

FR A1 A2 A3

(2,5,23,25;1,5,23,26) (20,21,23,24;20,21,23,24) (4,5,10,11;4,5,10,11) (0,0.5,2,3;0,0.5,2,3)

CR31

(0.6,0.75,0.95,1;0.55,0.75,0.95,1) (0.528,0.606,0.716,0.857;0.383,0.606,0.716,1) (0.32,0.392,0.499,0.637;0.159,0.392,0.499,0.68) (0.7,0.8,0.9,1;0.644,0.8,0.9,1)

CR12

FR A1 A2 A3

(0.6,0.73,0.9,1;0.55,0.73,0.9,1) (0.578,0.671,0.776,0.932;0.482,0.671,0.776,1) (0.644,0.737,0.842,0.965;0.581,0.737,0.842,1) (0.374,0.467,0.569,0.683;0.252,0.467,0.569,0.761)

CR32

(0,40,400,1100;0,40,400,1150) (40,45,55,60;40,45,55,60) (100,110,130,140;100,110,130,140) (1000,1050,1150,1200;1000,1050,1150,1200)

CR13

FR A1 A2 A3

(42,49,55,65;41,49,55,66) (48,50,54,56;48,50,54,56) (36,38,42,44;36,38,42,44) (41,43,47,49;41,43,47,49)

CR41

(0,0.5,2,3;0,0.5,2,4) (0,0.5,1,2;0,0.5,1,2) (0,1,3,4;0,1,3,4) (1,2,4,5;1,2,4,5)

CR21

FR A1 A2 A3

(0.45,0.65,0.85,1;0.4,0.65,0.85,1) (0.3,0.4,0.5,0.6;0.2,0.4,0.5,0.7) (0.578,0.671,0.776,0.932;0.482,0.671,0.776,1) (0.172,0.262,0.365,0.476;0.079,0.262,0.365,0.559)

CR42

(0,20,54,70;0,20,54,72) (66,68,72,74;66,68,72,74) (46,48,52,54;46,48,52,54) (21,23,27,29;21,23,27,29)

CR22

FR A1 A2 A3

(0.45,0.65,1,1;0.4,0.65,1,1) (0.578,0.671,0.776,0.932;0.482,0.671,0.776,1) (0.441,0.533,0.635,0.748;0.317,0.533,0.635,0.828) (0.172,0.262,0.365,0.476;0.079,0.262,0.365,0.559)

CR43

(0.5,0.65,0.9,1;0.45,0.65,0.9,1) (0.578,0.671,0.776,0.932;0.482,0.671,0.776,1) (0.239,0.328,0.431,0.542;0.126,0.328,0.431,0.626) (0.644,0.737,0.842,0.965;0.581,0.737,0.842,1)

CR23

FR A1 A2 A3

(0.5,0.75,0.95,1;0.45,0.75,0.95,1) (0.5,0.6,0.7,0.8;0.4,0.6,0.7,0.9) (0.441,0.533,0.635,0.748;0.317,0.533,0.635,0.828) (0.7,0.8,0.9,1;0.7,0.8,0.9,1)

CR44

(0,13,37,48;0,13,33,48) (46,48,52,54;46,48,52,54) (41,43,47,49;41,43,47,49) (31,33,37,39;31,33,37,39)

(a).

Fig. 14. Sensitivity analysis.

169

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157â€“172

(b).

Fig. 14. (continued) Table 19 Weighted information content based on environmental expertâ€™s evaluation. Table 18 Information contents.

FRs

FR11 FR12 FR13 FR21 FR22 FR23 FR31 FR32 FR41 FR42 FR43 FR44

0.000 0.306 0.000 1.862 0.000 0.595 0.853 0.000 0.000 3.158 0.000 5.466

0.000 0.000 4.158 0.000 0.552 1.154 4.417 0.000 0.415 0.000 3.568 1.851

3.100 2.790 1.019 4.060 4.060 0.000 0.000 3.855 1.263 0.000 0.000 0.000

Total I

1.020

1.343

1.679

Criteria weights

FR11 100 FR12 70 FR13 90 FR21 80 FR22 70 FR23 60 FR31 80 FR32 90 FR41 60 FR42 70 FR43 80 FR44 70 Weighted Total I

