Stepping type pedestrian navigation method based on bidirectional filtering smoothing technology

The invention provides a stepping type pedestrian navigation method based on a bidirectional filtering smoothing technology. The method comprises the following steps: acquiring pedestrian walking data; performing error correction on the acquired pedestrian walking data in a first forward algorithm channel, and performing strapdown inertial navigation calculation on the posture and speed by utilizing the corrected pedestrian walking data; simultaneously performing gait time phase detection and stepping interval partition, and performing open-loop Kalman filtering estimation on the posture error and speed error in an error domain when a support phase of the gait cycle is detected; smoothing the posture error and speed error estimated in the first forward algorithm channel in a backwards algorithm channel; and performing error correction on the posture and speed calculated in the first forward algorithm channel by utilizing the error estimation smoothed in the backwards algorithm channel in a second forward algorithm channel, integrating the speed subjected to error correction, and calculating the pedestrian walking position. According to the method disclosed by the invention, the accuracy, robustness and smoothness of the pedestrian navigation method can be improved.