8
V
http://doi.org/10.22214/ijraset.2020.5360
May 2020
International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 8 Issue V May 2020- Available at www.ijraset.com
Discrete-Time Controller Design for Pitch Channel Lukman Ahmed Omeiza1, Kateryna Kozak2, Salawudeen Ahmed Tijani3, Aikhonomu Oseyemen Daniel4 1
2
Department of Electrical Engineering, Federal Polytechnic Bida, Nigeria Department of Electrical Engineering, Ternopil National Technical University, Ukraine 3 Department of Electrical Engineering, University of Jos, Nigeria 4 Department of Electrical and Computer Engineering, Concordia University, Canada
Abstract: This research proposes a discrete time controller design for the pitch channel for a two degree of freedom helicopter using the root locus method. The proposed lead-lag controller uses zero and pole placement to increase the stability and controllability of the system. Simulation is provided to insure the validity of the proposed controller. Keywords: Two-degree-of-freedom, Helicopter, discrete time control, root locus, lead lag compensator I. INTRODUCTION The two degree of freedom (2 DOF) helicopter consists of a fixed base with two propellers that are driven by DC motors [1]. One propeller controls the elevation of the helicopter nose about the pitch axis and the other propeller controls the side to side motion about the yaw axis [1-2]. High resolution encoders are used to measure the pitch and yaw angles of the system. The 2-DOF helicopter recreates a behavior that is a subset of a real helicopter dynamics. The helicopter model is a Twin Rotor Multiple-inputs Multipleoutputs System. Helicopters has several non-linarites and open loop unstable dynamics as well as significant cross-coupling between their control channels which makes the control of such multi-input multi-output (MIMO) system a challenging task [3]. These nonlinearities and model uncertainties make designing a controller for helicopters an open research problem [2, 3]. The interest in this research problem has increased recently due to their potential military and civil applications [4]. Various approaches for stabilization and tracking control of helicopters have been reported in several literatures. A fuzzy control technique was presented in [3], a State Dependent Riccati Equation (SDRE) methodology in [5], back-stepping based approach in [5], and linear and non linear feedback control was presented in [6] among others. In this project the root locus method was used to design a lead lag controller using the linearized method of the system. Section 2 gives the system description and modeling, section 3 gives the design specification, section 4 gives the problem formulation and controller design, and section 4 gives the concluding remarks of the project. II. SYSTEM DYNAMICS & PROBLEM STATEMENT The two degrees of freedom (2DOF) helicopter system is a popular modeling tool due to its highly non-linear nature. The modeling and control tools of this system can be used in multiple areas such as aerospace [7, 8]. The system used in the model is a twin rotor single input single output system. The twin rotors are the yaw rotor and the pitch rotor which control the yaw and pitch of the system respectively. The system can be seen in figure 1.
Figure 1: 2DOF helicopter system.
ŠIJRASET: All Rights are Reserved
2192
International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 8 Issue V May 2020- Available at www.ijraset.com The free body diagram of the 2-DOF helicopter is illustrated in figure 2. The diagram illustrates the degrees of freedom for the helicopter using the two rotors. In this system the two degrees of freedom are around the yaw axis and pitch axis [9, 10]. The pitch angle increases positively, ( ̇ )> 0, when the nose is moved upwards, and the body rotates in the counter-clockwise (CCW) direction. The yaw angle increases positively, (̇ ) > 0 when the body rotates in the clockwise (CW) direction. When the pitch thrust force is positive the pitch increases, and when the yaw thrust force is positive the yaw increases [11].
Figure 2: Simple free-body diagram of 2-DOF Helicopter. The thrust forces acting on the pitch and yaw axes from the front and back motors are then defined [12, 13]. The non-linear equations of motion for the system are derived. Linearization can be used to simplify the non-linear dynamics of the system about a set of preselected equilibrium conditions and presented in the form: ̇= + = + The (linearized) state–space equations describing the system are: 0 1 0 0 0 0 37.2021 3.5306 −2.7451 −0.2829 0 0 ( )+ ̇( ) = () 0 0 0 1 0 0 0 0 0 −0.2701 2.3892 7.461 ( )=
1 0 0 0 0 0 1 0
( )
Where: =
̇
,
=
,
=
̇ The closed loop system presentation for the pitch channel is shown in figure 3.
Figure 3: Closed loop system (pitch channel).
©IJRASET: All Rights are Reserved
2193
International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 8 Issue V May 2020- Available at www.ijraset.com The transfer function of the system for the pitch channel is as follows. ,
( )=
( ) = ( )
37.2021 + 0.2830 + 2.7452
The closed loop system presentation for the yaw channel is shown in figure 4.
