四容水箱的小波神经网络广义预测控制  

Wavelet neural network generalized predictive control for quadruple-tank system

作　　者：宋清昆[1] 余杉钰[2] 韩笑[1] SONG Qingkun;YU Shanyu;HAN Xiao(School of Automation, Harbin University of Science and Technology, Harbin 150080, China;Shenzhen Clou Electronics Corporation Limited, Shenzhen, Guangdong 518057, China)

机构地区：[1]哈尔滨理工大学自动化学院,哈尔滨150080 [2]深圳科陆电子科技股份有限公司,广东深圳518057

出　　处：《计算机工程与应用》2016年第16期216-220,228,共6页Computer Engineering and Applications

摘　　要：针对四容水箱系统的多变量、大时滞、非线性及强耦合等特性,采用了小波神经网络广义预测控制(WNNGPC)策略。利用小波神经网络良好的函数逼近能力,对系统被控对象进行辨识,得到控制系统的预测模型,再结合广义预测控制良好的控制性能,达到对四容水箱系统的稳定控制。在系统辨识的过程中,采用的是改进的BP学习算法,这一算法能够快速平稳地修正网络权值和阈值,使预测输出平滑地趋近期望输出。在解决系统的耦合问题上,利用了模糊控制的通用逼近性,设计了模糊前馈补偿解耦。基于模糊补偿解耦的WNNGPC对四容水箱进行控制实验,并对比分析实验结果。结果表明,这一控制策略对四容水箱进行控制时取得了较好的控制效果。Considering the characteristic of the quadruple-tanks, such as multiple variables, great time-delay, misalignmentand coupling, the author applies the wavelet neural network generalized predictive control(WNNGPC). Predictive modelof the control system can be obtained through recognizing the system controlled object based on the good function approximationof the wavelet neural network. And with the combination of the good control performance of the generalized predictivecontrol, the quadruple-tanks system can achieve stability control. During the reorganization of the system, the authorapplies the optimal BP neural network. This algorithm can correct weights and thresholds of the network quickly, andmake the prediction output approaching the desired output smoothly. On solving the coupled problem of system, the authordesigns a fuzzy feed forward compensation decoupling by using the universal approximation of fuzzy control. Use theWNNGPC based on fuzzy compensation decoupling to do the experiments of the quadruple-tanks, and analyze the resultsof experiments. Through the experiments and analysis, it can be indicated that the control strategy achieves good controleffect of the quadruple-tanks.

关 键 词：四容水箱 小波神经网络 广义预测控制 模糊补偿解耦 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]