Robust decentralized control of interval DC linked hybrid microgrids network

Here, a robust sparse decentralized controller (RSDC) for uncertain DC-linked hybrid microgrids is designed. The robustness against load/generation uncertainty is guaranteed using Kharitonov's stability criteria and sparsity is promoted by appending an l(1) term to the typical Linear Quadratic Regulator (LQR). To improve the convergence rate, the differentiable term of the composite cost function is second order approximated, the coordinate descent technique is adopted, and the "active set" idea which limits the search to the effective regions is employed. In addition, the level of sparsity is automatically and optimally adjusted not to compromise the system's stability and robustness. The algorithm is applied to a 69-bus (consist of 7 hybrid microgrid) distribution system and the controller sparsity percentage, design convergence rate, and performances are compared with the H-2 and H-infinity designs. The results indicate that the proposed method is superior to the others in all aspects, which are presented by the numerical indices, the time-domain transient responses, and the closed-loop poles map.