基于神经网络PID控制的颗粒物料称量系统  

Particles weighting system based on neural network PID

作　　者：谢宇[1] 韩保红[1] 段云龙 

机构地区：[1]军械工程学院,石家庄050003 [2]燕山机械厂,张家口075041

出　　处：《国外电子测量技术》2013年第9期15-17,共3页Foreign Electronic Measurement Technology

摘　　要：颗粒状物料自动精确称量设备采用先粗称量再分粒精称量的新型控制方法,以保证对物料的高精度称量,其中粗称量控制精度与稳定性对整体控制精度及合格率有很大影响,传统控制方法难以满足控制要求。在粗称量控制系统中,利用BP神经网络与PID控制器相结合,可在线调节PID参数,提高称量系统控制的精度与稳定性,达到粗称量要求。仿真结果表明,神经网络PID整定收敛速度快,超调量小,跟踪误差小,控制效果良好,能够达到称量控制要求。A new control method is taken in the automatic weighting equipment of particles to make sure the high precision weighting.In this method,the particles firstly is weighted roughly and then is weighted accurately by counting the particles.The control precision and qualified rate are decided by the precision and stability of the rough weighting.The traditional control method can＇t meet the demand.The BP neural network which is combined with the PID controller,can adjust the parameters of PID controller online It can satisfy the demand of the rough weighting through improving the precision and stability of the system.The simulation results shows the PID controller based on BP neural network has fast convergence,small overshoot and small tracking error.The controller which has good effect can meet the demand of propellant weighting.

关 键 词：神经网络 PID 称量 精度 稳定 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]