Predictive load accommodation in a micro-grid

Abstract

This manuscript applies model prediction-based control (MPC) to accommodate the variable ac-natured load (AL) by optimally allocating the fuel cell (FC), battery, solar photovoltaic (SP) power in a four-port dc/dc converter (FPDDC). The presented work considers the FC, battery, SP, and AL at the first, second, third, and fourth ports of the FPDDC, respectively. The small signaled state-space representation of the FPDDC has been obtained through the improved cyclic averaged current. Then, to get the receding horizon-based MPC optimal control action, ∆u * [k] ; k = 0, 1, 2, 3, . . ., the above-mentioned model of the FPDDC is discretized. The ∆u * [k] has been obtained by formulating a performance index, Ja [k] based on an error minimization between the reference AL power demand, rpd [k] and the actual AL power, y [k]. The corresponding results have been obtained through the Simulink platform of the MATLAB. Further, the article takes sequential quadratic programming (SQP) into consideration to benchmark these solutions.