Objective Evaluation Of Powertrain Drivability On Light Duty Vehicle

Drivability reflects the subjective driving feeling of human-vehicle interaction in the longituadinal direction.The good property of drivability is the key to promote the competitiveness of products under the preconditions to meet the economy.At present,the domestic drivability evaluation research is focus in the field of subjective evaluation,the objective evaluation is almost blank.The traditional subjective evaluation is vulnerable to the testers and the environment,and it also has the disadvantage of the test process can not be reproduced,strong subjective and high cost.With the development of test and computer technology,the objective evaluation has become an important means of drivability evaluation.It summarizes the subjective and objective driving parameters of the driveline,and make the experimental study on the road test.To establish the subjective and objective evaluation index system,the relationship between the subjective feeling of driver and objective parameters of powertrain is being studied.The gear shift and drive away subjective evaluation of two test prototypre vehice was experiment with the road test,and the objective test parameters were recorded.The correlation between the subjective evaluation and objective parameters was analysed,and establishes the BP neural network prediction model and multivariate linear regression model,to evaluate objectively the drivability performance.By comparing the the relative error and consistency of two objective evaluation model to the subjective evaluation results,the accuracy and superiority of the BP neural network prediction model is being verified.It can indicate that the drivers' subjective drivability evaluation can be quantified as objective parameters.The BP neural network prediction model has the high consistency with the subjective evaluation and by using BP neural network to prediction the objective drivability is reasonable.The drivability objective evaluation based on the BP neural network has a broad development and application prospect,it fill the domestic blank in the research of objective evaluation of drivability and has strong theoretical research significance and practical use value.