Optimal targeting of nonlinear chaotic systems using a novel evolutionary computing strategy

Abstract

Control of chaotic systems to given targets is a subject of substantial and well-developed research issue in nonlinear science, which can be formulated as a class of multi-modal constrained numerical optimization problem with multi-dimensional decision variables. This investigation elucidates the feasibility of applying a novel population-based metaheuristics labelled here as Teaching-learning-based optimization to direct the orbits of discrete chaotic dynamical systems towards the desired target region. Several consecutive control steps of small bounded perturbations are made in the Teaching-learning-based optimization strategy to direct the chaotic series towards the optimal neighborhood of the desired target rapidly, where a conventional controller is effective for chaos control. Working with the dynamics of the well-known Hénon as well as Ushio discrete chaotic systems, we assess the effectiveness and efficiency of the Teaching-learning-based optimization based optimal control technique, meanwhile the impacts of the core parameters on performances are also discussed.Furthermore, possible engineering applications of directing chaotic orbits are discussed. Keywords:Chaos control; chaotic dynamics; optimization; computational intelligence;Teaching-learning-based optimization Highlights► Control of chaotic systems is formulated as constrained optimization problem.► Teaching-learning-based optimization is to direct the chaotic series.► The efficacy of TLBO based optimal control technique have been demonstrated.

Design of meta-heuristic computing paradigms for Hammerstein identification systems in electrically stimulated muscle models

“…These all are deterministic procedures with their own advantages, applications, and limitations while the stochastic techniques [29][30][31][32][33][34] are not extensively exploited for the parameter estimation of input nonlinear Hammerstein systems as yet. Few potential applications of these methodologies based on exploration and exploitation of artificial neural networks (ANNs), differential evolution (DE) and genetic algorithms (GAs) include fractional order systems [35][36], nonlinear singularly perturb systems [37], nonlinear pantograph systems [38], nonlinear prey-predator models [39], nonlinear chaotic systems [40][41], models of nonlinear optics [42], random matrix theory based application [43], thin film flow systems [44], thermal analysis of porous fin model [45], input nonlinear control autoregressive systems [46][47], active noise control systems [48][49] and control autoregressive moving average systems [50]. Beside these recently stochastic solvers are used to address viably the optimiza tio n problems arising in various domains such as astrophysics [51][52], atomic physics [53][54], plasma physics [55][56], thermodynamics [57], mechanics [58][59], nanotechnology [60][61], electric circuits [62][63], energy [64][65], power [66]…”