A predictive control algorithm for multi-sensor information fusion weighted by matrices

Aiming at the multi-sensor discrete-time linear time-invariant stochastic controllable system based on state space model,a predictive control algorithm for multi-sensor information fusion weighted by matrices is presented using the Kalman filtering method under the linear minimum variance optimal information fusion criterion.This algorithm combines the information fusion of Kalman filter with predictive control,avoiding to solve the complex Diophantine equation,and therefore,it can reduce the computational burden obviously.Comparing to the case of a single sensor,the accuracy of the predictive control has been evidently improved.A simulation example of a 3-sensor target-tracking controllable system shows that it is effective and correct.