Technical Research Of Data-driven Accessory Fault Data Value-added Service

With the advancement of science and technology and economic development,manufacturing enterprises continue to innovate and develop,China's car ownership is gradually increasing,and parts maintenance is increasing.As the source enterprise of the accessories,the supplier cannot grasp the fault law of accessories in time,and cannot take corresponding measures in time.At the same time,the lack of accurate prediction of the number of parts damage,the supplier can not improve the quality of after-sales service.In response to this problem,this paper conducts a technical study on data-driven accessory fault data value-added services.Through the comprehensive analysis of damaged accessories,it provides an opportunity to improve the quality of accessories and assist in corporate decision-making.At the same time,based on the prediction of the amount of damage to the parts,it provides conditions for the supplier to guide the after-sales spare parts and support the optimization service,and achieves the purpose of increasing the value of the fault data.This article takes BY enterprise as the research object.Under the new situation,Comprehensive research and accurate prediction of damaged accessories become the basis for suppliers' survival.The auto industry value chain collaborative cloud platform,as a third-party role,connects the upstream and downstream enterprises of the automobile and masters the comprehensive accessories data from production to after-sales.The stability of the platform operation is excellent,attracting a large number of small and medium-sized automobile enterprises,but the research on the damaged parts of the supplier in the collaborative cloud platform still needs to be improved.Therefore,this paper studies and predicts damaged parts based on the collaborative cloud platform framework.In order to improve the prediction accuracy of the model,this paper proposes the LSTM-SVR hybrid prediction model.By combining the LSTM network with the linear support vector regression model,the model can improve the overall prediction accuracy of the model.Based on the actual prediction results of BY enterprise's accessory damage,the performance of the LSTM-SVR hybrid model is verified by comparing with the LSTM network and the SVR model.This article is based on the B/S three-tier architecture and data visualization.According to the actual needs of BY enterprises,this paper developed the accessory fault analysis function module,the accessory analysis report function module,the faulty component accurate search function module,and the accessory damage prediction function module.Let the supplier fully grasp the damage status of the parts,timely optimize the problem of the parts failure,and predict the repair of the parts in time through the prediction of the number of damaged parts,improve service quality.