Normalized weights

0.109 0.076 0.098 0.087 0.076 0.065 0.087 0.098 0.065 0.076 0.087 0.076

0 0.306 0 1.862 0 0.595 0.853 0 0 3.158 0 5.466 0.954

0 0 4.158 0 0.552 1.154 4.417 0 0.415 0 3.568 1.851 1.387

3.1 2.79 1.019 4.06 4.06 0 0 3.855 1.263 0 0 0 1.771

the superiority of the proposed intuitionistic fuzzy FIA. It can be concluded that the trapezoidal intuitionistic fuzzy information approach may present drastic results when the hesitancy on the evaluation is larger.

approaches. However, 16 evaluations show a decrease while 4 evaluations indicate an increase. In two functional requirements (FR12 and FR22) information contents exhibit both an increase and a decrease. These changes are caused by the consideration of hesitancies of experts. Even these changes in our case have not affected the rankings, it may cause a significant change in the rankings in another case study. Besides, the consideration of different weights for FRs may put forward

8. Conclusion Multicriteria information axiom is a very useful tool for the evaluation of alternatives when the data of design requirements and functional 170

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157–172

Table 20 Evaluation results obtained from ordinary fuzzy information axiom approach. FRs

FR11 FR12 FR13 FR21 FR22 FR23 FR31 FR32 FR41 FR42 FR43 FR44 Total I

0.000 0.278 0.000 2.415 0.000 0.637 0.950 0.000 0.000 3.755 0.000 5.285 1.110

0.000 0.000 4.755 0.000 0.529 1.272 6.509 0.000 0.585 0.000 5.901 1.671 1.768

4.170 3.876 1.140 7.562 7.562 0.000 0.000 4.492 1.585 0.000 0.000 0.000 2.532

Chabuk, A., Al-Ansari, N., Hussain, H.M., Knutsson, S., Pusch, R., 2016. Landfill site selection using geographic information system and analytical hierarchy process: A case study Al-Hillah Qadhaa, Babylon, Iraq. Waste Manage. Res. 34 (5), 427–437. Chang, K.-H., Cheng, C.-H., Chang, Y.-C., 2008. Reliability assessment of an aircraft propulsion system using IFS and OWA tree. Eng. Optim. 40 (10), 907–921. Chen, D.P., Chu, X.N., Sun, X.W., Li, Y.P., 2015. A new product service system concept evaluation approach based on information axiom in a fuzzy-stochastic environment. Int. J. Comput. Integr. Manuf. 28 (11), 1123–1141. Djokanovic, S., Abolmasov, B., Jevremovic, D., 2016. GIS application for landfill site selection: a case study in Pančevo, Serbia. Bull. Eng. Geol. Environ. 75 (3), 1273– 1299. Donevska, K.R., Gorsevski, P.V., Jovanovski, M., Pesevski, I., 2012. Regional nonhazardous landfill site selection by integrating fuzzy logic, AHP and geographic information systems. Environ. Earth. Sci. 67, 121–131. El Baba, M., Kayastha, P., De Smedt, F., 2015. Landfill site selection using multi-criteria evaluation in the GIS interface: a case study from the Gaza Strip Palestine. Arab. J. Geosci. 8, 7499–7513. El Maguiri, A., Kissi, B., Idrissi, L., Souabi, S., 2016. Landfill site selection using GIS, remote sensing and multicriteria decision analysis: case of the city of Mohammedia, Morocco. Bull. Eng. Geol. Environ. 75 (3), 1301–1309. Ersoy, H., Bulut, F., 2009. Spatial and multi-criteria decision analysis-based methodology for landfill site selection in growing urban regions. Waste Manag. Res. 27, 489–500. Eskandari, M., Homaee, M., Mahmoodi, S., Pazira, E., Van Genuchten, M.T., 2015. Optimizing landfill site selection by using land classification maps. Environ. Sci. Pollut. Res. 22 (10), 7754–7765. Garg, H., 2016. Generalized intuitionistic fuzzy interactive geometric interaction operators using einstein t-norm and t-conorm and their application to decision making. Comput. Ind. Eng. 101, 53–69. Garg, H., 2017a. Generalized Pythagorean fuzzy Geometric aggregation operators using Einstein t-norm and t-conorm for multicriteria decision-making process. Int. J. Intell. Syst. 32 (6), 597–630. Garg, H., 2017b. Novel intuitionistic fuzzy decision making method based on an improved operation laws and its application. Eng. Appl. Artif. Intell. 