Design of Adaptive Kalman Consensus Filters (a-KCF)

This paper addresses the problem of designing an adaptive Kalman consensus filter (a-KCF) which embedded in multiple mobile agents that are distributed a 2D domain. The role such filters is to provide estimation states dynamic linear system through communication over wireless sensor network. It assumed each sensing device (embedded agent) provides partial state measurements and transmits information its instant neighbors topology. An algorithm then adopted enforce agreement on estimates among all connected agents. basis a-KCF design derived from classic filtering theorem; adaptation gain for local disagreement terms improves convergence associated difference between actual system, reducing it zero with appropriate norms. Simulation results testing performance confirm validation our design.