Research On Vehicle Longitudinal Vehicle Speed Estimation Based On Multi-sensor Information Fusion

The vehicle dynamic stability control(DSC)system includes an automobile brake anti-lock system,a traction control system,and an active yaw moment control system.The longitudinal velocity of the vehicle's center of mass is the key control parameter in the DSC control algorithm.Estimation algorithms are generally used for real-time estimation due to the difficulty of direct measurement.The research goal of this paper is to establish a method for estimating the longitudinal velocity of the car's center of mass.It has the characteristics of wide applicability,high precision and real-time operation.The algorithm established in this paper is mainly based on the cubature Kalman filter algorithm.Summarizes the design methods,design ideas and several basic problems in the design of Kalman filter: the stability of the filter,the error of the filter,the state initial value and how to select the statistical characteristics of the noise.The system model is the basis of the Kalman filter.A four-wheel three-degree-of-freedom car model is established in this paper.Using the magic formula tire model,the tire fitting force(including the longitudinal force and the lateral force)has a high accuracy of estimation and fitting.The basic signals of the DSC system are processed,and the existing sensors of the DSC system are used for multi-sensor information fusion.Considering the influence of vehicle driving conditions on key parameters such as wheel speed and acceleration,the driving state of the vehicle is estimated.According to the four-wheel three-degree-of-freedom vehicle model,a cubature Kalman filter based on the vehicle kinematics model and the vehicle dynamics model is constructed.The simulation verification and comparative analysis are carried out with Carsim and Simulink.Combined with practical experience in engineering,the confidence factor was introduced.Combining the advantages of the two methods,the two algorithms are combined to construct a dual-layered volumetric Kalman filter.Finally,the noise variance of the system and the variance of the model noise are designed to realize the adaptive algorithm.Finally,the actual vehicle test verifies the correctness,effectiveness and stability of the algorithm.The results are satisfactory.