60, 164–174. Garg, H., Agarwal, N., Tripathi, A., 2017. Choquet integral-based information aggregation operators under the interval-valued intuitionistic fuzzy set and its applications to decision-making process. Int. J. Uncertain. Quantif.. Garg, H., Arora, R., 2017. Generalized and group-based generalized intuitionistic fuzzy soft sets with applications in decision-making. Appl. Intell. 1–14. Gorsevski, P.V., Donevska, K.R., Mitrovski, C.D., Frizado, J.P., 2012. Integrating multicriteria evaluation techniques with geographic information systems for landfill site selection: A case study using ordered weighted average. Waste Manage. 32, 287–296. Hanine, M., Tikniouine, O., Boutkhoum, A., Agouti, T., 2016. Comparison of fuzzy AHP and fuzzy TODIM methods for landfill location selection. SpringerPlus 5, 501. Isalou, A.A., Zamani, V., Shahmoradi, B., Alizadeh, H., 2013. Landfill site selection using integrated fuzzy logic and analytic network process (F-ANP). Environ. Earth Sci. 68, 1745–1755. Kahraman, C., Cebi, S., 2009. A new multi-attribute decision making method: hierarchical fuzzy axiomatic design. Expert Syst. Appl. 36 (3), 4848–4861. Kahraman, C., Cevik Onar, S., Cebi, S., Oztaysi, B., 2017. Extension of information axiom from ordinary to intuitionistic fuzzy sets: an application to search algorithm selection. Comput. Ind. Eng. 105, 348–361. Kahraman, C., Cevik Onar, S., Oztaysi, B., 2015a. Engineering economic analyses using intuitionistic and hesitant fuzzy sets. J. Intell. Fuzzy Syst. 29 (3), 1151–1168. Kahraman, C., Cevik Onar, S., Oztaysi, B., 2015b. Fuzzy multicriteria decision making: A literature review. Int. J. Comput. Intell. Syst. 8 (4), 637–666. Kahraman, C., Cevik Onar, S., Oztaysi, B., 2016a. A comparison of wind energy investment alternatives using interval-valued intuitionistic fuzzy benefit/cost analysis. Sustainability 8 (2), Art. no. 118. Kahraman, C., Oztaysi, B., Cevik Onar, S., 2016b. A comprehensive literature review of 50 years of fuzzy set theory. Int. J. Comput. Intell. Syst. 9 (1), 3–24. Kahraman, C., Ruan, D., Doğan, I., 2003. Fuzzy group decision-making for facility location selection. Inform. Sci. 157, 135–153. Kaya, İ., Oztaysi, B., Kahraman, C., 2012. A two-phased fuzzy multicriteria selection among public transportation investments for policy-making and risk governance. Int. J. Uncertain. Fuzziness Knowl.-Based Syst. 20/01, 31–48. Koushik, Paul, Dutta, Amit, Krishna, A.P., 2014. A comprehensive study on landfill site selection for Kolkata City, India. J. Air Waste Manage. Assoc. 64 (7), 846–861. Kulak, O., Cebi, S., Kahraman, C., 2010. Applications of axiomatic design principles: A literature review. Expert Syst. Appl. 37 (9), 6705–6717. Kulak, O., Kahraman, C., 2005a. Fuzzy multi-attribute selection among transportation companies using axiomatic design and analytic hierarchy process. Inform. Sci. 170, 191–210. Kulak, O., Kahraman, C., 2005b. Multi-attribute comparison of advanced manufacturing systems using fuzzy vs. crisp axiomatic design approach. Int. J. Prod. Econ. 95, 415– 424. Kumar, S., Mohammad, I.H., 2013. Selection of a landfill site for solid waste management: An application of AHP and spatial analyst tool. J. Indian Soc. Remote Sens. 41 (1), 45–56.

