GPS/INS-based Unscented Kalman Estimation Method Of Vehicle Motion State

In this era of increasing scientific and technological development,the automotive field has also joined the ranks of intelligent development.This requires the car navigation system to have lower cost and higher accuracy.Common single navigation systems,such as GPS,are prone to loss of satellite signals during single navigation,which causes the problem of reduced positioning accuracy.INS has the phenomenon of error accumulation,and the effect of working alone for a long time is not good.The complementarity of the two in terms of navigation principles,error characteristics,etc.,makes the GPS/INS integrated navigation system emerge as the times require.For this reason,this article expects to design a high-precision integrated navigation system in intelligent vehicle navigation at a lower cost.First,analyze and model the inertial navigation system.Establish a suitable coordinate system,and derive the posture matrix under the coordinate system conversion,and get the inertial navigation system solution algorithm.The error equation and random error equation are deduced in detail to provide a basis for the subsequent establishment of the integrated navigation system model.Secondly,the integrated navigation algorithm of this paper is designed by the error equation of the inertial navigation system.For the most important part of data fusion technology,aiming at the nonlinear system model,this paper uses unscented Kalman filter to realize the fusion of navigation data.The loose combination mode and feedback correction method are selected to establish an error estimation model,and a 21-dimensional state equation and a 6-dimensional measurement equation are derived.According to the established system model,combined with the random error characteristics of the inertial device,the Allan variance method is used to identify the parameters of the system noise.Introduce the concept of precision factor to identify observation noise.The identification result is used in the design of the Kalman filter.Finally,the effectiveness of the integrated navigation algorithm designed in this paper is verified.Design a simulation environment and a real-vehicle test environment,and analyze the performance of the integrated navigation algorithm in the two environments to verify the effectiveness of the algorithm.In the simulation environment,the Car Sim vehicle simulation tool is used to obtain vehicle motion information under ideal conditions.On this basis,noise signals are added to simulate the state of vehicle movement under real conditions.Through the integrated navigation algorithm,the error estimation of the motion information containing the noise signal is carried out,and the navigation error and the accuracy of the integrated navigation algorithm are obtained.In the actual vehicle test environment,record the vehicle’s motion trajectory and collect its motion information,including angular velocity,acceleration,etc.The algorithm is also used to solve it and complete the state estimation to obtain the navigation error in the integrated navigation state,and compare it with the navigation error of the single navigation system.The results show that the error performance of the integrated navigation is significantly better than that of only one navigation method,which verifies the effectiveness of the algorithm.