Human Locomotion Implicit Modeling-Based Real-Time Gait Phase Estimation

Gait phase estimation based on inertial measurement unit (IMU) signals facilitates precise adaptation of exoskeletons to individual gait variations. However, challenges remain in achieving high accuracy and robustness, particularly during periods of terrain changes. To address this, we develop a gait phase estimation neural network based on implicit modeling of human locomotion, which combines temporal convolution for feature extraction with transformer layers for multi-channel information fusion. A channel-wise masked reconstruction pre-training strategy is proposed, which first treats gait phase state vectors (a three-dimensional representation composed of a polar-encoded phase value and its first-order time derivative) and IMU signals as joint observations of human locomotion, thus enhancing model generalization. Experimental results on datasets involving level walking, stair ascent/descent, slope ascent/descent, and transitions between level ground and these terrains demonstrate that the proposed method outperforms existing baseline approaches, achieving a gait phase RMSE of 2.729 ± 1.071 % and gait phase rate MAE of 0.037± 0.016% under stable terrain conditions with a look-back window of 2 seconds, and a phase RMSE of 3.215 ± 1.303% and rate MAE of 0.050 ± 0.023 % under terrain transitions. Hardware validation with one subject (N=1) wearing a hip exoskeleton further confirms that the algorithm can reliably identify gait cycles and key events, adapting to various continuous motion scenarios. This work lays the groundwork for more adaptive exoskeleton systems capable of robust real-time gait assistance in varied and dynamic environments.