基于BP神经网络的控制算法及在制药生产中的应用

The Control Algorithm Based on BP Neural Betwork and the Application in Pharmaceutical Production

制药工业是我国国民经济的主要产业,药品生产质量的要求也越来越严格.药品生产车间洁净区的环境条件是影响药品生产质量的主要因素之一,针对药品生产车间温度控制系统大滞后、大惯性、非线性,以及难以建立精确的数学模型等特点。在传统PID控制的基础上,基于神经网络具有任意非线形表达能力,以及可以通过对系统性能的学习来实现P I D参数的在线调整,确定最佳组合的P I D参数来实现车间温度的有效控制,本文通过仿真分析表明控制效果良好,具有调节时间短、无超调、稳态误差小等优点。

The pharmaceutical industry is the main economical industry of our country;the pharmaceutical production quality requirements are increasingly strict.The clean environment of pharmaceutical production workshop is one of the major factors that affect the quality of drug production;the temperature control system of workshop is always including time delay,big inertia,nonlinear,and,difficult to build accurate mathematical model and other characteristics.Based on the traditional PID control,the control algorithm based on BP neural network has the expressing ability of nonlinear,can realize online adjustment of PID parameters from the study of performance of the system,and determine the best combination of PID parameters to achieve the effective control of temperature workshop.The simulation analysis shows that the control effect is good,that is have short setting time,no overshoot,steady-state error is smaller,etc.