神经网络模糊多模型软测量在磨煤机存煤量测量方面的应用

Application of Fuzzy Multi-model Soft Sensor in Mill Load Measurement Based on Neural Network

基于钢球磨煤机的机理模型,采用神经网络模糊多模型软测量的方法解决球磨机存煤量测量问题,首先建立钢球磨煤机的机理模型,然后将FCM聚类与RBF神经网络多模型理论相结合深入探讨了神经网络模糊多模型软测量方法的实现,最后进行了球磨机存煤量测量的仿真试验,并与RBF神经网络单模型的仿真结果进行了比较.结果表明:神经网络模糊多模型软测量的预测输出的误差较小,训练速度更快,具有更好的泛化能力;将神经网络模糊多模型应用于球磨机存煤量的测量是可行的.

With mechanism model of ball mill,the problem of how to measure the coal load in ball mill has been solved by fuzzy multi-model soft sensing based on neural network.First,a mechanism model of ball mill was built up;then the specific measurement way of fuzzy multi-model soft sensor based on neural network was discussed in combination of FCM clustering with RBF multi-model theory;finally simulation tests were carried out for coal load measurement of ball mill,after which the simulated results were compared with that of RBF single-model simulation.Results show that the fuzzy multi-model soft sensor is more accurate in measurement,faster in training and stronger in generalization,which is therefore applicable for coal load measurement of ball mills.