A distributed set-membership estimator for linear systems with reduced computational requirements

AbstractAbstractIn this paper, a distributed set-membership estimator for linear full-coupled systems affected by bounded disturbances is presented. The estimator uses a multi-hop staircase decomposition, capturing the locally unobservable subspaces in a cascaded fashion with the information incoming from other agents involved in the network. Each agent has to find different sets for each subspace, that are mathematically described by zonotopes. The observer gains that minimize the size of those sets, i.e. the estimation uncertainty, can be designed in independent distributed steps by means of simple algebraic equations. Simulations are given to compare the proposed solution with others in the field. An important benefit of the proposed structure is the reduction of the computational requirements with respect to existing solutions.