基于粒子群模糊神经网络的丙烯腈收率软测量建模  

Soft-Sensor Modelling of Acrylonitrile Yield Based on Particle Swarm Optimization Fuzzy Neural Networks

作　　者：陈国初[1] 徐余法[1] 俞金寿[2] 

机构地区：[1]上海电机学院电气学院,上海200240 [2]华东理工大学自动化研究所,上海200237

出　　处：《系统仿真学报》2007年第23期5370-5372,共3页Journal of System Simulation

基　　金：教育部博士点专项基金项目(20030251003);上海市教委自然科学科研项目(05VZ01;06VZ002)资助

摘　　要：对粒子群优化算法与模糊神经网络的结合进行研究,提出粒子群模糊神经网络,并将其应用丙烯腈收率软测量建模。该方法采用模糊神经网络来构建丙烯腈收率软测量模型,用粒子群优化算法优化模糊神经网络的参数;并结合实际工艺,对所建软测量模型进行仿真研究。实验结果表明,该模型的性能优于粒子群神经网络模型,能够准确预测丙烯腈收率,具有较高的精度和良好的应用前景。By combining particle swarm optimization algorithm （PSO） with fuzzy neural networks （FNN）, a PSO fuzzy neural networks （PSOFNN） was proposed. Then PSOFNN was applied in soft-sensor modelling of acrylonitrile yield. The new method assumed that FNN was used to construct the soil-sensor model of acrylonitril yieM and PSO was employed to optimize parameters of FNN. Experiment results show that the model based on PSOFNN has higher precision and better performance than the model based on PSONN.

关 键 词：丙烯腈 收率 粒子群优化算法 模糊神经网络 软测量 模型 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]