International Journal of Bifurcation and Chaos

In this study, we investigate the occurrence of dragon-king extreme events in a three-dimensional autonomous Shimizu–Morioka oscillator. We observe that the bounded chaotic oscillations transition into large amplitude extreme events at a critical value of the system control parameter triggered by an interior crisis. These extreme events exhibit a unique distribution characterized by the probability distribution function. We performed laboratory experiments and conducted rigorous numerical simulations on the Shimizu–Morioka oscillator to validate our findings. The results from both approaches are in excellent agreement and confirm extreme behavior in this autonomous system. Our study represents the first comprehensive investigation of extreme events in the Shimizu–Morioka oscillator, integrating experimental observations and numerical simulations. Also, we observed the dragon-king extreme events in both experimental and numerical studies. These findings enhance our understanding of extreme events and their potential applications in chaos-based dynamical systems, contributing to advancing this field.

This study investigates a structurally unstable synchronization phenomenon observed in the two-coupled Izhikevich neuron model. As the result of varying the system parameter in the region of parameter space close to where the unstable synchronization is observed, we find significant changes in the stability of its periodic motion. We derive a discrete-time dynamical system that is equivalent to the original model and reveal that the unstable synchronization in the continuous-time dynamical system is equivalent to border-collision bifurcations in the corresponding discrete-time system. Furthermore, we propose an objective function that can be used to obtain the parameter set at which the border-collision bifurcation occurs. The proposed objective function is numerically differentiable and can be solved using Newton’s method. We numerically generate a bifurcation diagram in the parameter plane, including the border-collision bifurcation sets. In the diagram, the border-collision bifurcation sets show a novel bifurcation structure that resembles the “strike-slip fault” observed in geology. This structure implies that, before and after the border-collision bifurcation occurs, the stability of the periodic point discontinuously changes in some cases but maintains in other cases. In addition, we demonstrate that a border-collision bifurcation set successively branches at distinct points. This behavior results in a tree-like structure being observed in the border-collision bifurcation diagram; we refer to this structure as a border-collision bifurcation tree. We observe that a periodic point disappears at the border-collision bifurcation in the discrete-time dynamical system and is simultaneously replaced by another periodic point; this phenomenon corresponds to a change in the firing order in the continuous-time dynamical system.

In this paper, a type of modified dual memristive Shinriki oscillator is constructed with a flux-controlled absolute-type memristor and a voltage-controlled generic memristor, and the proposed oscillator with abundant dynamical behaviors, including the multistability and antimonotonicity, is comprehensively studied through dynamical distribution graphs, bifurcation diagrams, Lyapunov exponents and phase portraits. It is found that the passive/active state of memristor, which means different characteristics in the $v$–$i$ domain with positive and negative parameters of the elements, can affect the state of the oscillator. For example, if the memristor is active, the oscillator will change more frequently in the multistable region. Also, it is noted that, for inherent initial-related symmetry and circuit structures with duality, both phenomena have strong symmetric characteristics and opposite evolution trends modulated by values of corresponding components. Especially, the bubbles, which are symmetric about parameters with duality and own complex evolution laws, have rarely been explored in previous works. In addition, the memristive oscillator is modularized based on field programmable gate array (FPGA) technology, and the multiple coexisting attractors are captured, which verifies the accuracy of the numerical results.

This paper considers a dual-channel closed-loop supply chain (CLSC) consisting of a manufacturer who wholesales new products through the traditional retail channel and distributes remanufactured products via a direct (online) channel established by himself. Two dynamical Stackelberg game models are developed based on the assumption that the retailer is an adaptive player, and the manufacturer is a bounded rational player who may adopt a delay decision. The existence and locally asymptotic stability of the Nash equilibrium are examined. Moreover, the impacts of key parameters on the complexity characteristics of the models and the performance of chain members are studied by numerical simulation. The results reveal that the excessively fast price adjustment speeds of the manufacturer, the larger consumers’ discount perception for the remanufactured products, and the consumers’ preference for the direct channel have a strong destabilizing effect on the Nash equilibrium’s stability. Furthermore, the delay decision implemented by the manufacturer could be a stabilizing or destabilizing factor for the system. The manufacturer will tolerate a considerable profit reduction while the retailer gains more profits when the dual-channel CLSC system enters periodic cycles and chaotic motions.

