Random delayed Kalman fusion based on equivalent conversion for the multisensor system

Aiming at the linear time invariant or the parameters preliminary given for the multisensor target tracking system, this paper develops a random delayed Kalman filter fusion estimator based on equivalent conversion for multisensor system. This algorithm effectively uses the characteristics of Kalman filter statistical parameters which can be calculated out of line, and the calculated form of measurements weighted summation under Linear Minimum Mean Square Error (LMMSE) estimate. Firstly, it transformed the multisensor system into the single sensor order form based on the idea of remolding. Secondly, it utilizes the method of one step prediction estimate and measurements prediction residual compensation to get the optimal weighting coefficient though calculating out of line and adjusting online. Then, it can realize the optimal update of the random delay measurements. At last, Algorithm analysis and computer simulation indicates that this algorithm is validity and advantage.