Detection, estimation, and accommodation of loss of control effectiveness

In this paper, an adaptive Kalman filtering algorithm is developed for use to estimate the reduction of control effectiveness in a closed-loop setting. Control effectiveness factors are used to quantify faults entering control systems through actuators. A set of covariance-dependent forgetting factors is introduced into the filtering algorithm. As a result, the change in the control effectiveness is accentuated to help achieve a more accurate estimate more rapidly. A weighted sum-squared bias estimate is defined for the change detection. The state estimate is fed back to achieve the steady-state regulation, while the control effectiveness estimate is used for the on-line tuning of the control law. A stability analysis is performed for the adaptive regulator.