基于和声搜索优化多T-S模糊神经网络的聚合釜过程软测量建模

Soft-sensor modeling of polymerizing process based on multi T-S fuzzy neural network optimized with harmony search algorithm

根据多个模型相加可以提高整体预测精度和鲁棒性的思想,提出一种基于模糊C均值聚类算法的多T-S模糊神经网络模型对聚氯乙烯(polyvinylchlorid,PVC)聚合生产过程中的氯乙烯(vinyl chloride monomer,VCM)转化率和转化速率进行预测。首先采用主元分析来对软测量模型的辅助变量进行选择以降低模型维数,并提出和声搜索和最小二乘法相结合的混合优化算法来优化T-S模糊神经网络子模型的结构参数。仿真结果表明该模型能够显著提高PVC聚合过程中经济技术指标预测的精度和鲁棒性,可以满足聚合釜生产过程的实时控制要求。

According to the principle that multiple models can enhance the overall accuracy and robustness of predicative model,a multi T-S fuzzy neural network soft-sensing model based on fuzzy c-means（FCM） clustering algorithm is proposed to predict the conversion rate and velocity of VCM in PVC polymerizing process.Firstly,principal component analysis（PCA） method is adopted to select the auxiliary variables of the soft-sensing model in order to reduce the model dimensionality.Then a hybrid optimization algorithm utilizing harmony search（HS） and least square method is proposed to optimize the structure parameters of the T-S fuzzy neural network.Simulation results show that the proposed model can significantly enhance the predictive accuracy and robustness of the technical-and-economic indexes and satisfy the real-time control requirements of PVC polymerizing production process.