Application of high-efficiency experimental design method for predicting the performance and emission of diesel engine

In order to provide high-efficiency and reasonable experimental design methods for diesel engine performanceprediction and optimization, this study compared and analyzed the accuracy and efficiency of the regression modeldesigned and tested by different methods, researched the application of orthogonal design and uniform design inestablishing the multiple nonlinear regression predictive model of diesel engine performance and emission. Theregression model adopted stepwise regression analsysis method, which can extract factors that influence the predictiveparameter most obviously according to the test data, so as to build the most accurate and simple predictive model. Inaddition to the main effect of each factor, the interactions and nonlinear actions among factors were also considered inthe establishment of regression model, so as to make the model more accurate. Finally, the accuracy and efficiency ofprediction of the above-mentioned two kinds of regression models were compared and analyzed, and the rests showedthat the uniform design was more efficient in the predictive test research of diesel engine performance and emission.