Abstract

Dynamical behaviors of the 4D hyperchaotic memristive circuit are analyzed with the system parameter. Based on the definitions of fractional-order differential and Adomian decomposition algorithm, the numerical solution of fractional-order 4D hyperchaotic memristive circuit is investigated. The distribution of stable and unstable regions of the fractional-order 4D hyperchaotic memristive circuit is determined, and dynamical characteristics are studied by phase portraits, Lyapunov exponents spectrum, and bifurcation diagram. Complexities are calculated by employing the spectral entropy (SE) algorithm and C0 algorithm. Complexity results are consistent with that of the bifurcation diagrams, and this means that complexity can also reflect the dynamic characteristics of a chaotic system. Results of this paper provide a theoretical and experimental basis for the application of fractional-order 4D hyperchaotic memristive circuit in the field of encryption and secure communication.

1. Introduction

According to the principle of completeness with variable combination, Professor Chua predicted the existence of memristor in 1971 [1]. In 1976, he expounded the character of memristor, composition principle, and applications [2]. For a long time, the existence of element which satisfied the character of memristor was not discovered, so the study of memristor did not rise to the attention of scientific community and engineering circles. In 2008, the HP laboratory reported the realization of memristor firstly [3, 4], and, since then, the memristor has attracted much attention all over the world. Memristors are often divided into charge-controlled memristor and flux-controlled memristor. Both of them are typical nonlinear elements. It is expected to be an effective memory storage device in computers. It means lower power consumption and less thermal design to deal with, and it is easy to generate a chaotic vibration signal by employing this element. On one hand, we need to prevent the harm of chaos phenomenon in the application. On the other hand, the chaotic vibration signal by employing the memristor can be applied to many fields such as secure communication and aerospace industry. So researchers began to focus on the design and realization of memristive chaotic circuit [520]. In these literatures, only one memristor was applied in an independent circuit, and the dynamic characteristics of memristive chaotic system are related to the initial state of memristor, including unique nonlinear physics phenomenon. Zhang and Deng studied double-compound synchronization of six memristor-based Lorenz systems [21]. However, this article focuses on the synchronization method of a net with four Lorenz systems, rather than dynamic behaviors of a memristive circuit. When Bao et al. and Mou et al. applied more than one memristor in a single circuit [22, 23], they found that memristors would affect each other, and the dynamical behaviors of the circuit with more than one memristor become more complex. Compared with ordinary memristive chaotic systems, the memristive hyperchaotic systems have excellent security in communication because the memristive hyperchaotic sequence has better complexity and randomicity [2426].

Fractional calculus is more than 300-year-old topic. It can describe physical phenomena more accurately than that of the integer-order calculus, so fractional calculus has attracted more attention in various areas of applications such as physics, chemistry, bioengineering, signal processing, and control system [2729]. A very important area of applications is the chaos theory [3033]. At present, there are three typical methods to solve the fractional-order nonlinear system, such as frequency domain approximation [30], predictor-corrector method (PCM) [34], and Adomian decomposition method (ADM) [35]. The calculation precision of frequency domain approximations is limited. Whether this method accurately reflects the chaos characteristics of a fractional-order nonlinear system was questioned [36, 37]. For PCM, one can obtain more accurate results, but the calculation speed is too slow, and it consumes too much computer resources. Thus, it is unsuitable for engineering practice. Compared with the two solution approaches, ADM is capable of dealing with linear and nonlinear problems in time domain [38, 39]. He et al. concluded the characteristics of this method, such as high accuracy, fast convergence, and less computer resources consumption [40]. Recently, Li et al. proposed a new 4D hyperchaotic memristive circuit [41], which possesses abundant complex dynamics. The main feature of this system is having uncountable infinite number of stable equilibria, which is significantly different from other reported chaotic systems before. However, the article mainly analyzed the equilibrium states, and the analysis of dynamics characteristics of the 4D memristive hyperchaotic system is incomplete. Its fractional-order form has not been studied by now. Thus, it makes a great sense to study the dynamics characteristics of the system and the solution and dynamics of its corresponding fractional-order case. To our knowledge, no one has studied the fractional-order memristive hyperchaotic system. So we will employ ADM algorithm to solve the fractional-order 4D hyperchaotic system.

In this paper, we focus on dynamical characteristic of the fractional-order 4D hyperchaotic memristive circuit. It is organized as follows. The dynamical characteristics of integer-order 4D hyperchaotic memristive circuit are investigated in Section 2. In Section 3, ADM is introduced briefly, and the iterative algorithm of the fractional-order 4D hyperchaotic memristive circuit is deduced. In Section 4, the distribution of stable and unstable regions of the fractional-order 4D hyperchaotic memristive circuit is determined, and the dynamical characteristics of this system are analyzed. Finally, we summarize the results and indicate future directions.

