Issuu

Complex System with Flows and Synchronization Linfan MAO 1. Chinese Academy of Mathematics and System Science, Beijing 100190, P.R.China 2. Academy of Mathematical Combinatorics & Applications (AMCA), Colorado, USA E-mail: maolinfan@163.com

Abstract: A complex system S consists of m components, maybe inconsistence with m ≥ 2 such as those of self-adaptive systems, particularly the biological systems and usually, a system with contradictions, i.e., there are no a classical mathematical subfield applicable. Then, how can we hold its global and local behaviors or true face? All of us know that there always exists universal connections between things in the world, i.e., a topological → − graph G underlying parts in S . We can thereby establish mathematics over a graph family n→ o − → − G 1 , G 2 , · · · for characterizing the dynamic behaviors of system S on the time t, i.e., → − → − complex flows. Formally, a complex flow G L is a topological graph G associated with A+ vu (v, u) and A+ a mapping L : (v, u) → L(v, u), 2 end-operators A+ vu : L(v, u) → L uv : + Auv L(u, v) → L (u, v) on a Banach space B over a field F with L(v, u) = −L(u, v) and → − A+ vu (v, u) for ∀(v, u) ∈ E A+ (−L(v, u)) = −L G holding with continuity equations vu dxv = dt

LAvu (v, u) ,

u∈NG (v)

∀v ∈ V

→ − G ,

→ − where xv is a variable on vertex v for ∀v ∈ E G . Particularly, if dxv /dt = 0 for → − → − ∀v ∈ V G , such a complex flow G L is nothing else but an action flow or conservation

flow. The main purpose of this lecture is to clarify the complex system with that of contradictory system and its importance to the reality of a thing T by extending Banach or Hilbert n→ o − → − spaces to Banach or Hilbert continuity flow spaces over topological graphs G 1 , G 2 , · · ·

and establishing the global differential theory on complex flows, characterize the global dynamic behaviors of complex systems, particularly, complex networks independent on graphs, for instance the synchronization of complex systems by applying global differential on the → − complex flows G L .

Key Words: Complex system, Smarandache multispace, continuity flow, Banach space, Hilbert space, differential, Taylor formula, L’Hospital’s rule, mathematical combinatorics.

AMS(2010): 34A26, 35A08, 46B25, 92B05, 05C10, 05C21, 34D43, 51D20. §1. Introduction Is our mathematical theory can already be used for understanding the reality of all things in the world? This is a simple but essential question on the developing direction of mathematics, 1 An

invited J.C.& K.L.Saha Memorial Lecture in ICGMMCP-2017, December 05-07, 2017.

and it’s answer is not positive. All of us live in a world full of colors, encountering various phenomena such as those of gorgeous guppy or peacock shown in Fig.1 each day and cant help ourselves: Why are they looks like this, not that, i.e., the reality of things in the macro and micro world

Fig.1 For example, we all known or heard that birds flying in the sky and fishes swimming in the ocean from disorderly to orderly in the macro world, and also words that the superposition, i.e., a quantum particle is both in two or more possible states of being in the micro world. All of these show a beauty of the scenery in one’s eyes. It should be noted that the superposition not only appeared in the micro world but also in the macro world. For example, there is a most reluctant to answer question for a Chinese man. That is, if one’s wife and his mother fell simultaneously in a river, who will he save first, his mother or his wife? This Chinese question is also equivalent to the famous thought model of Schr¨odinger’s cat, which assumed that a cat, a flask of poison and a radioactive source are placed in a sealed box. If an internal monitor detects radioactivity, the flask is shattered, releasing the poison which kills the cat. Yet, when one looks in the box, one sees the cat either alive or dead, but not both alive and dead. Then, Schr¨odinger asked: Is the cat alive or dead? Certainly, the two questions both show that the superposition can be also happen in the macro world. Then, what is the reality of a thing and where do the complex systems come from? The word reality is the state of things as they actually exist, including everything that is and has been, whether or not it is observable or comprehensible. Can one really hold on the reality of things? Usually, a thing T is multilateral or complex, and so to hold on its reality is difficult for human beings, where the world complex implies the cognitive system on a thing T is complex, i.e., a system composes of many components which maybe interact with each other. A typical example for explaining the complex of cognitive system on a thing is the well-known fable “the blind men with an elephant ”. In this fable, a group of blind men heard that a strange animal, called an elephant had been brought to the town but none of them were aware of its shape and form. “We must inspect and know it by touch of which we are capable ”. The first person hand landed on the trunk, said: “this being is like a thick snake ”. The 2nd one whose hand reached its ear, claimed it like a kind of fan. The 3rd person hand was upon its leg, said the elephant is a pillar like a tree-trunk. The 4th man hand upon its side said: “the elephant is a wall ”. The 5th felt its tail, 2

described it as a rope and the last felt its tusk, stated the elephant is that being hard, smooth and like a spear. They then entered into an endless argument! “All of you are right” ! A wise man explained to them: “why are you telling it differently is because each one of you touched the different part of the elephant ”.

