Intelligent Vehicle Stability Control Based On Driver Driving Skill Identification

Since active safety technology can effectively avoid dangers such as sidereal slip,it is of great significance to avoid traffic accidents and improve traffic safety.Therefore,stability control has become a hot topic in current research.Considering the stability requirement of vehicle yaw motion,how to select the corresponding control strategy,objective function and constraint condition is the focus of this paper.In addition,in the existing studies,the design of control strategy does not fully consider the skill difference of the whole driver group.How to effectively the driving skills feature classification and accurate identification and placed in the vehicle lateral stability controller,on the basis of how the driver skill levels to consider in the controller design is the focus of this article,at the same time,the real-time update of security and stability of border how to estimate the tire-road friction coefficient was also the focus of this article.This paper makes the following research and analysis.This paper on the major projects of national natural science fund project four(No.61790564)"Extreme coordinated control of the vehicle movement under the condition of integration",the national natural science foundation of key projects of international(regional)cooperation and the exchange(No.61520106008)"For safety of electric cars rolling optimization of energy efficiency" and national natural science foundation of China youth fund project(No.61703176)"Fast calculation method of nonlinear model predictive control research and application" of funding,the following research and analysis.For the driver driving skill identification problem.First of all,a driving skill question and a self-evaluation quantization table are designed and then driving simulator and data acquisition software are used to collect and process driving data in the process of driving.Through data analysis,characteristic parameters that can represent the characteristics of driving skills are constructed to describe the driving skills of each identified segment.In order to reduce the computational workload and weaken the correlation between evaluation indexes,the PCA method was used to calculate and process the constructed driving skill characteristic matrix.In addition,K-meansclustering method is adopted to classify and identify driving skills.At last,the recognition results are analyzed by combining with the results of the driver's subjective questionnaire.In order to solve the problem of difficulty in estimating road adhesion coefficient,the nonlinear tire model and vehicle dynamics model are used to prove theoretically.And the mathematical function relationship between road adhesion coefficient and some vehicle state parameters is established.Secondly,the BP neural network is designed to train the relationship between the corresponding vehicle state and road friction coefficient,so as to obtain the calculation model of road adhesion coefficient.The estimated results are verified in different working conditions which can meet the accuracy requirements of steady-state safety boundary update under various pavement conditions.For the stability demand of vehicle yaw motion,the model predictive control strategy was adopted to transform the yaw stability control problem into an optimization problem.The optimal control variables,namely the additional front wheel rotation Angle and braking force exerted by each wheel,were solved through rolling optimization and acted on the steering/braking actuator of the vehicle respectively.Taking the tracking of the yaw velocity and the restriction of the actuator action as the optimization target,the stability boundary drawn by the phase diagram of ?-??as the constraint condition,and the simulation experiment under the given working condition can ensure the yaw stability of the vehicle under the action of the controller.Considering that most of the current control strategies are only aimed at general types of drivers' skill characteristics,the combination of driving skills and controller is converted into the objective function weight optimization problem according to the driver's skill identification results after the identification results are obtained,and the weight change is determined and adjusted in real time through fuzzy rules.On the other hand,the stable and safe boundary is updated by setting the coefficient of driving skill.Simulation results show that for skilled drivers,the actuator control action can be reduced on the basis of ensuring safety.For drivers who are unfamiliar with driving skills,part of driving comfort can be sacrificed to make the vehicle safer.Therefore,the vehicle stability control system with certain adaptability to the driver is realized.