Figure 4: Closed loop system (Yaw channel). The system for the yaw angle can be expressed in the following transfer function ( ) 7.461 = , ( ) = ( ) ( + 0.2701) Both part of the system needs to be controlled as per the desired specifications. III. DESIGN SPECIFICATIONS The desired controller for the pitch channel must have an overshoot of less then 20 percent, a setteling time of les the 16 seconds, and a rise time of less the 2 seconds [14-15]. The desired specifications are presented as follows: ≤ 20%, ≤ 16 , ≤2 , ( ) = 0, ( ) = 0, The specifications for the yaw controller are presented as follows: ≤ 20%, ≤ 16 , ≤2 , ( ) = 0, ( ) = 0, The response (θ(t)) to step disturbance must settle within 16 s. For this item, we define the settling time as follows. Let θmax = max|θ(t)| (t ≥ 0) [17]. The settling time is the time after which |θ(t)| < 0.02θmax. IV. CONTROLLER DESIGN 1) Part 1: Design of pitch channel controller. The continous system is digitized in order to create the discrete time controller. The discrete transfer function of the plant with a sampling time of 0.2 since the rise time must be less or equal two seconds [18, 19]. The discrete time plant transfer function using zero order hold is as follows: ,
( )=
0.1838 + 0.1821 − 1.945 + 0.972
The discrete feedback control system can be observed in figure4.
©IJRASET: All Rights are Reserved
2194
International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 8 Issue V May 2020- Available at www.ijraset.com
Figure 5: discrete system block diagram A. Percentage overshoot Calculations Table 1: Zeta vs. Overshoot table. ≤ 0.7 0.6 0.5 0.46
5% 10% 15% 20%
Since the design specification for the overshoot is ≤ 20%, then is selected from table 1 the dampening ratio can be select as = 0.7 in order to ensure that the system is dynamic and operates within the desired specification then the dampening ratio is chosen from the information gathered from figures 5 and 6.
Figure 6: the root locus with the poles and zeros of the open loop system.
©IJRASET: All Rights are Reserved
2195
International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 8 Issue V May 2020- Available at www.ijraset.com B. Rise time calculations. Since the design specification for the rise time is ≤ 2 , can be calculated using the rise time as follows: The rise time of the system as per the desired specifications is 2 sec. The equation for the rise time calculations is: 1.8 = = 2 as per the desired specification 1.8 ≥ → ≥ 0.9 2
With
can be calculated as follows
C. Settling time calculations The settling time of the system can be estimated using the below equation 4.6 = ∗ With = 0.7 = 0.9 4.6 = = 7.3 0.7 ∗ 0.9 This means that the settling time for our chosen parameters is estimated to be 7.3 sec which is less than the desired settling time of less than 16 sec.
D. Sampling time calculations The sampling frequency is calculated using the bandwidth frequency = (−1.196 × + 1.85) = (−1.196 × 0.7 + 1.85) ∗ 0.9 = 0.9115 / From the bandwidth frequency can be calculated as follows: = 30 × = 30 × 0.9115 = 27.3450 = 27.3450 = 2 → = 4.3521 1 = = 0.229 sec
which can be derived as follows.
For this project the rise time will be chosen as 0.15 sec to give the system more dynamic freedom in our calculation. E. Desired pole calculation Desired pole calculation: ,
= − = 0.9 ,
± (1 − ) = 0.7 , = 0.15
= −(0.7)(0.9) ± (0.9) (1 − (0.7) ) = −0.7200 ± 0.5400 In z domain = , ( . )× . . . × . . × . . = = = [cos(0.081) + sin(0.081)] The desired poles of the system are: = 0.9056 + 0.0876 = 0.9056 − 0.0876 The poles of the open loop system are: ,
,
( )=
©IJRASET: All Rights are Reserved
( + 0.9859) ( − 0.9491 + 0.2400)( − 0.9491 − 0.2400)
2196
International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 8 Issue V May 2020- Available at www.ijraset.com = 0.9491 + 0.2400 = 0.9491 − 0.2400 The zero of the open loop system is: = −0.9859 These poles and zeros can be seen in the root locus of the system:
Figure 7: root locus of the open loop system. In order to stabilize the system and meet the steady state error requirements a pole will be placed near the systems zero in order to decrease and counteract its effect on the system ( = − 0.9854) And the complex poles will be counteracted using zeros in the controller [20, 21, 22]. In order to use two zeros a lead lag compensator was used. ( − )( − ) ( )= ( − )( − ) After adding the selected poles or zeros ( − 0.9090 + 0.2396)( − 0.9090 + 0.2396) ( )= ( + 0.9854)( − ) To find the pole of the controller the desired poles were used to approximate the position of the controller.
Figure 8: Angle and magnitude criteria.