Fig.2 What is the philosophical meaning of this fable to human beings? It lies in that the situation of human beings hold on the reality of things is analogous to these blind men, i.e., a complex system. Usually, the reality of thing T identified with known characters on it at one time. For example, let νi , i ≥ 1 be unknown and µ1 , µ2 , · · · , µn known characters at time t. Then, the reality of thing T should be understood by

T =

n [

{µi }

i=1

[

 

[

k≥1



{νk } ,

i.e., a Smarandache multispace in logic with an approximation T ◦ =

(1.1)

n S

{µi } at time t ([22]),

i=1

which also implies that the cognition on the reality of a thing T is only an approximation, and also the complex, i.e., the reality of a thing T is nothing else but a complex system. Einstein once said the reality of things with that of mathematics: “As far as the laws of mathematics refer to reality, they are not certain; and as far as they are certain, they do not refer to reality ”. Why did Einstein say these words? Because we have no a mathematical subfield applicable to complex system, i.e., the reality of thing T, and generally, we get a contradictory system in mathematics. The main purpose of this lecture is applying the contradictory universality and the existence − → of universal connections between things T in the world, i.e., a topological graph G underlying o n− → − → its parts in philosophy to establish a global mathematics over a graph family G 1 , G 2 , · · · for characterizing dynamic behaviors of a system on the time t, i.e., complex flows such as those of to extend Banach or o Hilbert spaces to Banach or Hilbert continuity flow spaces over topological n− → − → graphs G 1 , G 2 , · · · to establish the global differential theory on complex flows and how to characterize the global dynamic behaviors of complex systems by the global differential theory over graphs. These results can be also applied to complex networks and analyze their dynamic 3

behaviors particularly, , for instance the synchronization of complex networks independent on − → graphs by applying global differential on the complex flows G L . For terminologies and notations not mentioned here, we follow references [1] for mechanics, [4]] for complex network, [5] for functional analysis, [7] for combinatorial geometry, [19] for biological mathematics, [21] for elementary particles, [6] and [30] for Smarandache systems and multispaces, and all phenomenons discussed in this paper are assumed to be true in the nature.

§2. Contradictory Systems The formula (1.1) implies that one’s recognition on a thing T is usually non-completed, which is the origination of contradiction. In classical logic, a contradictory system consists of a logical incompatibility between two or more propositions, which is abandoned without discussion in classical mathematics because a mathematical system should be compatibility in logic. However, different things are contradictory in the eyes of human beings. This is the reason why classical mathematics can not provides a complete recognition on things T . Usually, a physical phenomenon of a thing T is characterized by differential equations. If there is only one cell or one bird flying in the sky such as the flying bird, its dynamic behavior can be characterized easily by a orbit in the space, i.e., a differential equation x˙ = F (t, x),

(2.1)

Fig.3 where, x˙ = dx/dt, t is the time parameter and x is the position of bird in R3 . But how can one characterize the behavior of a complex system of m cells with m ≥ 2 in Fig.3? For example, a water molecule H2 O consists of 2 hydrogen atoms and 1 oxygen atom, and we have known the behavior of a particle is characterized by the Schr¨odinger equation i~

∂ψ ~2 2 =− ∇ ψ + Uψ ∂t 2m

in quantum mechanics ([21]). 4

(2.2)

Fig.4 Can we conclude this equation is absolutely right for atoms H and O in water molecule H2 O? Certainly not because equation (2.2) is always established with an additional assumption, i.e., the geometry on a particle P is a point in classical mechanics or a field in quantum mechanics. In this case, if equation (2.2) is true for toms H and O, we get three differential equations following  ~2 ∂ψO    −i~ = ∇2 ψO − V (x)ψO   ∂t 2mO   ∂ψH1 ~2 −i~ = ∇2 ψH1 − V (x)ψH1 (2.3)  ∂t 2mH1   2  ∂ψH2 ~   = ∇2 ψH2 − V (x)ψH2  −i~ ∂t 2mH2

on atoms H and O. Which is the right model on H2 O, the (2.2) or (2.3) dynamic equations? The answer is not so easy because the equation model (2.2) can only characterizes those of coherent behavior of atoms H and O in H2 O. Although equation (2.3) characterize the different behaviors of atoms H and O but it is non-solvable in mathematics ([14], [15]). Generally, when one wish to hold on the reality of a thing, i.e., a complex system, he usually get a contradictory system, which also implies that the mathematical known is not absolutely equal to the reality of a thing T. Thus, establish mathematics on non-mathematics, i.e., an envelope theory on mathematics for reality is needed ([11]). Now, are these contradictory systems meaningless for human beings? The answer is not! For example, let T1 , T2 , T3 , T4 and T1′ , T2′ , T3′ , T4′ be respectively two groups of horses running constraint with     x + y = 2 x=y          x + y = −2  x+y =4 (LES4N ) (LES4S )     x=2 x − y = −2         x−y = 2 y=2

on the earth. It is clear that (LES4N ) is non-solvable because x + y = −2 is contradictious to x + y = 2, and so that for equations x − y = −2 and x − y = 2. But system (LES4S ) is solvable with x = 2 and y = 2. Can we conclude that things T1′ , T2′ , T3′ , T4′ are x = 2, y = 2 and T1 , T2 , T3 , T4 are nothing? Certainly not because all of them are horses running on the earth, 5

Fig.5 and their solvability only implies the orbits intersection in R2 such as those shown in Fig.6. y

x=y

x=2 A -

O C

x − y = −2 x−y =2

y=2

x+y =2

x + y = −2

x+y =4 - x

(LES4N )

(LES4S ) Fig.6

Denoted by La,b,c = {(x, y)|ax+by = c, ab 6= 0} be points in R2 . We are easily know the beS haviors of horses T1 , T2 , T3 , T4 and T1′ , T2′ , T3′ , T4′ are nothings else but the unions L1,−1,0 L1,1,4 S S S S S L1,0,2 L0,1,2 and L1,1,2 L1,1,−2 L1,−1,−2 L1,−1,2 , i.e., Smarandache multispaces, respectively. Generally, let F1 , F2 , · · · , Fm be m mappings holding in conditions of the implicit mapping theorem and let SFi ⊂ Rn be a manifold such that Fi : SFi → 0 for integers 1 ≤ i ≤ m. Consider the equations   F1 (x1 , x2 , · · · , xn ) = 0      F (x , x , · · · , x ) = 0 2 1 2 n (2.4)  .......................      Fm (x1 , x2 , · · · , xn ) = 0 in Euclidean space Rn , n ≥ 1. Geometrically, the system (2.4) is non-solvable or not dependent m \ on SFi = ∅ or 6= ∅. i=1

Definition 2.1 A G-solution of system (2.4) is a labeling graph GL defined by V (G) = {SFi , 1 ≤ i ≤ n}; T E(G) = (SFi , SFj ) if SFi SFj 6= ∅ for integers 1 ≤ i, j ≤ n with a labeling T L : SFi → SFi , (SFi , SFj ) → SFi SFj .