2. 4D Hyperchaotic Memristive Circuit

2.1. Model of the 4D Hyperchaotic Memristive Circuit

Figure 1 shows the 4D hyperchaotic memristive circuit model, which consists of standard integrators, standard multipliers, linear resistors, linear capacitors, and a nonlinear active memristor. We can use , , and to indicate states of voltages. For the memristive circuit, there is an input from to by a flux-controlled memristor, and it is illustrated by in Figure 1. According to Chua’s definition, a memristor is a passive two-terminal circuit element described by a nonlinear i-v characteristic as follows: or , where , , , andφare the voltage, current, charge, and flux associated with the device. is the memristance defined as follows: . is the memductance defined as follows: . Here, we focus on the flux-controlled memristive system described by the following circuit equation: , where and are two positive constant parameters. According to the volt-ampere characteristics of each element and Kirchhoff’s current and voltage law, the differential equation (1) is obtained:where is a reference resistor and is a reference capacitor; then is the physical time, where is the dimensionless time. The parameters can be taken as follows: = = = C, = = R, , , and . By employing the normalized operation, (1) becomeswhere is a positive parameter indicating the strength of the memristor. We should note here that comes from the memristor and has the same function as the physical memductance mentioned above, but it is dimensionless, which would be convenient for the following discussion.

Setting the parameters , , , , , , the initial value of (2) is (1, 0, 1, 0), and the time step is  s; we get the hyperchaotic attractor as shown in Figure 2. In this case, the Lyapunov exponents of the system are = 0.2566, = 0.0674, and = 0, = −19.2935, and the Lyapunov dimension is DL = 3.017. Obviously the first two positive Lyapunov exponents imply that the 4D memristive circuit is hyperchaotic.

2.2. Dynamic Analysis with Different System Parameter

Fix the initial value of (2) to (1, 0, 1, 0), and change the parameter from 0 to 3.2. Other parameters are the same as mentioned above. We obtain the Lyapunov exponents and its corresponding bifurcation diagram as shown in Figure 3, where the last Lyapunov exponent is not displayed because it is always a big negative number. It shows that the Lyapunov exponents spectrum and bifurcation diagram are consistent. Figure 3(b) shows the routes to chaos of the system (2), and the system transforms into two-scroll hyperchaotic attractor from one-scroll chaotic attractor. When the circuit parameter increases further, the system transforms into three-scroll hyperchaotic attractor from two-scroll hyperchaotic attractor. Obviously, there are two periodic windows in the chaotic region at about and . It indicates that system (2) has abundant dynamical behaviors.

To display its dynamics further, phase portraits of different states with different parameter are presented in Figure 4.

3. Numerical Solution of Fractional-Order 4D Hyperchaotic Memristive Circuit

3.1. Adomian Decomposition Method

For a given fractional-order differential equation , here are variables and is the Caputo derivative operator of order [42, 43]. To obtain the following initial value problem, is separated into three terms [44, 45]:

Here, L and represent linear and nonlinear items, respectively, and are constants for autonomous systems, and is a specified constant. By applying the operator to both sides of (3), the following equation is obtained [46]:

is Riemann-Liouville fractional integral operator with order . For , , , , and real constant , the fundamental properties of the integral operator are described as follows [47]:

Based on ADM, the nonlinear terms of (4) are decomposed according towhere , = ; then the nonlinear terms are expressed as

So the solution of (3) is derived from

3.2. Solution of the Fractional-Order 4D Hyperchaotic Memristive Circuit

The equation of the fractional-order 4D hyperchaotic memristive circuit iswhere , , , and are the state variables, and is the order of fractional-order differential equation, where is the memductance defined as , and , , , , , and are the system parameters. According to (5) and (8), the discrete iterative formula of the system (9) is presented bywhere is iteration step size. Г(·) is Gamma function. Considering the fast convergence of this method, we truncate the first six terms of (10) in this paper. For the computer simulation, the iteration is expressed as follows:

According to (11), the chaotic sequences of the fractional-order 4D hyperchaotic memristive circuit are obtained with initial values, , , and appropriate parameters. Then we can analyze the dynamical characteristics of the system by using the chaotic sequences.

4. Dynamical Characteristics

4.1. Stability Analysis

By setting the left-hand side of (9) to zero, we can calculate the equilibrium points of the system. Obviously, is the one equilibrium point set for the system, is a real constant. The Jacobian matrix of the system (9) at equilibrium point set is described as follows:

The characteristic polynomial of (12) is , which has root at

The stability of the equilibrium can be investigated using Theorem  1: the fractional-order system is asymptotically stable if all the eigenvalues of the Jacobian matrix satisfy the condition [30]. It determines the stable and unstable regions as shown in Figure 5.

The black region in Figure 5 is the stable region that , while light region is the unstable region that . When , it is the stability of integer-order case.

