Research On Objectification Of Subjective Evaluation For Passenger Vehicle Handling Stability

As an important performance of the car chassis dynamics,the steering stability of the car is not only related to the ease of handling of the car,but also related to the stability of the car at high speed and the ability of the car to remain stable when subjected to external disturbances.Therefore,the stability of vehicle handling is closely related to the driving safety of the vehicle,and is known as the "lifeline of high-speed vehicles."In recent years,with the improvement of computer virtual simulation technology,the adjustment of vehicle steering stability has changed from relying entirely on subjective evaluation to comprehensive technology based on subjective evaluation,supplemented by objective testing and virtual simulation.Therefore,the objective evaluation of vehicle handling stability has become a particularly important part of this trend.The core of objective research on subjective evaluation of vehicle handling stability is to establish a consistency model of steering objective and objective evaluation.In previous studies,it was often the case that multiple subjective evaluation drivers correlated the subjective evaluation results of different levels of vehicles with the indicators in the existing objective test standards.Different from previous research ideas,this paper avoids the adverse effects of driver's differences and vehicle differences on the subjective evaluation results in the research process.Based on the enterprise resources,11 groups of test vehicles are replaced by a certain SUV tire model for subjective evaluation of handling stability.And objective test,the evaluation results of a professional driver of the company.Objectively evaluate the driver's subjective evaluation method and form a set of test methods for objective testing.The vehicle dynamics theory is used to analyze the subjective evaluation indicators one by one,and an objective index system that can express its meaning is established.The statistical analysis method was used to establish the correlation model between subjective evaluation index and objective index,and the weighting model of subjective evaluation system.Finally,the consistency model of steering objective and objective evaluation was established.The main research contents of this paper are as follows:First,a research on the subjective evaluation system based on the driver's steering stability is conducted.In this paper,the subjective evaluation system adopted by the driver is divided into four levels: total evaluation of operational stability,performance evaluation,evaluation items and evaluation indicators.Based on the theory of vehicle dynamics and the description of the driver,the mechanism meaning of the subjective evaluation index is studied,and the driving methods of each subjective evaluation project are clarified,which provides a basis for standardizing the objective test driving method and the extraction of objective indicators.Secondly,based on the subjective evaluation system,an objective evaluation index system is established.The objective evaluation project driving method is objectiveized to form a set of driving methods for objective tests,the purpose of which is to obtain objective indicators under the same vehicle motion state as in the subjective evaluation.According to the vehicle dynamics theory of each subjective evaluation index and the driver's description of the meaning of each subjective evaluation index,the corresponding objective indicators are extracted respectively,and an objective evaluation index system for describing subjective evaluation indicators is established.The objective test data collected is subjected to sensor zero drift and low pass filter pre-processing.The objective index is calculated by polynomial fitting,gain calculation,normalization processing,signal correlation analysis,etc.,and provides a data basis for establishing a consistency-related model for the objective and objective evaluation of steering stability.Thirdly,based on the theory of ridge regression analysis,a model of subjective evaluation index and objective indicator is established.The ridge regression analysis and the ridge map method are used to screen out the objective indicators with weak or no correlation with the subjective evaluation items,to solve the problem of multi-collinearity between objective indicators,and to establish a more stable and reliable regression model.A statistical method was used to perform a goodness-of-fit test(test),a confidence test(F test),and a variable significance test(t test)on the regression model.The results show that the confidence level of all regression models is above 90%,and the goodness-of-fit test index(corrected sample determination coefficient)is greater than 0.84.Then,based on the fuzzy analytic hierarchy process(FAHP),the subjective evaluation of each level of weight model is established.The comparison discriminant matrix is introduced,and the relative importance of the subjective evaluation indexes of each level is given by the driver's expert experience,so as to avoid the subjective and objective evaluation consistency caused by the driver's inability to accurately give the subjective evaluation index weight coefficient or the weight assignment error.The model is not accurate.Combined with the triangular fuzzy number theory,the uncertainty of the driver's subjective evaluation index is considered.According to his expert experience,the upper and lower limits of the triangular fuzzy number are given,and the fuzzy comparison discriminant matrix is established.The square root method is used to calculate the order of fuzzy comparison judgment matrix,and the weight coefficient of each level of the subjective evaluation system relative to the upper level is obtained.Finally,a complete subjective and objective consistency model of steering stability is established.The subjective evaluation index is combined with the objective index correlation model and the subjective evaluation hierarchy level weight relationship model.The subjective evaluation item score is obtained by inputting the objective index subjective evaluation index and the objective index correlation model,and the subjective evaluation item score is subjectively evaluated at each level.The weight relationship model input obtains a subjective evaluation of each level of score.The error formula is used to verify the accuracy of the model.The results show that the index error of each level of the model does not exceed the tolerance error limit(3.3%).