A Consensus-based Filtering Algorithm with State Constraints

This paper proposes a consensus-based filtering algorithm to solve a nonlinear information fusion problem in multi-agent system that the state vector of the target is constrained. This method is based on the Kalman-Consensus filtering algorithm, Extended Kalman filter and null space method. Simulations on target tracking prove the feasibility of the algorithm in dealing with state estimation of nonlinear systems under linear and nonlinear constraints. It is demonstrated that a significant increase in the tracking precision can be expected via the proposed method compared with traditional Kalman-Consensus filter.