Research On Sound Quality Evaluation And Optimization During Multi-conditions For Pure Electric Passenger Vehicles

In recent years,the market share of pure electric vehicles has been increasing rapidly.As an important branch of new energy vehicles,pure electric passenger cars have won the favor of more and more consumers because of the advantages of maintenance costs,air pollution,power performance and noise emissions.However,because of the differences in NVH characteristics and subjective preferences,the emphasis of noise evaluation of pure electric vehicle is different from that of traditional fuel vehicle,and the sound quality has a greater impact on the vehicle driving experience.Therefore,new evaluation method of interior noise for pure electric passenger vehicles is urgently needed.At the same time,the current evaluation of sound quality is mostly oriented to single stationary condition,and further study is needed for the evaluation of critical multi-condition.This paper is focused on the subjective and objective evaluation and improvement of sound quality for pure electric passenger vehicles in multiple working conditions.The vibration and noise data of several prototypes in various steady-state and non-steady-state conditions are collected by NVH test in semi-anechoic chamber.The orders and influencing factors of high-frequency gear whine noise of a representative vehicle are identified and investigated by the methods of spectrum analysis,order tracking analysis and transmission path analysis;and the source of abnormal noise is determined by test troubleshooting.A hierarchical comparison method is proposed for the subjective evaluation problem in multiple working conditions.And the subjective evaluation test of multi-operating conditions for pure electric passenger vehicles is designed and developed.With analysis of subjective and objective evaluation data,it is verified that the indicators commonly used could not effectively evaluate the high-frequency order noise of unsteady conditions.In response to this problem,a metric for pure tone analysis is introduced and the objective evaluation metric set of pure electric passenger car sound quality is established.Based on the results of subjective and objective evaluation,a multi-case noise sample library is established.The sound quality evaluation models based on neural network and support vector machine are established and their prediction accuracy is compared.Then the improving method of prediction accuracy is discussed.Finally,the modal and dynamic response of the gear-shafting system,the optimization of gear modification and the modal of the reducer housing are analyzed by modeling method.The mechanism of the gear whine noise is discussed,and the technical measures to improve the sound quality are proposed.The study shows that the established metric system of interior noise objective evaluation is effective and reliable for pure electric passenger vehicles.It is proved that the established multi-case sound quality evaluation model which unifying subjective and objective data has high accuracy.The study could be helpful for the efficiency and accuracy improvement of sound quality evaluation for pure electric passenger cars.And the optimization of sound quality provides a reference case for the research of sound quality improvement in engineering.