| This paper develops an approach for the integration of GPS, inertial measurements from accelerometers and gyros, and differential wheel speed sensors for land vehicle navigation. Incorporating differential wheel speed sensor information into land vehicle navigation provides a solution for eliminating large errors caused by vehicle roll and pitch while also reducing errors from sideslip. Extended Kalman and Unscented Filtering algorithms are designed with a six degree-of-freedom model. In order to incorporate differential wheel speed information properly, the effective wheel radius must also be estimated as part of the overall estimation approach. Simulation results show the performance of the filters for cases of GPS/INS with and without the differential wheel speed sensor. |
