Rough Set-Probabilistic Neural Networks Fault Diagnosis Method of Polymerization Kettle Equipment Based on Shuffled Frog Leaping Algorithm

Abstract

In order to realize the fault diagnosis of the polyvinyl chloride (PVC) polymerization kettle reactor, a rough set (RS)–probabilistic neural networks (PNN) fault diagnosis strategy is proposed. Firstly, through analysing the technique of the PVC polymerization reactor, the mapping between the polymerization process data and the fault modes is established. Then, the rough set theory is used to tackle the input vector of PNN so as to reduce the network dimensionality and improve the training speed of PNN. Shuffled frog leaping algorithm (SFLA) is adopted to optimize the smoothing factor of PNN. The fault pattern classification of polymerization kettle equipment is to realize the nonlinear mapping from symptom set to fault set according to the given symptom set. Finally, the fault diagnosis simulation experiments are conducted by combining with the industrial on-site historical datum of polymerization kettle, and the results show that the RS–PNN fault diagnosis strategy is effective.

Keywords

• polymerization kettle equipment
• fault diagnosis
• rough set
• probabilistic neural networks
• shuffled frog leaping algorithm