For example, the G-solutions of (LES4N ) and (LES4S ) are respectively labeling graphs C4L and K4L shown in Fig.7. L1,−1,−2

L1,0,2

L1,1,2

P D C

L1,1,−2

L1,−1,0

L1,1,4

L1,−1,2

C4L

P K4L

L0,1,2

Fig.7 Example 2.2 Let (LDES61 ) be a system of linear homogeneous differential equations   x ¨ − 3x˙ + 2x = 0       x ¨ − 5x˙ + 6x = 0      x ¨ − 7x˙ + 12x = 0   x ¨ − 9x˙ + 20x = 0      x ¨ − 11x˙ + 30x = 0      x ¨ − 7x˙ + 6x = 0

(1) (2) (3) (4) (5) (6)

d2 x dx and x˙ = . Clearly, (LDES61 ) is a non-solvable system with solution bases 2 dt dt {et , e2t }, {e2t , e3t }, {e3t , e4t }, {e4t , e5t }, {e5t , e6t }, {e6t , et } respectively on equations (1) − (6) and G-solution shown in Fig.8,

t 2t 2t 2t 3t e ,e e e ,e where x ¨=

6t e ,e

het i t

e6t

5t 6t e ,e

5t e

4t e ,e

3t e

3t 4t e ,e

4t e 5t

Fig.8 where h∆i denotes the linear space generalized by elements in ∆. A more interesting application of the G-solution is it can be applied to characterizing the global stability of differential equations (2.4), even it is non-solvable. See [8-12] for details. 7

§3. Complex Flows 3.1 Complex Flows Holding on the reality of things, i.e., complex systems enables us to present an element called complex flow in mathematics on oriental philosophy, i.e.,there always exist the universal contradiction and connection between things in the world. Then, what is a complex flow ? what is its role in understanding things in the world ? Definition 3.1 A continuity flow

− → − → G ; L, A is an oriented embedded graph G in a topological

space S associated with a mapping L : v → L(v), (v, u) → L(v, u), 2 end-operators A+ vu : + A+ uv (u, v) on a Banach space B over a field F L(v, u) → LAvu (v, u) and A+ : L(u, v) → L uv A+ vu

A+ uv

L(v, u) -

L(v)

L(u)

Fig.9

− → A+ vu (v, u) for ∀(v, u) ∈ E with L(v, u) = −L(u, v) and A+ G holding with vu (−L(v, u)) = −L continuity equation − X + → LAvu (v, u) = L(v) f or ∀v ∈ V G u∈NG (v)

such as those shown for vertex v in Fig.10 following

L(u1 )

L(v, u4 )

L(u1 , v)

u4 A1 L(u2 , v)

L(u2 )

L(v)

L(v, u5 )

L(u5 )

u5 L(v, u6 ) -

L(u3 , v) L(u3 )

L(u4 )

Fig.10

L(u6 )

with a continuity equation LA1 (v, u1 ) + LA2 (v, u2 ) + LA3 (v, u3 ) − LA4 (v, u4 ) − LA5 (v, u5 ) − LA6 (v, u6 ) = L(v), where L(v) is the surplus flow on vertex v. 8

Particularly, we have known continuity flows following: âˆ’ âˆ’ â†’ â†’ (1) L(v) = xË™ v , v âˆˆ V G . In this case, G ; L, A is said to be a complex flow, discussed in this lecture; âˆ’ âˆ’ â†’ â†’ (2) For v âˆˆ V G , xv is a constant vv dependent on v. In this case, G ; L, A is said to be an action flow, which was discussed extensively in [16] with applications [14]-[15], [17] to elementary physics and biological systems; âˆ’ â†’ (3) L(v) =constant independent on v, v âˆˆ V G , which is a special case of action flow âˆ’ â†’ called A0 G -flow and shown can be applied to synchronization of system in this lecture; âˆ’ â†’ âˆ’ â†’ (4) If A = 1V , G ; L, A is said to be a G -flow, which was discussed in [13]; âˆ’ â†’ (5) If A = 1V and V is a number field Z or R, G ; L, A is said to be a complex network, which was already discussed extensively in publications, for examples, [2]-[4] and [20]. âˆ‚ and For example, let the L : (v, u) â†’ L(v, u) âˆˆ Rn Ã— R+ with end-operators A+ vu = avu âˆ‚t âˆ’ â†’ avu : Rn â†’ R for any edge (v, u) âˆˆ E G in Fig.11 following. u

Fig.11 Then the conservation laws are partial differential equations ï£± âˆ‚L(t, u)1 âˆ‚L(t, u)2 âˆ‚L(u, v) ï£´ ï£´ 1 2 a + a = auv ï£´ tu tu ï£´ âˆ‚t âˆ‚t âˆ‚t ï£´ ï£´ 1 2 ï£´ ï£´ ï£² auv âˆ‚L(u, v) = avw1 âˆ‚L(v, w) + avw2 âˆ‚L(v, w) + avt âˆ‚L(v, t) âˆ‚t âˆ‚t âˆ‚t âˆ‚t 1 2 ï£´ ï£´ avw1 âˆ‚L(v, w) + avw2 âˆ‚L(v, w) = awt âˆ‚L(w, t) ï£´ ï£´ ï£´ âˆ‚t âˆ‚t âˆ‚t ï£´ 1 2 ï£´ ï£´ ï£³ a âˆ‚L(w, t) + a âˆ‚L(v, t) = a 1 âˆ‚L(t, u) + a 2 âˆ‚L(t, u) wt vt tu tu âˆ‚t âˆ‚t âˆ‚t âˆ‚t

