Multi-objective Optimization For Microgrid System Sizing Based On Levy-Harmony Searching Algorithm

Rural and island areas are generally far away from the national power grids,and the environments are usually uncertain and complex.Thus it is challenging to provide them with stable power supply.On the other hand,the electricity consumption of these areas is very small,while the cost of directly connecting them to the national power grid is too high and not practical.Note that there are rich natural resources in rural areas,so renewable energy and distributed generation are preferable to be adopted to build up the island microgrid systems(IMS)with wind,diesel and storage.This thesis presents a multi-objective optimization design method for island microgrids based on a novel Levy-Harmony Searching Algorithm.The optimization is achieved by improving the sizes configuration of wind turbines,photovoltaic,diesel generator capacity and energy storage,with comprehensive and balanced performance of the whole system.The main contributions of this work are as follows:Traditional island optimization design models use insufficient optimization indexes,so it is hard to reflect the comprehensive system performance.Based on reviewing and analyzing the existing methods of microgrids optimization,we select nine typical indexes for describing the system performances and constraints,and use them for sizing optimization.Most multi-objective fitness functions rely on prior knowledge and are usually hard to determine.Moreover,they are not applicable to the cases with many optimization indexes.Considering the fact that there exist some coupled conflicts between various types of indexes of microgrids,this thesis presents a triangle aggregation model for multi－objective fitness computation of island microgrids.This model has simple structure and is easy for realization.It relies less on empirical data and complex calculation.Moreover,the optimization indexes are placed equal priorities,so the conflicts can be well addressed and the comprehensive system performance can be balanced.To address the problem of existing optimization algorithms that there are too many parameters and confilcts between searching ability and convergence speed,this thesis blends the bias searching factor with the Levy flight mechanism,and formulates an improved bio-inspired approach named Levy-Harmony searching algorithm.It is found that adopting the Levy flight can successfully strengthen the global search ability of the algorithm,while adding the bias searching factor into the algorithm helps avoid repeated and unnecessary searching.Consequently,the searching efficiency convergence speed of the algorithm are effectively improved and better balanced.