In this study, we extend an SIS epidemic model by introducing a piecewise smooth incidence rate. By assuming that the demographic parameters are much smaller than the disease-related ones, the proposed model is converted to a slow–fast system. Utilizing the geometrical singular perturbation theory and entry-exit function, we prove the coexistence of two relaxation oscillations surrounding the unique positive equilibrium of the model. Numerical simulations are performed to verify our theoretical results. The phenomenon presented in this study can be a potential explanation for that several infectious diseases can re-emerge many years after being almost extinct.

The aim of modeling musculoskeletal systems is to understand the mechanisms of locomotion control in terms of neurophysiology and neuroanatomy. The complexity and unique nature of neuromuscular systems, however, make control problems in these systems very challenging due to several characteristics including speed and precision. Thus, their investigation requires the use of simple and analyzable methods. Consequently, taking into account the central pattern generator’s (CPG) function, we attempted to create a structured chaotic model of how human joints and muscles interact for the purpose of facilitating gait and rehabilitation in patients with incomplete spinal cord injury. The four muscle groups used in this model are gluteus, and hip flexor groups for flexion and extension of the hip joints as well as hamstring muscles and vasti muscles for flexion and extension of the knee joint. The results indicate that the output of the chaotic model of muscle and joint interactions in a healthy state would be chaotic, while in the incomplete spinal cord injury state, it would remain a fixed point. For model rehabilitation, afferent nerve stimulation is used in a CPG model; based on the modeling results, by applying coefficients of 1.98, 2.21, and 3.1 to the values of Ia, II, and Ib afferent nerves, the incomplete spinal cord injury model state is changed from a fix-point to periodic in a permanent fashion, suggesting locomotion with rehabilitation in our model. Based on the results obtained from the chaotic model of muscle and joint interactions as well as the comparisons made with reference papers, it can be stated that this model has acceptable output while enjoying simple computations and can predict different norms.

In this paper, a chaotic multiple-image encryption algorithm using block scrambling and dynamic DNA coding is designed. The algorithm can achieve simultaneous encryption of many different types of images. Firstly, a new block scrambling algorithm is proposed, which is divided into two scrambling modes according to the size of the initial block. And it combines with the cyclic shift operation to achieve scrambling. Secondly, dynamic DNA coding is utilized to diffuse images, which enhances the complexity and security of the proposed multiple-image encryption algorithm. Through the analysis of key space, correlation, information entropy, histogram, differential attacks and robustness, it is verified that the algorithm is safe and effective. Experimental results show that the new multiple-image encryption algorithm is suitable for multiple color and gray images encryption. The algorithm has excellent encryption performance, and can be applied in secure communication.

A liquid bath vibrating vertically can lead to the emergence of a self-propelled walking droplet on its free surface, which can exhibit chaotic motion. It is well-known that trajectories of a chaotic system are sensitive to its initial condition, known as the “butterfly-effect”, while its statistics normally remain stable to small disturbances: this type of chaos is called “normal-chaos”. However, a concept called “ultra-chaos” has been recently introduced, whose statistical features are unstable, i.e. extremely sensitive to small disturbances. Up to now, a few examples of ultra-chaos have been reported. In this paper, the influence of tiny disturbances on the motion of walking droplet is investigated. It is found that both normal-chaos and ultra-chaos exist in the motion of the walking droplet. Different from the normal-chaotic motion, even the statistical properties of the droplet’s ultra-chaotic motion are sensitive to tiny disturbances. Therefore, this illustrates once again that ultra-chaos indeed exists widely and represents a higher disorder compared with normal-chaos. The ultra-chaos as a new concept can widen our knowledge about chaos and provide us with a new point of view to study chaotic properties. It should be emphasized that, for an ultra-chaos, it is impossible to repeat any results of its physical experiments or numerical simulations even in the meaning of statistics! Unfortunately, reproducibility is a corner stone of modern science. Thus, the paradigm of modern scientific research might be invalid for an ultra-chaotic system.

In this study, we have developed a novel SIR epidemic model by incorporating fractional-order differential equations and utilizing saturated-type functions to describe both disease incidence and treatment. The intricate dynamical characteristics of the proposed model, encompassing the determination of the conditions for the existence of all possible feasible equilibria with their local and global stability criteria, are investigated thoroughly. The model undergoes backward bifurcation with respect to the parameter representing the side effects due to treatment. This phenomenon emphasizes the critical role of treatment control parameters in shaping epidemic outcomes. In addition, to understand the optimal role of the treatment in mitigating the disease prevalence and minimizing the associated cost, we investigated a fractional-order optimal control problem. To further visualize the analytical results, we have conducted simulation works considering feasible parameter values for the model. Finally, we have employed local and global sensitivity analysis techniques to identify the factors that have the greatest potential to reduce the impact of the disease.