which maybe solvable or not but characterize the behavior of things. 3.2 Extended Linear Space

n âˆ’ S âˆ’ â†’ âˆ’ âˆ’ â†’ â†’ âˆ’ â†’ âˆ’ â†’ â†’ âˆ’ â†’ Let G , G 1 , G 2 , Â· Â· Â· , G n be oriented graphs embedded in S with G = G i , i.e., each G i i=1 âˆ’ â†’ be a subgraph of G for integers 1 â‰¤ i â‰¤ n. In this case, these is naturally an embedding âˆ’ â†’ âˆ’ â†’ âˆ’ â†’ 1 âˆ’ â†’L2 âˆ’ â†’Ln Î¹ : G i â†’ G . Can we construct linear space by reviewing continuity flows G L 1 , G2 , Â· Â· Â· , Gn not only labeling graphs but mathematical elements? The answer is yes! âˆ’ â†’ Let V be a linear space over a field F . A vector labeling L : G â†’ V is a mapping with

− → − → L(v), L(e) ∈ V for ∀v ∈ V ( G ), e ∈ E( G ). Define

and

− →L1 − → 2 − → − → L1 [ − → \− → L1 +L2 [ − → − → L2 G1 + GL G1 G2 G2 \ G1 2 = G1 \ G2

(3.1)

− → − → λ · G L = G λ·L

(3.2) − →L → − L1 → − L2 for ∀λ ∈ F . Clearly, if G , G 1 , G 2 are continuity flows with linear end-operators A+ vu and − → − →L1 − →L2 − →L + Auv for ∀(v, u) ∈ E G , G 1 + G 2 and λ · G are continuity flows also. If we consider each − →Lb → − − → → − → → b − b − continuity flow G L i a continuity subflow of G , where L : G i = L( G i ) but L : G \ G i → 0 − → for integers 1 ≤ i ≤ n, and define O : G → 0, then we get the following result. − → + Theorem 3.1([18]) If A+ vu and Auv are linear end-operators for ∀(v, u) ∈ E G , all con− → − → − → tinuity flows on oriented graphs G 1 , G 2 , · · · , G n naturally form a linear space, denoted by D E V − → G i, 1 ≤ i ≤ n ; +, · over a field F under operations (3.1) and (3.2). Particularly, for action flows, we get the following result. − → G ; L, A with linear end-operators A ∈

Theorem 3.2([13], [16]) Let G be all action flows O(V ). Then

dimG = (dimO(V ) × dimV )

→ − β( G)

if both V and O(V ) are finite. Otherwise, dimG is infinite. − → Particularly, if operators A ∈ V ∗ , the dual space of V on graph G , then → − 2β( G )

dimG = (dimV )

− − →

− →

− → → where β G = ε G − G + 1 is the Betti number of G .

− → 1 − →L2 − → 1 − → 1 − →L2 − → 2 − → − → Notice that G L 6= G L or G L 6= G L 1 + G2 1 + G2 2 if and only if G 1 6 G 2 with − → − → − → − → 1 − → − → L1 : G 1 \ G 2 6→ 0 or if G 2 6 G 1 with L2 : G 2 \ G 1 6→ 0, and generally, we say a continuity − → 1 − →L2 − →Ln flow family { G L 1 , G 2 , · · · , G n } is linear irreducible if for any integer i, [− − → → G i 6 Gl l6=i

with

− → [− → Li : G i \ G l 6→ 0, l6=i

where 1 ≤ i ≤ n. We know the following result on linear generated sets. n− → 1 − →L2 − →Ln o Theorem 3.3([18]) Let V be a linear space over a field F and let G L be 1 , G2 , · · · , Gn − → an linear irreducible family, → V for integers 1o ≤ i ≤ n with linear operators A+ vu , Li : Gni− − → → →L2 − →Ln L1 − + Auv for ∀(v, u) ∈ E G . Then, G 1 , G 2 , · · · , G n is an independent generated set of 10

D− EV → G i , 1 ≤ i ≤ n , called basis, i.e.,

D− EV → dim G i , 1 ≤ i ≤ n = n.

3.3 Extended Commutative Rings Furthermore, if V is a commutative ring (R; +, ·), can we extend it over oriented graph family − → − → − → { G 1 , G 2 , · · · , G n } by introducing operation “ + ” with (3.1) and operation “ · ” following: − →L1 − → 2 − → − → L1 [ − → \− → L1 ·L2 [ − → − → L2 G1 · GL G1 G2 G2 \ G1 , 2 = G1 \ G2

where L1 · L2 : x → L1 (x) · L2 (x) ? The answer is yes! We get the following result: n− → 1 − →L2 → − Ln o Theorem 3.4([18]) Let (R; +, ·) be a commutative ring and let G L be a 1 , G2 , · · · , Gn − → + linear irreducible family, Li : G i → R for integers 1 ≤ i ≤ n with linear operators Avu , A+ uv − D− ER → → for ∀(v, u) ∈ E G . Then, G i, 1 ≤ i ≤ n ; +, · is a commutative ring. 3.4 Banach or Hilbert Space − → We have shown that G V is a Banach space, and furthermore, Hilbert space if V is a Banach − → or Hilbert space for an oriented graph G embedded in topological space S in [13] and [16]. D− EV − → 1 − →L2 → − Ln → Generally, let { G L G i , 1 ≤ i ≤ n , where V is a Banach 1 , G 2 , · · · , G n } be a basis of D− E → space with a norm k · k. Can we extend Banach space V over G i , 1 ≤ i ≤ n ? And similarly, E D− → can we extend Hilbert space V over G i , 1 ≤ i ≤ n ? The answer is yes! EV D− → − → For ∀ G L ∈ G i , 1 ≤ i ≤ n , define − →L G =

kL(e)k

→ − e∈E G

or D− →L1 − → 2E G1 , GL 2

→ − → − e∈E G 1 \ G 2

hL1 (e), L1 (e)i

→ − T→ − e∈E G 1 G2

hL1 (e), L2 (e)i +

→ − → − e∈E G 2 \ G 1

hL2 (e), L2 (e)i .

