Sensor less Finite Control Set Model Predictive Control for Stabilizing DC-Grid Voltage in Islanded DC-Microgrids

Using a linear Kalman filter approach(LKFA), this study proposes sensor-less finite-control-set model predictive control for stabilizing DC-grid voltage estimation in direct current microgrids (DCMs). In controlling many parallel power converters in DCMs, the proposed control method eliminates the need for an inductance current sensor. The second prediction is derived from the state-space model so that an efficient algorithm may be performed. The dynamic model of DCMs in the second prediction is converted into a stationary linear stochastic discrete time-invariant system to standardize the state estimation design. The DC-grid voltage reference is then computed adopting LKFA using a state feedback control law based on the dynamic algebraic Riccati equation with integral action to guarantee zero steady-state error during transient responses. Under load demand variation, the efficacy of the proposed approach is shown using a sensor-less control algorithm. Robustness against parametric uncertainty, DC-grid voltage estimation, and equally estimated power-sharing amongst DERs have all been achieved, as have the fast convergence.