摘要
The purpose of this research is to perform an economic dispatch, formulate an optimisation model, and determine optimal operating strategies for stand-alone microgrid systems. The stand-alone microgrid is an off-grid system that is common in places where it is impossible to connect to the grid. Its economic dispatch, however, involves several constraints that impede the possibility of obtaining a feasible solution. To resolve this, the properties of the stand-alone modular microgrid mathematical model are developed to foreshorten the feasible regions of each variable. Furthermore, a new concept that involves the regeneration of a new feasible variable set is proposed. This new set replaces the infeasible variable set without the necessity of using the penalty function or regenerating entire solutions to reduce runtime. To integrate the two new foregoing concepts, which are anticipated to overcome the obstacle of infeasibility and exhibit better economic dispatch, an improved genetic algorithm is proposed. The performance of the proposed improved genetic algorithm is demonstrated through its comparison with simplified swarm optimisation, improved simplified swarm optimisation, and particle swarm optimisation on data obtained from the microgrid systems used in DongAo Island. The results show that the improved genetic algorithm, which integrates the two new proposed concepts, can handle problems with numerous constraints and is capable of obtaining a high-quality solution.