Research On Key Technology For Vehicle Autonomous Navigation System

Due to a series of advantages, inertial navigation has been widely applied in the vehicle autonomous navigation systems. To improve the accuracy and reliability of the vehicle navigation systems, one way is to design high-precision inertial navigation system, which needs optimal navigation algorithms and/or higher level instruments, the former one is better taking into consideration the cost. Another way is to construct suitable integrated navigation systems to achieve a certain balance between performance and cost.In this paper several issues on the vehicle autonomous navigation systems are discussed as follows:1. Inertial navigation integration algorithm is deduced, in which dynamic angular rate/acceleration effect is considered (coning in attitude integration, sculling in velocity integration, scrolling in position integration). Two different precision algorithms, iterative algorithm and simplified algorithm, are designed based on the former mentioned digital algorithm for different applications. Based on the study of typical vehicle driving conditions, two kinds of vehicle trajectory generation techniques are proposed. The performance of the two navigation algorithms is compared through simulation using the vehicle tracks.2. A four-wheel vehicle model with three degrees of freedom and tire model for tire force modeling is established and verified. Vehicle state estimator for yaw rate, longitudinal speed and lateral speed is designed; two difference methods, Extended Kalman filter (EKF) and Unscented Kalman filter (UKF), is applied to estimate the vehicle state. Comparison of the effects of the two filters is conducted through the simulation. The vehicle state estimation is built for the integrated navigation system as the sub-model.3. The error characteristics of inertial navigation are studied, the influence of sensor error on position accuracy is analyzed. Dead reckoning error equation and vehicle dynamic model error equation is established. INS-DR integrated navigation system is designed and simulated, so is the INS-VDM integrated navigation system. The results show that the estimation for both integrated navigation are biased, however, both reduce the navigation error. Since no additional sensor devices are required, INS-VDM integrated navigation is a better option for vehicle navigation.