Optimization of RBF Neural Network PID Controller for Reactor Based on Improved WOA Algorithm

In the field of industrial control, traditional PID control algorithms often face challenges in meeting the requirements of systems characterized by significant time delays and nonlinear behaviors. To address this issue, this paper presents a method for optimizing PID controller parameters that integrates an enhanced whale optimization algorithm (WOA) with a radial basis function (RBF) neural network. The improved WOA effectively optimizes the initial values of the center position, weight, and width of the RBF neural network through its synergy with the RBF architecture. As a result, the RBF neural network generates PID controller parameters that enable faster and more accurate convergence. Specifically, the enhanced WOA incorporates a logistic chaotic map to increase diversity within the initialization population; it employs a nonlinear piecewise convergence factor to accelerate convergence speed while minimizing oscillatory behavior; furthermore, an adaptive inertia weight factor is introduced to refine the position update process further. The reactor used in chemical reactions serves as our control object for conducting experiments on PID controller parameter optimization. Experimental results indicate that compared to both traditional PID algorithms and those optimized using WOA alone, the PID controller designed via our improved WOA integrated with RBF neural networks demonstrates superior antiinterference capabilities. This result confirms that the temperature control system operates under effective control conditions, highlighting the advantages of this optimization algorithm in enhancing PID controller design.