©IJRASET: All Rights are Reserved
2197
International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 8 Issue V May 2020- Available at www.ijraset.com 0.0876 ( ) = 2.65°( 0.9859 + 0.9056 0.0876 = ( ) = 2.65°( 0.9854 + 0.9056 0.24 − 0.0876 = 360 − = 271.3°( 0.909 − 0.9056 0.0876 + 0.2369 = = 89.33°( 0.909 − 0.9056 0.24 + 0.0876 = 180 − = 90.6°( 0.949 − 0.909 0.2369 − 0.0876 = 360 − = 271.27°( 0.949 − 0.909 =
) ) ) ) ) )
∢ − ∢ = −180 ( ( + + )− + + ) = −180 2.65° + 89.33° + 271.25° − 90.6° − 271.3° − 2.65° − = −180 = 180 − 133.93 = 46.065 0.0876 tan(46.065) = 0.9056 − = 0.99 After some trial and error the pole was selected as 0.99 as to not be placed on the unit circle but be in a position to stabilize the system. The final lead lag controller is as follows: ( − 0.9090 + 0.2396)( − 0.9090 + 0.2396) ( )= ( + 0.9854)( − 0.99) The closed loop system with the controller is ( ) ( ) ( )= 1+ ( ) ( ) The characteristic equation (C.E) = 1 + |
( )||
( )
( ) = 0.0
( )| = |−1|
|( − 0.9090 + 0.2396)( − 0.9090 − 0.2396)| |0.1838 + 0.1821| . =1 |( + 0.9854)( − 0.99)| | − 1.945 + 0.972|
By solving the characteristic equation, the gain is found to be 0.34. After some trial and error the gain was chosen to be 2.52 (K = 2.52) in order to reach the systems specifications [23-24]. The discrete time controller can be written as follows: ( ) = 2.52 ∗
( ) = 2.52 ∗
( − 0.9090 + 0.2396)( − 0.9090 + 0.2396) ( + 0.9854)( − 0.99) − 1.817 + 0.88366 − 0.0046 + 0.9755
The closed loop transfer function is ( ) 1.035 − 0.8598 − 0.9392 + 0.9014 = ( ) − 0.8683 − 0.8682 + 0.9083 − 0.03365
©IJRASET: All Rights are Reserved
2198
International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 8 Issue V May 2020- Available at www.ijraset.com The closed loop root locus is shown in the following figure
Figure 9: the root locus for the closed loop system F. The response of the closed–loop system (θ[n]) to unit step reference input. In order to check if the system meets the required criteria the step response is simulated in MATLAB.
Figure 10: The step response of the closed loop system.
©IJRASET: All Rights are Reserved
2199
International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 8 Issue V May 2020- Available at www.ijraset.com The result of the closed loop unit step response can be observed in table 2. Table 2: closed loop system specifications 8.2441
Overshoot Rise Time Settling Time Steady state error
0 3.9000 0
The steady state error of the system is 0.0 since a pole was placed on the unit circle. 1 = 1+ = lim ( ) ( ) →
− 1.817 + 0.88366 − 0.0046 + 0.9755 0.4106 + 0.4048 ( )= − 1.898 + 0.9584 2.52 ∗ ( − 1.817 + 0.88366)(0.4106 + 0.4048) = lim = ∞ → ( − 0.0046 + 0.9755)( − 1.898 + 0.9584) So = 0. ( ) = 2.52 ∗
G.
The response of the closed–loop system (θ[n]) to unit step disturbance
( ) ( ) = ( ) 1+ ( ) ( )
Figure 11: the response of the system to a unit step disturbance.
