Reactive power control of hybrid systems using improved coyote optimizer

• Published in: Concurrency and computation Vol. 35; no. 28
• Main Authors: Chen, Lin, Yi, Xianzhong, Zhou, Yuanhua, Liu, Lijun, Liu, Hangming, Razmjooy, Saeid
• Format: Journal Article
• Language: English
• Published: Hoboken, USA John Wiley & Sons, Inc 25.12.2023; Wiley Subscription Services, Inc
• Subjects: Algorithms; Energy costs; fuel cell; Fuel cells; Heuristic methods; Hybrid systems; Hydrogen; improved coyote optimizer; Matching; Optimization; photovoltaic; Power control; Reactive power; reactive power optimization; Weather
• ISSN: 1532-0626; 1532-0634
• DOI: 10.1002/cpe.7859

Summary A new method of reactive power control by photovoltaic and hydrogen is presented in this study. Photovoltaic has been employed for harvesting the hydrogen which is based on considering the weather conditions. The proposed system includes a combination of photovoltaic, hydrogen, and fuel cell along with a DG to connect to the grid and to improve the supply power quality. The main contribution of this paper is to direct an improved metaheuristic algorithm, called improved coyote optimization (CO) algorithm for achieving a proper DG placement. The improved version of the CO algorithm has benefitted from a spiral policy that is derived from Whale optimization algorithm. This process makes better control for the social behavior of the coyotes. Reactive power optimization (RPO) has been established after the size selection objective function. Big data technology is also used for improving the historical solution matching‐based RPO appliance. Cosine distance is used for measurement purposes of the historical solution matching‐based similarity technique during the computation time of conventional RPO and PVH‐FC features. As a result of using the suggested CO, the costs for electricity and losses are reduced by around 86.6% and 26.9%, respectively. Additionally, realized profits showed that applying the suggested strategy reduced the overall cost from 9.315e6 to 4.435e6 units. After optimization, the network loss and power are finally reduced to 1325 Kvar and 1371 kW, respectively. The results show that the suggested RPO technique has higher speed of calculation in comparison with some latest algorithms. Achievements also show that the suggested method provides a proper and optimal solution for RPO.