Study On Vehicle Handling-stability Evaluation Approach Based On Humanvehicle-road Closed-loop System

In the booming automobile industry, car safety, energy conservation and environmental protection have become three research focus in automobile research fields, where security is of top priority considering vehicle performance. Because of the research of vehicle passive safety, human injury has been sharply reduced. However the enhancement of passive safety can’t prohibit traffic accidents from happening. With the rapid development of technology, vehicle active safety has been paid more and more attention and has occupied a very important role during automobile design and engineering. Therefore, the test and evaluation method of active safety is becoming one of the most important research fields in vehicle engineering area.The handling and stability of vehicle is one of the key aspects of vehicle active safety performance and is influenced by vehicle structures, road conditions, drivers and other factors. In order to evaluate the performance of the vehicle handling and stability, it is necessary to carry out the handling and stability tests. The handling and stability tests are based on subjective evaluation, and are affected by the difference of individual driver. Hence, this thesis is aiming to analyze the performance of vehicle handling and stability with an novel objective test method, expecting to carry on the analysis combining virtual simulation and road test, instead of lots of real experiments by drivers purely, and to reducing the experimental danger and the individual differences of drivers. Furthermore, the weight coefficients of evaluation index are analyzed. Based on the study of the consistency between subjective and objective evaluation, an objective evaluation method is proposed.To be more specific, this dissertation is organized in following structure.First, the driver model of driver–vehicle–road closed–loop test system is introduced. Based on the Preview–Followller theory, the Preview–Compensation Optimization Neural Network driver model is built through the analysis of the vehicle and driver status feedback.To describe those drivers and analyze their behaviors, the key factors of driving characteristics is proposed and analyzed. The selection and determination of driver model parameters is the basis for predicting the accuracy of driver behaviors. The reference model, preview time, reaction time and inertial time are used to determine the parameters of the driver model. The vehicle model can be obtained through open test identification,this model is compared with the test data of steering step wheel angle under different speeds and lateral accelerations. The human–vehicle–road closed–loop test system is established based on vehicle model identification and MSC Car Sim software. The system is analyzed on split roads. The co-simulation results show that the human–vehicle–road system has good performance which can be used to analyze the vehicle handling and stability.Secondly, design and optimization of the equipment for the handling and stability test are proposed. In order to test the vehicle handling and stability, it is necessary to analysis the technical parameters of test equipment. The combination of test device is a crucial aspect of test evaluation. The open–test and closed–test methods are studied; the test variable and the requirements are proposed through analyzing the test standards. The test fixtures are improved during the tests. The best combination scheme is found, which provides a reliable test data and reduces the redundancy of test variable.Thirdly, the study on the objective test of vehicle handling and stability is carried out. Based on the extensive reviews of the vehicle test methods, the human–vehicle–road closed test system can be used during the road test combining virtual simulation and road test. Vehicle model and driver model are analyzed. The test system of road test is built, the error of lateral displacement is measured, then the error and affecting factors are analyzed, and the error is reduced through adopting the compensation factor. The pre-given trajectory is obtained through gathering the real trajectory by the driver in the double lane change test or slalom test, and is inputted into the human-vehicle-road system. Driver handling input under typical conditions is obtained through optimization algorithm, the uncertainty of driver and proving ground is coupled with steering angle. The test with the steering angle is carried out through the steering robot on proving ground, and results are compared with the results of real driver test. The results show that the test method has finished the task successfully in many tests and obtained satisfactory effect.Finally, studies of the handling stability objective evaluation are presented. The evaluation criteria of subjective and objective tests which represent the characteristics of handling and stability are analyzed based on the stability test methods. Objective evaluation indexes are assigned through multi–strategy assignment methods. The preference coefficients of subjective and objective are found. The weight coefficients of subjective and objective evaluation are studied and the objective evaluation method is determined. The results of subjective and objective evaluation are obtained through the experimental data under typical working conditions on proving ground and the consistency of subjective and objective evaluation is analyzed, the effectiveness of objective evaluation is proven.