Fault Diagnosis System For Polyurethane Insulation Board Production Line

In recent years,with the gradual development of science and technology,the continuous improvement of customer requirements and the increasing competitive pressure,the polyurethane insulation board production line equipment is gradually improved towards the direction of large-scale,systematic and intelligent development and meanwhile its automation degree is getting higher and higher.Because the polyurethane insulation board production line needs to consume a large amount of electricity when it starts,and the chemical raw materials for production are also expensive,once the failure occurs,if not handle in time,it will bring great losses.Since the application of fault diagnosis technology in polyurethane insulation board production line is very few and the original diagnosis system,which can't diagnose some complex and uncertain faults accurately,has poor self-adaptability and lacks learning function,this study intends to diagnose and monitor the faults of polyurethane insulation board production line system by adopting a diagnostic method in which neural network and expert system are combined.Firstly,the fault diagnosis method based on neural network is emphatically analyzed.The diagnosis object of the method is mainly aimed at the polyurethane high-pressure foaming machine,which is the core equipment of the whole production line.In order to diagnose the faults timely and accurately,BP neural network is selected to diagnose the faults.This paper adopts the adaptive immune genetic algorithm to optimize the BP neural network,which overcomes the shortcomings of original BP neural network.This paper designs a MATLAB simulation experiment using the theory of BP Neural Network before and after optimization and the experimental data of historical fault data,The results show that the improved algorithm has faster diagnosis speed and higher accuracy.Secondly,diagnosis method based on the fault tree is introduced.The diagnosis object is other equipment of polyurethane insulation board production line.Based on the previous fault mechanism analysis,the fault tree model is established,and the weight of each basic event to the fault occurrence is calculated by fault tree analysis method and the analytic hierarchy process.Then,the expert system of polyurethane insulation board production line is constructed.In the part of fault tree analysis diagnosis,the production representation is used to describe the fault rules,and the knowledge base of the expert system is established.In the reasoning mechanism,the knowledge of the production rules is reasoned and diagnosed by forward reasoning.In the diagnosis part of the neural network,knowledge is acquired by case study and stored in the structure of the neural network.The reasoning process is completed by the neural network.Finally,the overall design scheme of the diagnosis system is presented,which includes hardware design and software design.Using Visual Studio 2012,SQL Server 2012 and MATLAB develops a fault diagnosis software system for polyurethane insulation board production line.The simulation results show that the diagnostic system has high practicability.