requirements can be obtained. Vague and subjective evaluations for DRs and FRs can be captured by using fuzzy sets. The proposed trapezoidal intuitionistic fuzzy information axiom could handle these type of assessments including the hesitancy of the experts and successfully evaluated the landfill alternatives. The proposed model utilizes the common areas of trapezoidal intuitionistic fuzzy numbers for DRs and FRs. A careful calculation for common areas is needed in the proposed model since you may face various shapes of intersection of DRs and FRs. The proposed model also includes new defuzzification methods for intuitionistic fuzzy sets and a modified aggregation method. A real landfill site selection problem has been solved by using 4 main criteria, 12 sub-criteria, and 3 alternatives. Sensitivity analyses showed that the obtained rankings are sensitive to the weights of the criteria. This requires a careful evaluation process in the implementation of the proposed method. For further research, instead of intuitionistic fuzzy sets, we suggest other extensions of fuzzy sets such as type-2 fuzzy sets, hesitant fuzzy sets, Pythagorean fuzzy sets, or neutrosophic sets to be used in this work. The obtained results can be compared with our results. References Afzali, A., Sabri, S., Rashid, M., Samani, J.M.V., Ludin, A.N.M., 2014. Inter-municipal landfill Site Selection Using Analytic Network Process. Water Resour. Manage. 28 (8), 2179–2194. Alavi, N., Goudarzi1, G., Babaei1, A.A., Jaafarzadeh, N., Hosseinzadeh, M., 2013. Municipal solid waste landfill site selection with geographic information systems and analytical hierarchy process: a case study in Mahshahr County, Iran. Waste Manage. Res. 31 (1), 98–105. Arkoc, O., 2014. Municipal solid waste landfill site selection using geographical information systems: a case study from Corlu, Turkey. Arab. J. Geosci. 7 (11), 4975–4985. Atanassov, K.T., 1986. Intuitionistic fuzzy sets. Fuzzy Sets and Systems 20, 87–96. Atanassov, K.T., 1999. Intuitionistic Fuzzy Sets, Theory and Applications. Physica-Verlag, Heidelberg. Bahrani, S., Ebadi, T., Ehsani, H., Yousefi, H., Maknoon, R., 2016. Modeling landfill site selection by multi-criteria decision making and fuzzy functions in GIS, case study: Shabestar, Iran. Environ. Earth Sci. 75 (4), 337. Beskese, A., Demir, H.H., Ozcan, H.K., Okten, H.E., 2015. Landfill site selection using fuzzy AHP and fuzzy TOPSIS: a case study for Istanbul. Environ. Earth Sci. 73 (7), 3513–3521. Bilginol, K., Denli, H.H., Seker, D.Z., 2015. Ordinary least squares regression method approach for site selection of automated teller machines (ATMs). Proc. Environ. Sci. 26, 66–69. Bora, M., Neog, T.J., 2012. A note on intuitionistic fuzzy sets over two universes. Int. J. Basic Sci. Soc. Sci. 1 (1), 1–14. Cebi, S., Celik, M., Kahraman, C., 2010. Structuring ship design project approval mechanism towards installation of operator system interfaces via fuzzy axiomatic design principles. Inform. Sci. 180 (6), 886–895. Cebi, S., Celik, M., Kahraman, C., 2012. An axiomatic decision aid for marine equipment selection in ship machinery design and installation projects. J. Mult.-Valued Logic Soft Comput. 18, 223–245. Cebi, S., Kahraman, C., 2010. Developing a group decision support system based on fuzzy information axiom. Knowl.-Based Syst. 23 (1), 3–16. Cevik Onar, S., Oztaysi, B., Otay, I., Kahraman, C., 2015. Multi-expert wind energy technology selection using interval-valued intuitionistic fuzzy sets. Energy 90 (Part 1), 274–285.

