Step length estimation with wearable sensors using a switched-gain nonlinear observer

This paper focuses on step length estimation using inertial measurement sensors. Accurate has a number of useful health applications, including its use in characterizing the postural instability Parkinson’s disease patients. Three different sensor configurations are studied sensors shank and/or thigh human subject. The problem several challenges due to unknown bias, misalignment body and desire minimum A nonlinear is formulated that aims estimate angle, bias parameters lengths. observer designed Lyapunov analysis requires solving an LMI find stabilizing gain. It turns out global stability over entire operating region can only be obtained by switched gains, one gain for each piecewise monotonic output function. Experimental results presented performance compared with gold standard reference measurements from infrared camera capture system. An innovative technique utilizes three shown provide accuracy nearly equal four-sensor configuration.