An online stochastic MPC-based fault-tolerant optimization for microgrids

Abstract

Due to the current energy dependence of society, the availability and correct functioning of microgrids are strategic issues to be dealt with. Currently, energy management systems are very focused on achieving systems that present a remarkable optimization of demand as well as a high degree of fault tolerance. This paper presents a novel Control Reconfiguration framework to manage faults based on Model Predictive Control (MPC). The proposal comprises fault detection, isolation and reconfiguration. The Fault Detection and Isolation method identifies the faults based on parity equations, structured residuals and stochastic thresholds, while the reconfiguration is focused on adapting the control law to the faulty scenario. Therefore, two different model predictive controllers are involved: MPC-1 drives the microgrid to the correct values and MPC-2 carries out the control reconfiguration, optimizing a multi-criteria objective function where outputs are the values of criteria. The new decision variables of the fault reconfiguration are the selection of mitigation actions to be performed to reduce the effects of faults. A novel formulation of this problem is provided. Experiments have been carried out on a real microgrid located in the laboratory to show the benefits of the method.