Collaborative Harris hawks optimization and fuzzy neural network for dual-stage PID control in solar tracking systems: A dynamic performance analysis

To address the issues of dynamic lag and parameter conservatism in traditional proportional-integral-derivative (PID) control for solar tracking systems, a dual-stage collaborative optimization controller (HHO–FNN–PID) is proposed, integrating Harris hawks optimization (HHO) and fuzzy neural network (FNN) in two sequential phases: HHO–PID and FNN–PID. In the initial phase, HHO–PID rapidly identifies the feasible ranges of PID parameters and accelerates error convergence. But as the error falls below a predefined threshold, it is substituted by FNN–PID for parameter fine-tuning, to realize the dynamic compensation and overshoot suppression. Simulations show that HHO–FNN–PID reduces rise time by 75.5%, limits overshoot to 3.1%, and shortens disturbance recovery by 74.4% compared to traditional PID. Comparison with other controllers, it improves the response speed and stability of PID controller through the phased cooperation mechanism. This provides a valuable reference for designing high-precision control systems in photovoltaic tracking applications.