基于改进卡尔曼滤波的车辆定位精度仿真研究

Simulation of Vehicle Positioning Accuracy Based on an Improved Kalman Filter

研究车辆定位优化问题,关于定位精度是衡量车辆定位性能参数,针对精确地对车辆位置信息进行测量,传统方法测量误差大。为解决上述问题,提出了利用改进的模糊逻辑自适应控制器(FLAC)调整扩展卡尔曼滤波器(EKF)的方法。应用与GPS/DR信号融合,控制器通过残差的均值、理论和实际协方差的差值及差值变化率三个参数来确定滤波器的发散程度,并利用模糊推理规则抑制滤波器的发散,使滤波器不断执行最优估计。仿真结果表明,算法与常规EKF相比,定位精度、鲁棒性和可靠性等有较大的改善,尤其是有大误差偏差的情况时,优越性更为显著,满足了定位系统高精度、低成本的要求,具有一定的实用价值。

Positioning accuracy is one of the most critical parameters of the vehicle positioning system performance.How to accurately measure the vehicle position information has been a research focus.This paper presented a method of new fuzzy logic adaptive controller adjusting extended Kalman filter based on the information fusion for GPS/DR.The degree of filter divergence was determined by monitoring the three parameters of the residuals mean,theoretical and practical difference between the covariance and the change rate of the difference.The filter divergence was inhibited by using the fuzzy inference rules.Thus the optimal estimation was continued implemented by the filter.Compared with the conventional EKF,the simulation results show that the positioning precision,robustness and reliability are all improved.The superiority of the more significant occurs in the case of large deviation error.It can meet the requirements of positioning system with high accuracy,low cost and certain practical value.