摘要： 在 Labview组态软件、西门子PLC300和温度传感器构成的温度控制系统基础上，提出一种新的黄酒发酵温度控制系统，将带交叉因子的粒子群优化(PSO)算法应用到BP神经网络(MyPSO-BP)比例积分微分(PID)控制中。改进的PSO算法初始化神经网络的权重和阈 值，可以更好地在线整定PID参数，增强系统的稳定性和鲁棒性，减小误差。对系统进行Matlab仿真实验，结果表明，该系统相较于传统的神经网络PID控制器具有更好的温度控制性能。

关键词: 黄酒发酵温控, BP神经网络, 粒子群优化, 比例积分微分控制, 交叉因子, 惯性权重

Abstract: A novel temperature control method for rice wine fermentation control system is proposed, based on the system consisting of Labview, SIMATIC PLC300 and sensors. In this algorithm, cross-factor Particle Swarm Optimization(PSO) is applied to the BP neural network(MyPSO-BP) Proportion-Integral-Differential(PID) control. The improved particle swarm algorithm initializes the weights and thresholds of neural network, which can be an effective approach to tuning PID parameters on-line, enhancing system stability and robustness and reducing errors. Simulation experiment is performed by Matlab, and simulation result shows that this method compared with the traditional neural network PID controller has a great advantage in temperature control performance.

Key words: rice wine fermentation temperature control, BP neural network, Particle Swarm Optimization(PSO), Proportion-Integral-Differential (PID) control, crossover operator, inertial weight

中图分类号: 

• TP334.4