171

C. Kahraman et al.

Engineering Applications of Artificial Intelligence 67 (2018) 157–172 Torabi-Kaveh, M., Babazadeh, R., Mohammadi, S.D., Zaresefat, M., 2016. Landfill site selection using combination of GIS and fuzzy AHP, a case study: Iranshahr, Iran. Waste Manage. Res. 34 (5), 438–448. Tuşe, Della A., 2016. A trapezoidal intuitionistic fuzzy mcdm method based on some aggregation operators and several ranking methods. Stud. Univ. Babeş-Bolyai Inform. LXI (1), 23–40. URL 1. https://atikyonetimi.ibb.gov.tr/faaliyet-haritasi/#toggle-id-1. URL 2. http://www.hurriyet.com.tr/lodos-ve-poyrazin-yaptiklari-12472532. Uyan, M., 2014. MSW landfill site selection by combining AHP with GIS for Konya, Turkey. Environ. Earth Sci. 71, 1629–1639. van Haaren, R., Fthenakis, V., 2011. GIS-based wind farm site selection using spatial multicriteria analysis (SMCA): Evaluating the case for New York State. Renew. Sustain. Energy Rev. 15 (7), 3332–3340. Vasiljević, T.Z., Srdjevic, Z., Bajcetic, R., Miloradov, M.R., 2012. GIS and the analytic hierarchy process for regional landfill site selection in transitional countries: A case study from Serbia. Environ. Manage. 49, 445–458. Wang, G., Lib, Q., Guoxuea, L., Chen, L., 2009. Landfill site selection using spatial information technologies and AHP: A case study in Beijing, China. Journal of Environmental Managemen 90, 2414–2421. Yager, R.R., 2009. Some aspects of intuitionistic fuzzy sets. Fuzzy Optim. Decis. Mak. 8, 67–90. Yager, R., 2013. Pythagorean fuzzy subsets, 2013 Joint IFSA World Congress and NAFIPS Annual Meeting (IFSA/NAFIPS), p. 57-61. http://dx.doi.org/10.1109/IFSA-NAFIPS. 2013.6608375. Yal, G.P., Akgun, H., 2013. Landfill site selection and landfill liner design for Ankara, Turkey. Environ. Earth Sci. 70, 2729–2752. Yal, G.P., Akgun, H., 2014. Landfill site selection utilizing TOPSIS methodology and clay liner geotechnical characterization: a case study for Ankara, Turkey. Bull. Eng. Geol. Environ. 73, 369–388. Yildirim, V., 2012. Application of raster-based GIS techniques in the siting of landfills in Trabzon Province, Turkey: a case study. Waste Manage. Res. 30, 949–960. Zadeh, L., 1965. Fuzzy sets. Inf. Control 8, 338–353. Zhang, X., Liu, P., 2010. Method for aggregating triangular fuzzy intuitionistic fuzzy information and its application to decision making. Technol. Econ. Dev. Econ. 16 (2), 280–290.

Lee, K.-H., Jun, S.-O., Pak, K.-H., Lee, D.-H., Lee, K.-W., Park, J.-P., 2010. Numerical optimization of site selection for offshore wind turbine installation using genetic algorithm. Curr. Appl. Phys. 10 (2), 302–306. Melo, A.L.O., Calijuri, M.L., Duarte, I.C.D., Azevedo, R.F., Lorentz, J.F., 2006. Strategic decision analysis for selection of landfill Sites. J. Surv. Eng. 132, 83–92. Moghaddas, N.H., Namaghi, H.H., 2011. Hazardous waste landfill site selection in Khorasan Razavi Province, Northeastern Iran. Arab. J. Geosci. 4, 103–113. Nas, B., Cay, T., Iscan, F., Berktay, A., 2010. Selection of MSW landfill site for Konya, Turkey using GIS and multi-criteria evaluation. Environ. Monit. Assess. 160, 491–500. Nazari, A., Salarirad, M.M., Bazzazi, A.A., 2012. Landfill site selection by decision-making tools based on fuzzy multi-attribute decision-making method. Environ. Earth Sci. 65, 1631–1642. Rahmat, Z.G., Niri, M.V., Alavi, N., Goudarzi, G., Babaei, A.A., Baboli, Z., Hosseinzadeh, M., 2016. Landfill site selection using GIS and AHP: a case study: Behbahan, Iran. KSCE J. Civ. Eng. 21 (1), 111–118. Ramjeawon, T., Beerachee, B., 2008. Site selection of sanitary landfills on the small island of Mauritius using the analytical hierarchy process multi-criteria method. Waste Manage. Res. 26, 439–447. Sener, B., Suzen, M.L., Doyuran, V., 2006. Landfill site selection by using geographic information systems. Environ. Geol. 49, 376–388. Sener, S., Sener, E., Nas, B., Karagüzel, R., 2010. Combining AHP with GIS for landfill site selection: Acase study in the Lake Beysehir catchment area (Konya, Turkey). Waste Manage. 30, 2037–2046. Shahabi, H., Keihanfard, S., Ahmad, B.B., Amiri, M.J.T., 2014. Evaluating Boolean, AHP and WLC methods for the selection of waste landfill sites using GIS and satellite images. Environ. Earth Sci. 71, 4221–4233. Sirikijpanichkul, A., Van Dam, K., Ferreira, L., Lukszo, Z., 2007. Optimizing the location of intermodal freight hubs: An overview of the agent based modelling approach. J. Transp. Syst. Eng. Inf. Technol. 7 (4), 71–81. Suh, N.P., 1990. The Principles of Design. Oxford University Press, New York. Suh, N.P., 2001. Axiomatic Design: Advances and Applications. Oxford University Press, New York. Suh, N.P., 2005. Complexity Theory and Applications. Oxford University Press, NY.

172