Design of automatic tracking trolley: Optimizing PID parameter adaptability via BP neural network for path tracking accuracy improvement

Aiming at the problems of fixed parameters, poor adaptability to path deviation changes, and inability to meet high-precision path tracking requirements of traditional PID controllers for intelligent tracking trolleys, an adaptive PID control algorithm based on BP neural network is proposed. This method uses the nonlinear mapping and self-learning ability of BP neural network to dynamically adjust Kp, Ki, Kd according to three system states (current error, error integral, error derivative), so as to improve the trolley’s adaptability to different path types (straight line, sharp turn, S-curve). The simulation results show that compared with traditional PID, the proposed algorithm shortens adjustment time by 35.7% (from 420ms to 270ms), reduces maximum overshoot by 27.8% (from 22.3% to 16.1%), and maintains stable tracking when system parameters are perturbed by ±20%.