Research on PID Control of BP Neural Networks Based on PSO Optimization

In many industrial production fields, the proportional integral derivative (PID) control algorithm is widely applied by the rapid evolution of its control theory. However, conventional PID controllers faced certain difficulties in parameter tuning, real-time online adjustment, and processing of complex nonlinear systems. To solve this problem, the researchers organically combine the PID control algorithm and other mature control strategies to form a new control theory. In this paper, we propose a neural network PID control method based on particle swarm optimization (PSO). This method combines the adaptive learning ability of the BP neural network with the PID controller, and corrects PID control parameters in real time to improve the system stability. In order to overcome the problem that the bp-pid adaptive process is apt to be a problem of the local optimum and the convergence speed slow, optimization is carried out using an improved PSO algorithm. Experimental results show that this method can reduce system overshoot effectively and reduce steady-state error and drastically reduce the steady-state time compared to traditional PID control and bp-pid control. Therefore, the method studied in this paper can greatly improve system control accuracy and stability.