To address the problems of low pass rate and sudden change in speed of unmanned vehicles using the traditional timed elastic band (TEB) algorithm in a narrow space environment, a pose auxiliary point TEB algorithm (pose auxiliary point TEB, PAP-TEB) is proposed. By adding pose auxiliary points on both sides of the narrow space, the fixed-point planning algorithm is used to improve the passing rate of the narrow space environment. Secondly, the output speed of autonomous navigation and fixed-point navigation is optimized by using speed interpolation control and S-shaped speed planning. The method guarantees the smoothness of the output curve and the stability of the navigation process. Finally, the performance of the PAP-TEB algorithm is evaluated through the comparative simulation test of the unmanned vehicle operating system and the two-wheel differential unmanned vehicle experiments. Results show that the pass rate of the PAP-TEB algorithm is increased by about 50% in the 55 cm narrow channel scene. In the through-the-gate experiment, the PAP-TEB algorithm improves the pass rate by more than 20% compared to the traditional TEB algorithm, and the smaller the width, the more obvious the lifting effect. In addition, the fluctuation of the speed curve of the unmanned vehicle is small, and the smooth motion of the unmanned vehicle can be realized.