Neighborhood-perturbed and nonlinearly enhanced whale optimization for high-precision trajectory tracking and stability control in four-steering-wheel AGVs

This study proposes an integrated control strategy combining a nonlinear multi-swarm improved spiral whale optimization algorithm (nMISWOA) with PID control to enhance trajectory tracking and stability for heavy-duty four-steering-wheel AGVs in complex industrial environments. First, significant enhancements were introduced to the conventional whale optimization algorithm through the incorporation of cubic chaotic mapping, adaptive random factors, and neighborhood perturbation mechanisms, effectively improving initial population diversity, exploration–exploitation balance, and local convergence performance. Second, a dual-layer optimization control architecture based on nMISWOA-PID was constructed in conjunction with an enhanced nonlinear dynamic model of four-steering-wheel AGVs. This framework enables adaptive parameter tuning and dynamic compensation for the controller. Simulations on CEC 2022 benchmarks show nMISWOA outperforms five peers (WOA, GGO, GWO, MFO, SSA) in convergence accuracy, variance, and optimal solutions, particularly in 20-dimensional problems. The nMISWOA-PID controller reduced overshoot by over 30% in step response tests compared to other algorithm-based controllers. Physical experiments demonstrated over 20% lower tracking error and 12.87% higher positioning accuracy versus industry standards. The proposed strategy provides theoretical and practical foundations for deploying high-load AGVs in complex scenarios.