Abstract: The sufficient conditions for the filter's stability and an analytical measure for performance improvements are given.based on an analysis of the effect of the kinematic constraint on system's stochastic observability.For the stochastic system of the dynamic equation and the kinematic constraint,the upper and lower bounds on its information matrix are given,via eigenvalue analysis.It is shown that the system remains its uniformly complete controllability and observability after incorporating the kinematic constraint,which provides the sufficient conditions for the filter's stability.The decrements in the variances of the estimation errors are used to measure the performance improvements.Monte Carlo simulation verifies the conclusion of theoretic analysis.