Nonlinear analysis, circuit design, and chaos optimisation application of multiscroll chaotic attractors based on novel locally active non-polynomial memristor

This paper presents a novel local active non-polynomial memristor model capable of generating multiscroll attractors through simple voltage adjustments. Unlike traditional methods, this model extends attractors without modifying the system’s terms. The memristor model is incorporated into the Sprott-B system, producing one-dimensional, two-dimensional, and three-dimensional non-polynomial memristor multiscroll chaotic attractors (1D, 2D, 3D-NPMMSCAs). The comprehensive dynamic analysis, which includes Lyapunov exponents, bifurcation diagrams, and stability studies, reveals complex system behaviors, such as adjustable and controllable coexisting attractors. The existence of non-polynomial memristor multiscroll chaotic attractors (NPMMSCAs) is validated through circuit simulations, and their feasibility is confirmed via implementation on an STM32 microcontroller. Additionally, the 1D-NPMMSCAs-Grey Wolf Optimizer (1D-NPMMSCAs-GWO) enhances convergence speed, stability, and search capability, outperforming the standard Grey Wolf Optimizer (GWO) in CEC2005 tests. When applied to integrated circuit defect image segmentation, the 1D-NPMMSCAs-GWO combined with the Otsu algorithm improves global search efficiency, robustness, and defect detection accuracy.