Research on Adaptive Precise Discharge Control System for Hoppers Based on RBF- PID

Abstract The dynamic weighing of material discharge from a funnel-shaped hopper has nonlinear and dynamic time-varying characteristics, which makes it difficult to control the accuracy of the discharge weighing. To improve the accuracy of the material discharge control system and its adaptability to complex operating conditions, this paper proposes an adaptive PID control method based on radial basis function (RBF) neural networks to construct a precise material discharge control model for hoppers. By constructing a three-layer RBF neural network model to identify the Jacobian information of the weighing and discharge system online, and based on this, the three parameters of the PID controller (proportional, integral, and derivative) are adjusted in real time, dynamically adjusting the opening of the sector gate at the discharge port to precisely control the instantaneous mass flow of granular materials. MATLAB simulation results show that, compared with traditional PID control, this method can significantly reduce system overshoot and improve control accuracy by nearly 50–70% while slightly improve response speed, thereby improve control system synthetical performance largely; Furthermore, it can maintain stable control under different target quantities, demonstrating excellent robustness and adaptability, thereby effectively enhancing the overall performance of dynamic weighing systems.