(2.10)

Then we are easily know also that − → → − − → Theorem 3.5([18]) Let G 1 , G 2 , · · · , G n be oriented graphs embedded in a space S and V D− EV → + a Banach space over a field F . Then G i , 1 ≤ i ≤ n with linear operators A+ vu , Auv for − D− EV → → ∀(v, u) ∈ E G is a Banach space, and furthermore, if V is a Hilbert space, G i , 1 ≤ i ≤ n 11

is a Hilbert space too. Therefore, we can consider calculus and differentials on Hilbert space

D− EV → G i, 1 ≤ i ≤ n .

Now, if L is kth differentiable to t on a domain D ⊂ R, where k ≥ 1 and we define − → dGL − → dL = G dt dt

and

− →L − → G dt = G 0

Ldt

Then, what will happens? We can generalize Taylor formula on differentiable functions in D− EV → G i, 1 ≤ i ≤ n following. D− ER×R − → → Theorem 3.6(Taylor)([18]) Let G L ∈ G i , 1 ≤ i ≤ n and there exist kth order derivative − → + of L to t on a domain D ⊂ R, where k ≥ 1. If A+ vu , Auv are linear for ∀(v, u) ∈ E G , then n

k − →L − → →L′ (t0 ) →L(k) (t0 ) → t − t0 − (t − t0 ) − −k − G + ··· + G + o (t − t0 ) G , G = G L(t0 ) + 1! k! → −k − b of L that for ∀t0 ∈ D, where o (t − t0 ) G denotes such an infinitesimal term L

lim

t→t0

b u) L(v,

(t − t0 )k

f or

− → ∀(v, u) ∈ E G .

− →LC Particularly, if L(v, u) = f (t)c with LC : vu , where cvu is a constant, denoted by f (t) G − → (v, u) → cvu for ∀(v, u) ∈ E G and f (t) = f (t0 ) +

then

(t − t0 ) ′ (t − t0 )2 ′′ (t − t0 )k (k) f (t0 ) + f (t0 ) + · · · + f (t0 ) + o (t − t0 )k , 1! 2! k!

− → − → f (t) G LC = f (t) · G LC .

− → This formula for continuity flow G L enables one to find interesting results and formulas − → − → on G L by f t G such as those of the following. Corollary 3.7 Let f (t) be a k differentiable function to t on a domain D ⊂ R with 0 ∈ D and − → − → − → + f (0 G) = f (0) G . If A+ vu , Auv are linear for ∀(v, u) ∈ E G , then − − → → f (t) G = f t G .

For examples, → − → t2 − → → − → − → t− tk − et G = et G = G + G + G + · · · + G + · · · 1! 2! k!

and for a real number α if |t| < 1, − → − → α − → αt − → → α(α − 1) · · · (α − n + 1)tn − G + tG =G+ G + ···+ G + ··· . 1! n! §4. Synchronization Independent on Graphs How can we characterize the behavior of a self-adaptive system with cells m ≥ 2, for instance a flock of m birds ? A natural way for characterizing the behavior of m birds is to collect all dynamic equations of cells, i.e.,   x˙1 = F1 (t, x1 )      x˙ = F (t, x ) 2 2 2 (4.1)  ···············      ˙ xm = Fm (t, xm )

to characterize the global behavior of the system.

However, birds or generally, cells in a self-adaptive system are interacted each other. The system (1.2) only is a collection of equation of each cell, not a global characterizing of the biological system in space. Including the interaction of cells enables one to apply m geometrical points in R3 and characterizing the system by a system of differential equations following  P  x˙1 = F1 (t, x1 ) + Hj (xj → x1 )    j6=1   P    x˙2 = F2 (t, x2 ) + Hj (xj → x2 ) j6=2

   ····································    P   Hj (xj → xm )  x˙m = Fm (t, xm ) +

(4.2)

j6=m

where Fi : R3 → R3 is generally a nonlinear function characterizing the external appearance of ith cell and Hj (xj → xi ) is the action strength of the jth cell to the ith cell in this system for integers 1 ≤ i, j ≤ m. Then, what is the synchronization of a self-adaptive system ? The synchronization characterizes the behavior of a self-adaptive system from disorderly to orderly, such as those of birds flock or fishes shoal. By system (4.2) of differential equations, the synchronization nature is formally defined following. Definition 4.1([4]) The system (4.2) is said to be complete synchronization if lim kxi (t) − xj (t)k = 0

t→∞

for all integers i, j = 1, 2 · · · , m, where k · k is the Euclidean norm. In the past decades, many researchers discussed the synchronization of (4.2) in case of 13

F1 = F2 = · · · = Fm and Hj = H, i.e., a network of m identical nodes with a constant coupling c and action strength H(xj ) of node xj to xi for i, j = 1, 2, · · · , m such as those shown in the following model ([2-4, 20])  m P   x˙1 = f (x1 ) + c a1j H(xj )    j=1   m  P   x˙2 = f (x2 ) + c a2j H(xj ) j=1

