定量称重包装系统RBF神经网络PID控制研究

针对定量称重包装系统具有惯性、滞后、非线性时变且无法建立精确模型等特点,分析物料动态称量响应过程的动态特性及影响定量称重包装精度的相关因素和误差来源,提出1种基于RBF神经网络PID的定量称重包装控制策略。利用具有任意非线性表达能力及较强自学习能力的RBF神经网络寻求最佳的PID参数,并通过Matlab仿真验证控制策略的有效性。结果表明,与传统的PID控制相比,RBF神经网络PID控制策略具有较强的抗干扰能力,可显著改善定量称重包装系统的控制效果。

A Study of Quantitative Weighing Packaging System Based on RBF Neural Network PID Control

For quantitative weighing packing system with inertia,hysteresis,non-linear and not establishing a precise time-varying characteristics of the model,based on analysis of the dynamic characteristics of the material in response to dynamic weighing processes,and related factors and sources of error of affecting the accuracy of quantitative weighing packaging,this article presents a quantitative-based RBF neural network PID control strategy of weighing and packing.Using with any nonlinear self-learning skills and a strong ability to seek the best RBF neural network PID parameters,Matlab simulation verifies the effectiveness of the proposed control scheme.The results show that RBF neural network PID control strategy has stronger anti-jamming capability than the traditional PID control,and significantly improve the control effect of the quantitative weighing packaging system control.