用于多源信息中制导的MM-LS时间配准算法

针对目标做匀速直线运动以及时间配准周期与雷达采样周期成整数比两个假设限制了最小二乘时间配准算法应用范围的问题，结合多模型的思想，研究了无限制条件下的多模型最小二乘时间配准算法．在推导匀加速模型、加加速模型最小二乘时间配准算法的基础上，由最小均方误差准则得出多模型最小二乘时间配准算法．利用各模型求解得到的一阶、二阶、三阶导数，进一步通过外推算法，解决了时间配准周期与雷达采样周期比为非整数这一问题．仿真表明，当目标的机动过载从1g至9g变化时，雷达的探测误差从10m至100m变化，多模型最小二乘时间配准算法虽然不能取得最好的结果，但是可以确保得到值得信赖的时间配准结果．

MM-LS time registration algorithm in middle guidance using multi-sensor information

The application of LS time registration has been restricted by both the assumption of the uniform motion of the target and the assumption that the ratio of the registration period to the sensor sampling period is an integer. The MM-LS time registration algorithm under no restriction has been studied in this paper. Based on the model of the uniform accelerated motion and accelerated linear motion, the MM-LS time registration algorithm has been obtained with the rule of MMSE. With the first order, second order, and third order derivative achieved through each model, the problem of non-integer ratio is solved by the extrapolating algorithm. Simulations indicate that the performance of MM-LS is always reliable when the maneuver overload of the target changes from 1g to 9g, and that the precision of radar changes from 10m to 100m.