   ···························    m  P   amj H(xj )  x˙m = f (xm ) + c

(4.3)

j=1

T (1) (2) (n) where xi = xi , xi , · · · , xi ∈ Rn is the state vector, f is generally a nonlinear function satisfying a Lipschitz condition, H : Rn → Rn is the inner action and A = [aij ]m×m is the outer coupling matrix defined by aij = aji = 1 if there is a connection between nodes i and j, i 6= j, otherwise, aij = aji = 0 and the diagonal elements with i = j are defined by aii = −

j=1,j6=i

aij = −

aji = −ki ,

i = 1, 2, · · · , m

j=1,j6=i

where ki is the degree of node i. Hence, the matrix A is actually the negative Laplacian matrix. Today, we have known the synchronization of (4.3) is so dependent on eigenvalues λ2 and λm of matrix A ([2-4, 20]), which classified the regions leading to the synchronization of (4.3), called synchronized region into 4 cases following by a master stability function ([20]): Type I. Synchronized region is (α1 , ∞). In this type, the synchronization of (4.3) is determined by λ2 , i.e., if cλ2 > α1 , the system (4.3) is synchronized. Type II. Synchronized region is (α2 , α3 ) ⊂ (0, ∞). In this type, the synchronization of α3 α2 (4.3) is determined by λ2 and λm , i.e., if <c< , the system (4.3) is synchronized. λ2 λm Type III. Synchronized region is the union of several intervals of (0, ∞), for instance S (α2 , α3 ) (α4 , α5 ) α6 , ∞). Type IV. Synchronized region does not exist.

But, the criterions I-IV were so strange that the synchronization is a global behavior of individuals in a self-adaptive system and can not be completely dependent on its underlying graph or in other words, the eigenvalues of matrix A. However, they appears because of one’s assumptions on system (4.3), i.e., the synchronization of a self-adaptive system should be independent on the underlying structure of individuals in general. Can we view a self-adaptive system as a mathematical element and characterize the synchronization of system ? The answer − → is positive, i.e., by complex flows G L ! − → Notice that the synchronization state of a complex flow G L is nothing else but a non-zero − → A0 flows, i.e., L(v) = v 6= 0 for ∀v ∈ V G . − → − → − → − → − → Example 4.2 Let G = C n or P n for an integer n ≥ 1. If there is an A0 G -flow on C n such 14

as those shown in Fig.12. v1

fn 6

vi+1

Fig.12 We are easily know that f1 âˆ’ fn = f2 âˆ’ f1 = f3 âˆ’ f2 = Â· Â· Â· = fi+1 âˆ’ fi = Â· Â· Â· = fn âˆ’ fnâˆ’1 by the definition of A0 -flow, which only have solutions f1 = f2 = Â· Â· Â· = fn . Thus, it is a zero A0 flows. âˆ’ â†’ âˆ’ â†’ Similarly, if there is an A0 G -flow on P n such as those shown in Fig.13.

fiâˆ’1

viâˆ’1

fi+1

fnâˆ’1

vi+1

vnâˆ’1

Fig.13 We are easily know that âˆ’f1 = f2 âˆ’ f1 = Â· Â· Â· = fi âˆ’ fiâˆ’1 = Â· Â· Â· = fnâˆ’1 âˆ’ fnâˆ’2 = fnâˆ’1 by the definition of A0 flow, which only have solutions f1 = f2 = Â· Â· Â· = fn = 0. Thus, it is a zero A0 flows also. Dâˆ’ ERÃ—Rn âˆ’ â†’ â†’ A complex A0 flow G L exists in G i , 1 â‰¤ i â‰¤ n if and only if L(v) = F (t, x) for âˆ’ â†’ âˆ’ â†’ âˆ€v âˆˆ V G , where F (t, x) is independent on the interaction in G L . i.e., the system of continuity equations âˆ’ X + â†’ LAvu (v, u) = F (t, x), âˆ€v âˆˆ V G (4.4) uâˆˆNG (v)

with the same solvable differential equation dxv = F (t, x) dt âˆ’ â†’ characterizing the behavior of variables on v âˆˆ V G , which is homogenous.Thus, we know 15

the following result by definition. Dâˆ’ ERÃ—Rn âˆ’ â†’ â†’ Theorem 4.3 A complex A0 flow G L exists in Hilbert space G i , 1 â‰¤ i â‰¤ n if and only if the ordinary differential equation dx = F (t, x) (4.5) dt Dâˆ’ ERÃ—Rn â†’ is solvable in Hilbert space G i , 1 â‰¤ i â‰¤ n . Such a solution is usually called a multispace solution of (4.5). Dâˆ’ EV âˆ’ â†’ âˆ’ â†’ 1 â†’ Definition 4.4 Let G L , G L âˆˆ G , 1 â‰¤ i â‰¤ n with L, L1 dependent on a variable t âˆˆ i 1 âˆ’ â†’ [a, b] âŠ‚ (âˆ’âˆž, +âˆž) and linear continuous end-operators A+ vu for âˆ€(v, u) âˆˆ E G . For t0 âˆˆ [a, b] and any number Îµ > 0, if there is always a number Î´(Îµ) such that if |t âˆ’ t0 | â‰¤ Î´(Îµ) â†’L1 âˆ’ â†’L âˆ’ â†’ 1 âˆ’ â†’ âˆ’ then G 1 âˆ’ G < Îµ, then, G L is said to be converged to G L as t â†’ t0 , denoted by 1 âˆ’ âˆ’ â†’ 1 âˆ’ â†’ âˆ’ â†’ â†’ lim G L = G L . Particularly, if G L is a continuity flow with a constant L(v) for âˆ€v âˆˆ V G 1 tâ†’t0 âˆ’ â†’ 1 âˆ’ â†’ and t0 = +âˆž, G L 1 is said to be G -synchronized. These is a well-known result on liner operators following, which is useful to determining the synchronization of systems. Theorem 4.5([5]) Let B1 , B2 be Banach spaces over a field F with norms k Â· k1 and k Â· k2 , respectively. Then, a linear operator T : B1 â†’ B2 is continuous if and only if it is bounded, or equivalently, kT(v)k2 < +âˆž. kTk := sup kvk1 06=vâˆˆB1

cvu

According to Theorem 4.5, if A+ liner continuous operator there must be a constant vu is âˆ’ â†’ + such that kAvu k â‰¤ cvu for âˆ€(v, u) âˆˆ E G . Let cmax G1 G =

max

(v,u)âˆˆE(G1 )

c+ vu ,

max

(v,u)âˆˆE(G1 )

c+ vu .

