Self-Tuning Extended Kalman Filter Parameters to Identify Ankle's Third-Order Mechanics.

The estimation of the human ankle's mechanical impedance is an important tool for modeling human balance. This work presents the implementation of a parameter-estimation approach based on the State-Augmented Extended Kalman Filter (AEKF) to infer the human ankle's mechanical impedance during quiet standing. However, the AEKF Filter is sensitive to the initialization of the noise covariance matrices. In order to avoid a time consuming trial-and-error method and to obtain a better estimation performance, an algorithm based on Genetic Algorithms (GA) is proposed for tuning the measurement noise Rk and process noise covariances Q of the Extended Kalman filter (EKF). Results using simulated data show the ef?cacy of the proposed algorithm for parameter-estimation of a third-order biomechanical model. An experimental test with real data on human subjects is also presented. The results suggest that age is a factor that influences human balance capability.