The stability and cost-effectiveness of power supply has been a pressing issue in areas such as isolated islands where power resources are relatively scarce and natural resources is abundant. Conventional stand-alone microgrids mostly rely on the non-dominated sorting genetic algorithm(NSGA-Ⅱ) for capacity allocation

which has slightly insufficient local search capability when dealing with multi-objective optimization problems with real loads. In order to overcome this limitation

the improved strength Pareto evolutionary algorithm(SPEA2) is used to optimize the capacity allocation of wind-PV-diesel-battery stand-alone microgrid

which takes the economic cost

loss-of-load probability

and carbon emission as the optimization objectives

to achieve a more comprehensive and efficient capacity allocation. By importing the weather and load data of an isolated island and generating the real Pareto frontier of the independent microgrid with wind

PV

diesel and storage

the analysis results of SPEA2 are compared with that of multi-objective search based on indicator selection(IBEA) and NSGA-Ⅱ algorithms. Compared with the NSGA-Ⅱ algorithm

the anti generational distance evaluation IGD index of the SPEA2 increases by 46.83%

the spatial evaluation method Spacing index rises by 60.28%

and the real Pareto coverage CPF index grows by 35.14%

indicating the SPEA2 shows a more excellent performance. Finally

the parameters of each part are reasonably configured according to the results of capacity optimization. It shows that the joint output meets the load demand

which provides a new way of thinking for the energy management of isolated islands and other areas with scarce power resources

and also provides a valuable reference for the optimal design of multi-energy microgrids.