Online OPF Using Combined MOGA-ETS to Minimize Losses and Extend Battery Lifetime in Micro-Grid

In this paper, an Optimal Power Flow in Micro-Grid Operation is proposed. It is based on a learning algorithm combining prediction and optimization methods (Multi-objective Genetic Algorithm - Evolving Takagi-Sugeno) for implementing two objective functions i.e. minimizing losses and extending battery lifetime in online condition. This Micro-Grid operates in DC including the interest of redundancy i.e. parallel circuits for supplying loads from photovoltaic panels and batteries. The batteries use two way operations as energy generation and energy storage. It has been tested using PV power generation data and load data in a region. It is also demonstrated the comprehensive comparisons with some other learning algorithms. The results illustrate a higher online performance with optimal solution in many cases with the efficiency are higher than 97%. Moreover, reducing a high amount of CPU-time and large disk space for saving data can be achieved by the proposed approach.