Research On The Correlation And Quality Prediction Between Engine Tightening Process And Cold Test Vibration

Since the engine is a core part of an automobile,the quality of the automobile is affected by the quality of the engine,which is guaranteed by a scientific assembly process.The problem is that the engine assembly process nowadays lacks the process optimization program driven by the production data.Based on the engine assembly data,this paper studied the correlation between engine assembly process and the performance of engine by taking correlation analysis between the engine’s assembly tightening data and the cold test vibration data.Finally the prediction of engine production quality was realized.The main research in this paper include following parts:This paper analyzed the engine assembly line based on the production data.As the vibration peak of cylinder head and exhaust system being selected as the analysis object of vibration component,the correlation model of engine assembly process-quality-performance was put forward according to the engine assembly data,by analyzing the engine assembly process,the cold test mechanism and vibration.The selection of the influence factor of cold test vibration was carried out by the engine assembly process as the screening rule.At first,the torque standard deviation that is relevant to the dependent variable was selected by comparing the different variables according to the impact factors’ data structure.And then by means of mathematical statistical method,the correlation between the vibration peak of each station bolt,cylinder head and exhaust system was analyzed,and the weight of each assembly bolt was ranked.Finally,the correlation analysis was conducted to provide a guidance for adjustment and optimization of process parameters.Supervised learning algorithms such as BP neural network and support vector machine was used to analyze the production data.That means the quality prediction of cylinder head and exhaust system in engine cold test was realized by establishing a scientific quality prediction model,which could be used to improve the engine assemble quality and reduce the repair rate.In addition,the model can also be used to verify the correlation of cold test vibration factors on cold test vibration and lay a foundation for the software of predicting the assembly quality of engine.To make it possible for the model to guide the engine assembly process,The engine quality prediction analysis software has been established to realize the production data on-line acquisition,prediction and classification as well as the calculation of the engine qualification rate of the production data.The software was proved to work well as the prediction results are close to the actual data with a form of statistical graphical and report demonstrated on the Web.Due to this fact,the software was acknowledged by corporation and was put into use already.