CMOPEO-OP: Constrained multi-objective population extremal optimization-based optimal planning of standalone microgrids

A standalone microgrid is crucial for delivering electricity to residential and remote regions by combining various distributed generation sources. A major challenge in academia and engineering is developing an optimal planning approach for distributed generations and storage system in a standalone microgrid, taking into account multi-objective performance metrics like reliability, cost and environment. In this paper, we propose a novel constrained multi-objective population extremal optimization-based optimal planning approach for standalone microgrids, called CMOPEO-OP. We first transform the optimal planning problem into a constrained multi-objective optimization problem, where annualized cost of the system, loss of power supply probability (LPSP), and fuel emissions are considered as the three objective functions to be minimized simultaneously. To ensure the reliability of system operation, we use LPSP as a constraint. The decision variables include the number of components in standalone microgrids, e.g., photovoltaics (PV) panels, wind turbines, batteries, and diesel generators, the height of the wind turbine tower, and the inclination angle of PV panels, where the first four are integer variables and the last two are real variables. Then, to effectively and efficiently deal with this problem, we propose the ɛ adaptive trade-off model (ɛ-ATM) to deal with constraint and combine a population extremal optimization mechanism. In the search process, due to the introduction of ɛ, ɛ-ATM mechanism can better balance the relationship between constraints and objective values. Additionally, an elaborated hybrid mutation operation and an external archive update mechanism are developed in CMOPEO-OP. The performance of the CMOPEO-OP method is illustrated on a standalone microgrid by comparing with multi-objective particle swarm optimization-based optimal planning (MOPSO-OP) method and multi-objective differential evolution-based optimal planning (MODE-OP) method. The experimental results show that the CMOPEO-OP method has better performance than MOPSO-OP and MODE-OP in terms of three performance metrics, i.e., hypervolume indicator, spacing metric, and inertia-based diversity metric, and planning solutions for standalone microgrids.