©IJRASET: All Rights are Reserved
2200
International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 8 Issue V May 2020- Available at www.ijraset.com The steady state error of the system is 0.0 to a step input disturbance. ( ) = lim → 1+ ( ) ( ) = lim(1 + ( ) ( )) = ∞ →
=0 Table 3: closed loop system specifications (with step response to the disturbance) Overshoot 2.0032e+03 Rise Time 0 Settling Time 10.0500 Steady state error 0
From the table it can be seen that the settling time of systems reaction to a step disturbance is less than 16 seconds which means that the system specifications has been met. H. Obtain motor voltage vp[n] in response to step reference input in the pitch channel. From the step response the peak voltage can be derived [25, 26]. From this the maximum size of step input that does not result in motor saturation is calculated as follows. =
8 = 7.46 1.0720
V. CONCLUSION In this project a controller for a 2-DOF helicopter was designed using the root locus method. The poles and zeros of the lead lag controller were strategically placed to allow the maximum controllability and stability of the system. Simulation results were presented to show the result of the proposed controller under various conditions. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]
M. Hernandez-Gonzalez, A. Alanis, E. Hernandez-Vargas, Decentralized discrete-time neural control for a Quanser 2-DOF helicopter, Appl. Soft Comput.12(8)(2012)2462–2469. B. Zheng, Y. Zhong, Robust attitude regulation of a 3-DOF helicopter benchmark: Theory and experiments, IEEE Trans. Ind. Electron. 58 (2) (2011) 660–670. B. Kadmiry, D. Driankov, A Fuzzy gain-scheduler for the attitude control of an unmanned helicopter, IEEE Trans.Fuzzy Syst.12 (4) (2004) 502–515. A. Bogdanov, E. Wan, State-dependent Riccati equation control for small autonomous helicopters, J.Guid. Control Dyn. 30 (1) (2007) 47 – 60. I.A. Raptis, K.P. Valavanis, W.A. Moreno, A novel nonlinear backstepping controller design for helicopters using the rotation matrix, IEEE Trans. Control Syst. Technol. 19 (2) (2011) 465–473. El-Gendy, E. M., Saafan, M. M., Elksas, M. S., Saraya, S. F. and Areed, F. F. (2019). New Suggested Model Reference Adaptive Controller for the Divided Wall Distillation Column. Industrial and Engineering Chemistry Research, 58(17), 7247-7264. Zhang J, Mei X, Zhang D, Jiang G and Liu Q (2013). Application of decoupling fuzzy sliding mode control with active disturbance rejection for MIMO magnetic levitation system. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, 227(2), 213-229. Kinnaert M (1995) Interaction measures and pairing of controlled and manipulated variables for multiple input-multiple-output systems: a survey. Journal A, 36(4), 15-23. Van de Wal M and de Jager B (1995) Control structure design: A survey. In American Control Conference, Proceedings (Vol. 1, pp. 225-229). IEEE. Bristol E (1966) On a new measure of interaction for multivariable process control. IEEE transactions on automatic control, 11(1), 133-134. Khaki-Sedigh A and Moaveni B (2009) Control configuration selection for multivariable plants (Vol. 391). Springer. Halvarsson B (2010). Interaction Analysis in Multivariable Control Systems: Applications to Bioreactors for Nitrogen Removal Acta Universitatis Upsaliensis. Uppsala Dissertations from the Faculty of Science and Technology 92. 162 pp. Uppsala. ISBN 978-91-554-7781-3. Bequette, B. W. (2003). Process control: modeling, design, and simulation. Prentice Hall Professional Samadi B and Rodrigues L (2014). A sum of squares approach to backstepping controller synthesis for piecewise affine and polynomial systems. International Journal of Robust and Nonlinear Control, 24(16), 2365-2387. Nuthi P and Subbarao K (2015). Experimental verification of linear and adaptive control techniques for a two degrees-of-freedom helicopter. Journal of Dynamic Systems, Measurement, and Control, 137(6), 064501. Khayati K (2015). Multivariable adaptive sliding-mode observer-based control for mechanical systems. Canadian Journal of Electrical and Computer Engineering, 38(3), 253-265. Roman R C, Precup R E and David R C (2018). Second order intelligent proportional-integral fuzzy control of twin rotor aerodynamic systems. Procedia computer science, 139, 372-380.
©IJRASET: All Rights are Reserved
2201
International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 8 Issue V May 2020- Available at www.ijraset.com [18] Chang C M and Juang J G (2014). Real time TRMS control using FPGA and hybrid PID controller. In 11th IEEE International Conference on Control & Automation (ICCA) (pp. 983-988). IEEE. [19] Zeghlache S and Amardjia N (2018). Real time implementation of non linear observer-based fuzzy sliding mode controller for a twin rotor multi-input multioutput system (TRMS). Optik, 156, 391-407. [20] McFarlane D and Glover K (1990) Robust Controller Design Using Normalized Coprime Factor Plant Descriptions (Lecture Notes in Control and Information Sciences). [21] Blaˇziˇc S (2013). On periodic control laws for mobile robots. IEEE transactions on industrial electronics, 61(7), 3660-3670. [22] Taka´cs A´ , Kova´cs L, Rudas I, Precup R E and Haidegger T (2015). Models for force control in telesurgical robot systems. Acta Polytechnica Hungarica, 12(8), 95-114. [23] Apkarian J, Levis M, Fulford C (2012). Usermanual of 2-DOF helicopter experiment setup and configuration. Ontario, Canada: Quanser. [24] Halsey KMand Glover K (2005). Analysis and synthesis of nested feedback systems. IEEE transactions on automatic control, 50(7), 984-996. [25] [19] Hernandez-Gonzalez M, Alanis A Y and Hernandez-Vargas E A (2012). Decentralized discrete-time neural control for a Quanser 2-DOF helicopter. Applied Soft Computing, 12(8), 2462-2469. [26] Samadi B and Rodrigues L (2014). A sum of squares approach to backstepping controller synthesis for piecewise affine and polynomial systems. International Journal of Robust and Nonlinear Control, 24(16), 2365-2387.
©IJRASET: All Rights are Reserved
2202