In this paper, we consider a single-species reaction–diffusion–advection population model with nonlinear boundary condition in heterogenous space. We not only investigate the existence, nonexistence and stability of positive steady-state solutions through a linear elliptic eigenvalue problem by means of variational approach, but also verify the existence of steady-state bifurcations at zero solution through Crandall and Robinowitz bifurcation theory and discuss the stability of bifurcations, which can lead to Allee effect when the bifurcation is subcritical.

In the current study, we have formulated an $SI$ epidemic model incorporating the influence of Allee effect on the dynamics of the model system in space and time. During the spread of the disease, inhibition among the susceptible and infected individuals is seen, which is measured using the Crowley–Martin type incidence function. Extensive analysis explores the system’s stability and different bifurcation scenarios, including Hopf, saddle-node, transcritical, and Bogdanov–Takens. We also examine how these bifurcations react to parameter variations within the proposed model. The temporal model has further been extended to a spatial one to investigate the impact of the Allee effect on the formation of patterns for different values of the cross-diffusion coefficient. The well-posedness of the stationary solution and the global stability of the spatial system are derived. Also, the Turing instability, Hopf, and Turing bifurcation are calculated considering the transmission rate as critical parameter. In a heterogeneous environment, the spatial distribution of the susceptible population shows a complex structure with flat tabular surface and a few almost circular holes in surface plots for both strong and weak Allee effects. To further explore the role of Allee effect on pattern development for various cross-diffusion coefficient values, the temporal model has been spatially extended. Additionally, the transmission rate is taken into account while calculating the Turing instability and Hopf and Turing bifurcations. We ran numerical simulations to validate our analytical results for both spatial and nonspatial models. Our current model system can explain the recent occurrence of respiratory distress syndrome in endangered penguin species.

The dynamics of a delay Sel’kov–Schnakenberg reaction–diffusion system are explored. The existence and the occurrence conditions of the Turing and the Hopf bifurcations of the system are found by taking the diffusion coefficient and the time delay as the bifurcation parameters. Based on that, the existence of codimension-2 bifurcations including Turing–Turing, Hopf–Hopf and Turing–Hopf bifurcations are given. Using the center manifold theory and the normal form method, the universal unfolding of the Turing–Hopf bifurcation at the positive constant steady-state is demonstrated. According to the universal unfolding, a Turing–Hopf bifurcation diagram is shown under a set of specific parameters. Furthermore, in different parameter regions, we find the existence of the spatially inhomogeneous steady-state, the spatially homogeneous and inhomogeneous periodic solutions. Discretization of time and space visualizes these spatio-temporal solutions. In particular, near the critical point of Hopf–Hopf bifurcation, the spatially homogeneous periodic and inhomogeneous quasi-periodic solutions are found numerically.

We study discontinuous piecewise linear differential systems formed by linear centers and/or linear Hamiltonian saddles and separated by a nonregular straight line. There are two classes of limit cycles: the ones that intersect the separation line at two points and the ones that intersect the separation line in four points, named limit cycles of type ${II}_2$ and limit cycles of type ${II}_4$, respectively. We prove that the maximum numbers of limit cycles of types ${II}_2$ and ${II}_4$ are two and one, respectively. We show that all these upper bounds are reached providing explicit examples.

The neural network activation functions enable neural networks to have stronger fitting abilities and richer dynamical behaviors. In this paper, an improved fractional-order discrete tabu learning neuron (FODTLN) model map with a nonlinear periodic function as the activation function is proposed. The fixed points of the map are discussed. Then, the rich and complex dynamical behaviors of the map under different parameters and order conditions are investigated by using some common nonlinear dynamical analysis methods combined with the fractional-order approximate entropy method. Furthermore, it is found that fractional-order differential operators affect the generation of multiscrolls, and the model has infinitely many coexisting attractors obtained by changing the initial conditions. Interestingly, attractor growth and state transition are found. Finally, the map is implemented on the DSP hardware platforms to verify the realizability. The results show that the map exhibits complex and interesting dynamical behaviors. It provides a fundamental theory for the research of artificial neural networks.

In this paper, we investigate the existence of one type of homoclinic double loops (i.e. figure-eight loops) in a family of planar sector-wise linear systems with saddle–saddle dynamics. We obtain necessary and sufficient conditions for the existence of a figure-eight loop. Moreover, we prove that such systems can have simultaneously three types of invariant sets: a figure-eight loop, a homoclinic loop and three different types of periodic orbits. We also provide an example to show that a crossing limit cycle can bifurcate from this figure-eight loop.