Application of improved extended Kalman filtering algorithm in velocity determination with GPS

Xueli FENG,Fuwu YAN,Jie HU
DOI: https://doi.org/10.16652/j.issn.1004-373x.2016.08.008
2016-01-01
Abstract:Dynamic velocity determination had to meet the higher requirement from the vehicle electronic control system. The algorithm combining finite?difference algorithm with extended Kalman filtering algorithm has higher accuracy in GPS veloci?ty determination than GPS location difference velocity determination method,but there is a larger measuring error during the fast turning of car. The finite?difference operation is adopted to achieve the priori error covariance matrix and the posterior error cova?riance matrix of filtering to enhance the convergence of the filtering process,and then the finite?difference extended Kalman fil?tering algorithm is used to calculate car’s real?time location,so as to get the real?time speed of the vehicle. The experiment re?sults show that the improved algorithm has higher measuring accuracy than that of the extended Kalman filtering velocity determi?nation algorithm.
What problem does this paper attempt to address?