Impact of scaled fitness functions on a floating‐point genetic algorithm to optimise the operation of standalone microgrids. Issue 8 (20th March 2019)

• Record Type: Journal Article
• Authors: Batool, Munira; Shahnia, Farhad; Islam, Syed M.
• Is Part Of: IET renewable power generation. Volume 13:Issue 8(2019)
• Page Start: 1280
• Page End: 1290
• Publication Date: 2019-03-20
• Subjects: distributed power generation -- genetic algorithms -- floating point arithmetic -- power markets -- power generation economics -- power generation reliability -- costing -- nonlinear programming; standalone hybrid remote area power systems -- floating‐point genetic algorithm -- FP‐GA -- optimal control variables -- nonlinear optimisation problem -- scaled fitness functions -- standalone microgrids -- MG -- electricity pricing -- fossil fuels; Renewable energy sources -- Periodicals; 333.79405
• DOI: 10.1049/iet-rpg.2018.5519
• Languages: English
• ISSNs: 1752-1416

Abstract

Abstract : Standalone hybrid remote area power systems, also known as microgrids (MGs), can provide reasonably priced electricity in geographically isolated and the edge of grid locations for their operators. To achieve the reliable operation of MGs, whilst consuming minimal fossil fuels and maximising the penetration of renewables, the voltage and frequency should be maintained within acceptable limits. This can be accomplished by solving an optimisation problem. Floating‐point genetic algorithm (FP‐GA) is a heuristic technique that has a proven track record of effectively identifying the optimal solutions. However, in addition to needing appropriate operators, the solver needs a fitness function to yield the most optimal control variables. In this study, a suitable fitness function is formulated, by including the operational, interruption and technical costs, which are then solved with an FP‐GA, with different combinations of operators. The developed fitness function and the considered operators are tested for the non‐linear optimisation problem of a 38‐bus MG. Detailed discussions are provided on the impact, which different operators have upon the outcomes of the fitness function.