Study On Vehicle Dynamics Key Parameters Identification And Driving State Estimation

As the application of vehicle active safety control technology becomes more and more widespread,and how to obtain the vehicle dynamics parameters and driving state parameters accurately and stably is crucial to the research of active safety control strategies.The method of obtaining vehicle state information with limited measurement accuracy or difficulty to measure at low cost by vehicle state parameter estimation has received a lot of attention from scholars.Based on vehicle dynamics theory,the ride comfort parameters,handling dynamics parameters,tire cornering stiffness,road friction coefficient,and other parameters involved in vehicle dynamics control were identified in this paper,to provide the basis for active safety control technology of the vehicle.The main research work is as follows.1.Establishing an accurate equivalent model of the suspension system is the basis for the development of suspension control strategies.In this paper,a quarter-vehicle multibody dynamics model based on a double-wishbone suspension is firstly built in ADAMS/View.A 2-DOF suspension equivalent vibration model considering suspension nonlinear spring and nonlinear damping is established.Based on displacement step input simulation experiments,a homotopy optimization method is proposed to identify the equivalent model parameters of the suspension system.Combined with the random road simulation experiments,the linear and nonlinear models are compared in time and frequency domains to verify the effectiveness of the homotopy optimization algorithm and the accuracy of the equivalent model of the nonlinear suspension system.2.To address the problem of difficult to measure handling dynamics parameters,a 3-DOF maneuvering dynamics model considering roll motion is built,and a step-by-step identification strategy is proposed by using the front wheel step input simulation experiment of ADAMS/Car full car multi-body model.Firstly,a virtual sensor is established by combining the adaptive network-based fuzzy inference system(ANFIS)and the unscented Kalman filter method.The actual sensor data and virtual sensor data are used as the input to further identify the vehicle roll stiffness,roll damping,and other parameters by the unscented Kalman filter.Combined with the vehicle simulation test results for research and analysis to verify the effectiveness of the identification strategy.3.The tire-ground interaction is an important factor affecting vehicle handling stability.Based on the suspension parameters and the handling dynamics parameters,the Extended Kalman filter algorithm is used to identify the coaxial equivalent cornering stiffness in realtime by combining the 2-DOF handling dynamics model.A 3-DOF vehicle handling dynamics model considering longitudinal motion is further developed,and the ground adhesion coefficient of the four-wheeled vehicle is estimated in real time by using the Unscented Kalman filter combining with the Car Sim-Simulink co-simulation.The results show that the identification results under different adhesion coefficient roads are basically consistent with the Car Sim preset values,and the algorithm has high accuracy and effectiveness.