Smart Grid Production Cost Optimization by Bellman Algorithm

Abstract

This study focuses on the application of the Bellman algorithm, one of the algorithms of dynamic programming for energy management in a microgrid with five distributed energy sources. The management strategy of the microgrid is based on long-term planning. The latter requires that each of the distributed sources follow an optimal active power profile established the day before for the next day (D-1). Thus, the optimization strategy proposed in this study belongs to the deterministic optimization strategies, whose objective is to find the optimal control of power flows in the multi-sources micro grid. The optimal control means how to distribute the power flows among the different components of the microgrid in order to ensure the balance between production to demand with the lowest cost. The performance of dynamic programming is relative to the refinement of the discretization of the planning trajectory and reference signals. Therefore, the quality of the optimum found is sufficiently guaranteed at the expense of a consequent demand on computing and memory capacity. The aim of this study is to establish a management by the application of the Bellman algorithm (dynamic programming) and to compare it with the strategy of the rules-based management to open later on other meta-heuristic techniques such as the genetic algorithm. The obtained results showed that the implementation of the proposed energy management strategy resulted in a 40% saving on the electricity bill.