Here, we consider the simple case = = = = q, where the fractional-order system has a commensurate order. According to Theorem  2, suppose that the unstable eigenvalues of scroll saddle points are [30]. The necessary condition to exhibit the chaotic attractor of (9) is the eigenvalues remaining in the unstable region. The condition for the commensurate derivatives order is .

4.2. Phase Portraits

Setting the parameters , , , , , , and , the initial value of (9) is (1, 0, 1, 0). We get the chaotic attractor as shown in Figure 6. In this case, the Lyapunov exponents of the system are = 0.8528, = 0.3133, and = 0, = −59.55 and the Lyapunov dimension is . The fractional-order 4D hyperchaotic memristive circuit is in a hyperchaotic state as shown in Figure 6. Obviously, in this case, the largest Lyapunov exponent is much bigger than that of integer-order system.

4.3. Lyapunov Exponents Spectra and Bifurcation Diagram

According to the LE spectra calculation algorithm [48], we let , , , , , and ; the initial value of (9) is (1, 0, 1, 0). The Lyapunov exponents spectra and bifurcation diagram of the fractional-order 4D hyperchaotic memristive circuit by changing simultaneously are shown in Figure 7, where the last Lyapunov exponent is not displayed because it is always a big negative number.

All of the dynamical behaviors of the system with different parameter are summarized in Table 1. It shows that system (9) has abundant dynamical behaviors.

To display its dynamics further, phase portraits of different states with different parameter are presented in Figure 8. With the increase of the parameter , the system transforms from hyperchaotic attractor to limit cycle. Obviously, there is a periodic window in the chaotic region when = (2.0394–2.0407).

For above circuit parameters and , the Lyapunov exponents spectra and bifurcation diagram of the fractional-order 4D hyperchaotic memristive circuit by changing simultaneously are shown in Figure 9, where the last Lyapunov exponent is not displayed because it is always a big negative number. The Lyapunov exponents spectra and the bifurcation diagram match very well. Both of them show that the system has complex behaviors. Obviously, once the system is in hyperchaotic state, with the increase of the parameter q, the largest Lyapunov exponent becomes smaller and smaller until in the . So the fractional-order hyperchaotic system has a higher complexity compared to its corresponding integer-order system.

All of the dynamical behaviors of the system with different parameter are summarized in Table 2. It shows that system (9) has abundant and complex dynamical behaviors.

To further display its dynamics, - phase portraits of different states with different parameter are presented in Figure 10. The route to limit cycle from sink can be observed. With the increase of the parameter , the system transforms from chaotic attractor to hyperchaotic attractor. Obviously, there are several periodic windows in the chaotic region.

4.4. Spectral Entropy (SE) and Complexity

Complexity measure is an important reference to measure dynamics of a chaotic system. If a chaotic system is used in information security, it can reflect the security of the system to some extent. At present, there are several algorithms for measuring the complexity of chaotic sequences, including intensive statistical [49], multiscale entropy (MSE) [50], spectral entropy (SE) algorithm [51], and C0 complexity arithmetic [52]. Among them, C0 and SE complexity algorithms have less parameters, faster calculation speed, and higher accuracy. So we employ the SE algorithm to measure the complexity of the fractional-order 4D hyperchaotic system.

Let , , , , , and ; the initial value of (9) is (1, 0, 1, 0). According to the arithmetic spectral entropy (SE) and C0 complexity arithmetic, the SE and C0 complexity of the fractional-order 4D hyperchaotic memristive circuit by changing and simultaneously are shown in Figure 11.

Figures 7, 9, and 11 show that the Lyapunov exponents spectrum, bifurcation diagram, and complexity are consistent. Figures 11(a) and 11(b) show that the change trends of SE complexity and C0 complexity are consistent, but the values are different. It is mainly because of the differences defined by the SE and C0 definitions, which mean that complexity can also reflect the dynamic characteristics of a chaotic system.

5. Conclusions

In this paper, the dynamical characteristics of integer-order 4D hyperchaotic memristive circuit are analyzed firstly. Based on Adomian decomposition method (ADM), the numerical solution of a fractional-order 4D hyperchaotic memristive system is investigated and all parameters values of the system are determined. The equilibrium points are obtained, and the distribution of stable and unstable regions of the system is determined. Dynamical characteristics are studied by phase portraits, Lyapunov exponents spectrum, bifurcation diagram, SE complexity, and C0 complexity. We found that the dynamical characteristics of fractional-order 4D hyperchaotic memristive system are more complex than that of the integer-order system, and the SE and C0 complexity could reflect the dynamical characteristics quite well. According to the variation of the system dynamics, it is found that the fractional-order hyperchaotic system has better application prospect than its corresponding integer-order chaotic system in the field of chaotic secure communication. Next, we will try to study its implementation in hardware.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.

Acknowledgments

This work supported by the National Natural Science Foundation of China (Grant nos. 61161006 and 61573383) and Provincial Natural Science Foundation of Liaoning (Grant no. 20170540060).