Research On The Estimation Of Handling And Stability State For Commercial Vehicle

Highway transportation is the main mode of passenger and freight transportation in our country and it plays a fundamental role in national economy. This role is growing as the process of road network, and especially the highway network continues to accelerate. Passenger coach and tractor semi-trailer are two main commercial motor vehicle in passenger and freight transportation, they play an important role in highway transportation industry. Handling and stability of passenger coach and tractor semi-trailer has direct effects on people's life property safety and development of road transport market. The assurance and improvement of handling and stability of the two type vehicles could reduce the occurrence of major and extraordinarily serious traffic accidents and the loss of life and property, which is of great realistic significance.In order to ensure the safety of vehicles in market, newly developed prototype vehicle must be evaluated through handling and stability tests. Efficient and reliable stability control system could improve vehicle active safety. However, both the evaluation of vehicle handling stability and the effective working of stability control system require vehicle state as input information. Accurate state information can improve the reliability of evaluation result and the performance of stability controller. Based on a lot of references in related fields, estimation of handling and stability for passenger coach and tractor semi-trailer is studied in this paper. The main contents and results are listed as follows:First, Extended Kalman Filter, Unscented Kalman Filter and Particle Filter are chosen as three candidate algorithms for commercial vehicle handling and stability state estimation. Based on learning and summarized theories of the three algorithms, algorithm programs are compiled. The performances of three algorithms are compared through simulation. Simulation results show that EKF algorithm is better than UKF and PF in accuracy and real-time quality.Second, a commercial vehicle handling and stability state detection system is built. The hardware of state detection system include two GPS receivers, a three-axis angular rate gyroscope acceleration, a steering wheel angle sensor, a processing terminal and a power supply systems. Through experiments the coordinate system of gyro is conformed and the steering wheel angle sensor is calibrated. Finally, a handling and stability state detection system for commercial vehicle is integrated, and also the stability and reliability of the system are tested.Third, two state estimation models of handling and stability for commercial vehicle are established. The state detection system can not measure tire force, and it is more practical to establish kinematics based state estimation model. At first, a state estimation model for passenger coach is established based on kinematics analysis. State variables in the model include longitudinal and lateral displacement variation in the vehicle coordinate system, longitudinal velocity, lateral velocity, yaw angle and yaw rate. Longitudinal acceleration and lateral acceleration are taken as the control variables. On the basis of passenger coach, the state estimation model of handling and stability for tractor semi-trailer is also built.Finally, handling and stability tests of a commercial vehicle are carried out, including the single lane change test, double lane change test, pylon course slalom test and stable circle test. EKF, UKF and PF algorithms are applied to estimate vehicle states in the tests. Estimation results of the three algorithms are all acceptable, which could be strong evidence to correction of the estimation model established. Then estimation results of the three algorithms are compared and a conclusion is drawn that EKF algorithm is more suitable than both UKF and PF for state estimation of vehicle handling and stability, which is similar to the results achieved in simulation tests. Information fusion technology is also applied to estimate vehicle states, and results carried out through fusion estimation are better than results estimated only by GPS information.