Aiming at the problem that odometer measurement error leads to the reduction of integrated navigation accuracy, a fuzzy adaptive algorithm based on the combination of system working state and filter innovation state is proposed. According to the change of the innovation, the fuzzy rules are determined, and the output gain of the odometer is corrected, so that the innovation is always kept near the zero mean value, and the observation noise variance is corrected by using the corrected innovation, so that the deviation of the navigation positioning is reduced. The simulation results show that this algorithm can improve the accuracy of integrated navigation and positioning.