âˆ’ â†’ 1 âˆ’ â†’L Then, we get an equivalent condition for lim G L 1 = G following. tâ†’t0

âˆ’ â†’ 1 âˆ’ â†’L Theorem 4.6 lim G L 1 = G if and only if for any number Îµ > 0 there is always a number Î´(Îµ) tâ†’t0 âˆ’ â†’ âˆ’ â†’ such that if |t âˆ’ t0 | â‰¤ Î´(Îµ) then kL1 (v, u)k < Îµ for (v, u) âˆˆ E G 1 \ G , k(L1 âˆ’ L)(v, u)k < Îµ âˆ’ âˆ’ â†’ Tâˆ’ â†’ â†’ âˆ’ â†’ âˆ’ â†’ 1 âˆ’ â†’L for (v, u) âˆˆ E G 1 G and k âˆ’ L(v, u)k < Îµ for (v, u) âˆˆ E G \ G 1 ,i.e., G L is an 1 âˆ’ G âˆ’ â†’L1 âˆ’ â†’L infinitesimal or lim G 1 âˆ’ G = O. tâ†’t0

Proof Clearly, → − L − L1 → G1 − G → → → − → − L1 − \→ − L1 −L − → − −L = G1 \ G G + G1 + G \ G1 X X A′ + A′ + + vu L vu (v, u) + L vu − LA + = (v, u) vu 1 1 u∈NG \G (v) 1

≤

c+ vu kL1 (v, u)k

u∈NG T G (v) 1

c+ vu k (L1

− L) (v, u)k +

u∈NG T G (v) 1

u∈NG \G (v) 1

u∈NG\G (v)

c+ vu k

A+ −L vu (v, u)

− L(v, u)k

u∈NG\G (v) 1

− − − → → →L → 1 − and kL(v, u)k ≥ 0 for (v, u) ∈ E G and E G 1 . If G L 1 − G < ε, we are easily − → − → max knowing that kL1 (v, u)k < cmax G1 G ε for (v, u) ∈ E G 1 \ G , k(L1 − L)(v, u)k < cG1 G ε for − → T− → − → − → (v, u) ∈ E G 1 G and k − L(v, u)k < cmax G1 G ε for (v, u) ∈ E G \ G 1 . − → − → Conversely, if kL1 (v, u)k < ε for (v, u) ∈ E G 1 \ G , k(L1 − L)(v, u)k < ε for (v, u) ∈ − − → T− → → − → E G 1 G and k − L(v, u)k < ε for (v, u) ∈ E G \ G 1 , we easily find that − → →L1 − G 1 − G L

u∈NG1 \G (v)

u∈NG\G1 (v)

≤

′+ A vu L1 (v, u) +

A+ −L vu (v, u)

′+ A vu A+ − Lvuvu (v, u) L1

c+ vu k (L1 − L) (v, u)k

u∈NG1 T G (v)

c+ vu kL1 (v, u)k +

u∈NG1 \G (v)

u∈NG1

c+ vu k − L(v, u)k

G (v)

u∈NG\G1 (v)

−

− →

→

max →

→

max

→ −

→ \ −

→ −

→ [ − < G 1 \ G cmax G cG1 G ε + G \ G 1 cmax G cG1 G ε. G1 G ε + G 1 G1 G ε = G 1

This completes the proof.

An application of Theorem 4.6 enables us to get a result on synchronization of complex flows following, which is independent on the underlying structure of cells of a self-adaptive system. − − → → Theorem 4.7 A complex flow G L with linear continuous end-operator A+ vu for ∀(v, u) ∈ E G − → is G -synchronized if and only if for any number ε > 0 if t ≥ N (ε) then kL(v) − L(u)k < ε for − → ∀v, u ∈ V G , i.e., flows on vertex are synchronized.

− → − → 0 Proof By definition, if G L is synchronized, there must be a non-zero A0 flow G L 0 and − → − → L L0 a number N (ε) such that G − G 0 < ε if t ≥ N (ε), which implies that kL(v, u)k < ε − − → − → →T− → for u ∈ V G \ G 0 , k(L − L0 )(v, u)k < ε for u ∈ V G G 0 and k − L0 (v, u)k < ε for − → − → u ∈ V G 0 \ G by Theorem 4.6. 17

Therefore, kL(v)k =

u∈NG (v)

for ∀v ∈ V

+ Avu (v, u) ≤ L

max c+ vu kL(v, u)k ≤ |NG (v)|cGG0 ε

u∈NG (v)

− → G by applying Theorem 4.6. Thus,

kL(v) − L(u)k ≤ kL(v)k + kL(u)k ≤ (|NG (v)| + |NG (u)|)cmax GG0 ε < ε

− → for ∀v, u ∈ V G if t ≥ N

ε . cmax GG0 (|NG (v)| + |NG (u)|)

Conversely, if there is a number ε > 0 such that kL(v) − L(u)k < ε if t ≥ N (ε) for − → − → ∀v, u ∈ V G , we are easily know that lim L(v) = lim L(u) = v for ∀v, u ∈ V G . Let t→∞ t→∞ − → − → − → lim G L = G L0 . Then, G L0 is a non-zero A0 flow by definition and follows that

t→∞

− → →L − G − G L0 =

u∈NG (v)

≤

+ Avu L (v, u)

→ − (v,u)∈E G

1 2

v∈V



if t ≥ N 



→ − G

cmax G k(L − L0 )(v, u)k

−

→

cmax k(L − L0 ) (v)k ≤ G cmax G G ε < ε

, i.e., an infinitesimal which completes the proof.

