Research And Application Of Equipment Preventive Maintenance Policies Based On Fault Prediction

With the increasing scale of modern industrial production,the precision,automation and complexity of the mechanical equipment used in the production become higher and higher.Those advanced equipment can greatly improve production efficiency,improve product quality and save human resources,but also bring higher maintenance costs.If the equipment is running in the process of failure,it will causes the equipment to stop and result in a high maintenance cost,or it may even have impact on the entire production system and result in personal safety and significant economic losses.According to statistics,downtime and maintenance costs of equipment failure have accounted for 15%~40% of the cost of production.Therefore,how to predict the failure rate of the equipment timely and effectively before the equipment failure and on this basis to develop the effective preventive maintenance strategies to avoid equipment failure,has become an urgent problem in enterprises.Taking the single equipment for the research object,we propose a equipment failure prediction method which based on the grey rough set and BP neural network.On this basis,the preventive maintenance strategy is optimized to maintain the scientific management of equipment and ensure the normal operation of equipment.The main contents of this paper include the following aspects:Firstly,we have a general study about the preventive maintenance strategy based on fault prediction and define the theory and method of equipment fault prediction and the types and characteristics of equipment maintenance strategy.On this basis,we analyze the limitations of existing research and obtain the key problems to be solved in this paper.Then,focusing on the key problems,we put forward the research ideas and technical route of the paper.Secondly,we propose the method of equipment fault prediction based on the grey rough set and BP neural network.The grey incidence analysis method is used to simplify the horizontal data of the decision table of fault data,and the method of rough set is used to reduce the conditional attributes of the decision table of fault data,and remove the redundant and invalid data and attributes of the decision table.On this basis,we construct a BP neural network prediction model.Thirdly,we propose the preventive maintenance strategy based on fault prediction.Taking the minimum total maintenance cost,including the risk of equipment failure,as the optimization objective,and preventive maintenance failure rate threshold as the decision variable,then we introduce the equipment risk assessment model to analyze the loss caused by equipment failure.By calculating the minimum total cost of maintenance,the failure rate threshold of equipment preventive maintenance,maintenance interval and maintenance methods are obtained.Finally,the paper takes the numerical control lathe equipment of A enterprise workshop as the research object,and apply the methods in paper to the practice of the enterprise.Based on the prediction of equipment failure rate,the optimal preventive maintenance failure rate threshold,maintenance interval and the corresponding maintenance measures are obtained.The preliminary application results show the reliability and validity of the method described in this paper.