Chaotic fractional discrete neural networks based on the Caputo h-difference operator: stabilization and linear control laws for synchronization

Ahlem Gasri,Adel Ouannas,Amina Aicha Khennaoui,Giuseppe Grassi,Taki-Eddine Oussaeif,Viet-Thanh Pham
DOI: https://doi.org/10.1140/epjs/s11734-022-00442-8
2022-01-29
The European Physical Journal Special Topics
Abstract:<p class="a-plus-plus">Chaotic dynamics and synchronization in fractional systems described by non-integer order difference operators have attracted increasing attention in recent years. However, very few papers have been published regarding the chaotic behaviors of fractional discrete-time neural networks and their synchronization properties. A novel chaotic <em class="a-plus-plus">h</em>-fractional discrete neural network model is presented in this paper. This model is described using the <em class="a-plus-plus">h</em>-fractional Caputo difference operator of order <span class="a-plus-plus inline-equation id-i-eq1"><span class="a-plus-plus equation-source format-t-e-x"><span class="mjpage"><svg xmlns:xlink="http://www.w3.org/1999/xlink" width="1.232ex" height="1.676ex" style="vertical-align: -0.338ex;" viewBox="0 -576.1 530.5 721.6" role="img" focusable="false" xmlns="http://www.w3.org/2000/svg"><g stroke="currentColor" fill="currentColor" stroke-width="0" transform="matrix(1 0 0 -1 0 0)"> <use xlink:href="#MJMATHI-3BD" x="0" y="0"></use></g></svg></span></span></span>. The paper also presents a novel theorem, which assures that two chaotic fractional discrete neural networks achieve synchronized dynamics using very simple linear control laws. Additionally, the chaotic dynamics of the network are stabilized at the origin via a suitable controller. Finally, simulation results are carried out to demonstrate the feasibility of the theoretical results developed herein.</p><svg xmlns="http://www.w3.org/2000/svg" style="display: none;"><defs id="MathJax_SVG_glyphs"><path stroke-width="1" id="MJMATHI-3BD" d="M74 431Q75 431 146 436T219 442Q231 442 231 434Q231 428 185 241L137 51H140L150 55Q161 59 177 67T214 86T261 119T312 165Q410 264 445 394Q458 442 496 442Q509 442 519 434T530 411Q530 390 516 352T469 262T388 162T267 70T106 5Q81 -2 71 -2Q66 -2 59 -1T51 1Q45 5 45 11Q45 13 88 188L132 364Q133 377 125 380T86 385H65Q59 391 59 393T61 412Q65 431 74 431Z"></path></defs></svg>
What problem does this paper attempt to address?