→

− cmax

− → − →L1 − → − → Denoted by △ = G 1 − G L in Definition 4.1. Then, △ is an infinitesimal by Theorem − → 4.5, denoted by o t G . We therefore know a conclusion following by Theorem 4.7, which − → completely changed the notion that synchronization dependent on the structure of G . − → Theorem 4.8 A continuity flow G L with liner continuous end-operator A+ vu for ∀(v, u) ∈ − → − → − → 0 E G is G -synchronized if and only if there is a non-zero A0 flow G L such that 0 − − →L − → 0 −1 → G = GL + o t G , 0

− → independent on the structure of G . Notice that

 d  X dt

u∈NG (v)

for ∀v ∈ V

A+ vu



(v, u) =

u∈NG (v)

d A+ L vu (v, u) dt

− → d ln |t| G and = t−1 . We get the following result on a synchronized complex flow. dt 18

− − → → Theorem 4.9 A complex flow G L with liner continuous end-operators A+ vu for ∀(v, u) ∈ E G − → − → is G -synchronized if and only if there is a non-zero A0 flow G L0 such that − →L − → − → G = G L0 + o ln |t| G .

− → Particularly, if each A+ is a constant for ∀(v, u) ∈ E G , we get the conclusion following. vu

− − → → Corollary 4.10 A complex flow G L with A+ = c , a constant for ∀(v, u) ∈ E G is vu vu − →L0 synchronized if and only if there is a non-zero A0 flow G such that − →L − → − → G = G L0 + o ln |t| G . + For example, let A+ vi vi+1 = 1, Avi vi−1 = 2 and

f1 + 2i−1 − 1 F (t, x) fi = 2i−1 − → for integers 1 ≤ i ≤ n in Fig.12. We have known C fn with f : (vi , vi+1 ) → fi is a non-zero A0 − → flow. Construct a complex flow C L n by letting f1 + 2i−1 − 1 F (t, x) n! L : (vi , vi+1 ) → + i 2i−1 t and L∆ : (vi , vi+1 ) →

n! . ti

− − → ∆ → − → − → Notice that C L = o t−1 C n . We therefore known that the complex flow C L n n is G synchronized by Corollary 4.10. However, by the master stability functions in [20] we can only − → conclude that it is difficult to attain the synchronization for C n , n ≥ 3. §5. Conclusion The reality of a thing T is essentially a complex system, even a contradictory system in the eyes of human beings, and there are no a mathematical subfield applicable until today. Thus, a new mathematical theory should be established for holding on the reality of things in the world. For this objective, the mathematical combinatorics, i.e., mathematics over graphs and particularly, − → the mathematics on complex flows G L is a candidate because every thing T is not isolated but connected with other things in the world, and a complex system or a contradictory system in − → classical is nothing else but a mathematics over a graph G . References [1] R.Abraham and J.E.Marsden, Foundation of Mechanics (2nd edition), Addison-Wesley, Reading, Mass, 1978. 19

[2] R.Albert and A.L.Baraba´si, Topology of evolving networks: local events and universality, Physical Review E, Vol.85(24)(2000), 5234-5240. [3] A.L.Baraba´si and R. Albert, Emergence of scaling in random network, Science, Vol.286 (5439) (1999),509-520. [4] G.R.Chen, X.F.Wang and X.Li, Introduction to Complex Networks – Models, Structures and Dynamics (2 Edition), Higher Education Press, Beijing, 2015. [5] John B.Conway, A Course in Functional Analysis, Springer-Verlag New York,Inc., 1990. [6] Linfan Mao, Smarandache Multi-Space Theory, The Education Publisher Inc., USA, 2011. [7] Linfan Mao, Combinatorial Geometry with Applications to Field Theory, The Education Publisher Inc., USA, 2011. [8] Linfan Mao, Global stability of non-solvable ordinary differential equations with applications, International J.Math. Combin., Vol.1 (2013), 1-37. [9] Linfan Mao, Non-solvable equation systems with graphs embedded in Rn , Proceedings of the First International Conference on Smarandache Multispace and Multistructure, The Education Publisher Inc. July, 2013. [10] Linfan Mao, Geometry on GL -systems of homogenous polynomials, International J.Contemp. Math. Sciences, Vol.9 (2014), No.6, 287-308. [11] Linfan Mao, Mathematics on non-mathematics - A combinatorial contribution, International J.Math. Combin., Vol.3(2014), 1-34. [12] Linfan Mao, Cauchy problem on non-solvable system of first order partial differential equations with applications, Methods and Applications of Analysis, Vol.22, 2(2015), 171-200. − → [13] Linfan Mao, Extended Banach G -flow spaces on differential equations with applications, Electronic J.Mathematical Analysis and Applications, Vol.3, No.2 (2015), 59-91. − → [14] Linfan Mao, A new understanding of particles by G -flow interpretation of differential equation, Progress in Physics, Vol.11(2015), 193-201. [15] Linfan Mao, A review on natural reality with physical equation, Progress in Physics, Vol.11(2015), 276-282. [16] Linfan Mao, Mathematics with natural reality–action flows, Bull.Cal.Math.Soc., Vol.107, 6(2015), 443-474. [17] Linfan Mao, Biological n-system with global stability, Bull.Cal.Math.Soc., Vol.108, 6(2016), 403-430. [18] Linfan Mao, Hilbert flow spaces with operators over topological graphs, International J.Math. Combin., Vol.4(2017), 19-45. [19] J.D.Murray, Mathematical Biology I: An Introduction (3rd Edition), Springer-Verlag Berlin Heidelberg, 2002. [20] L.M.Pecora and T.L. Carrol, Master stability functions for synchronized coupled systems, Physical Review Letters, 80(1998), 2109-2112. [21] Quang Ho-Kim and Pham Xuan Yem, Elementary Particles and Their Interactions, SpringerVerlag Berlin Heidelberg, 1998. [22] F.Smarandache, Paradoxist Geometry, State Archives from Valcea, Rm. Valcea, Romania, 1969, and in Paradoxist Mathematics, Collected Papers (Vol. II), Kishinev University Press, Kishinev, 5-28, 1997.