Sensor fusion applied to position estimation and gait identification of an ankle-foot orthosis

This paper presents a sensor fusion strategy to estimate the absolute position and to identify the gait events of an active ankle-foot orthosis (AAFO). Sensor fusion is implemented using a Kalman filter with the signals of a gyroscope and an accelerometer, both present in an inertial measurement unit (IMU), besides the signals of three force sensitive resistors (FSRs) mounted in a sole of shoe. For correcting the gyroscope-based position, the redundant accelerometer-based position must be used only under reliable conditions, which are obtained by a proposed switching logic to enable the Kalman filter. Experimental results of the AAFO confirm the efficiency of the proposed sensor fusion strategy.