A distributed and cooperative supervisory estimation of multi-agent nonlinear systems

In this work, we propose a framework for supervisory cooperative estimation of multi-agent nonlinear systems. We introduce a group of sub-observers, each estimating certain states conditioned on certain given input, output, and state information. The cooperation among the sub-observers is supervised by a discrete-event system (DES). The supervisor makes decisions on selecting and configuring a set of sub-observers, so that the overall integrated sub-observers are able to successfully estimate all the states of the system. In cases when certain changes in the uncertainties take place, the supervisor reconfigures the set of selected sub-observers so that the impact of these uncertainties on the estimation performance is minimized. Our proposed method is applied to a nonlinear industrial process, and the simulations results obtained validate our analytical work.