Performance comparison of heuristic algorithms for optimization of hybrid off-grid renewable energy systems

Abstract

Hybrid renewable energy systems have been widely acknowledged as a clean, affordable and reliable mechanism to generate electricity and to accomplish global sustainable development goals. In this study, first, an operating strategy and an optimization problem are developed for a hybrid, off-grid, solar-wind system based on pumped hydro battery storage, and then a non-linear optimization problem is described for the considered system. To solve the optimization problem, four different optimization techniques are employed i.e. ant colony (ACO), firefly algorithm (FA), particle swarm optimization (PSO) and genetic algorithm (GA) and their performance is compared using statistical parameters like relative error, mean absolute error and root mean square error. Each optimization technique’s working principle is discussed in detail and formulated considering the proposed optimization problem. The exploration and exploitation behavior of each algorithm is comprehensively analyzed explaining that ACO and FA have higher exploitation behavior, while GA and PSO have more exploration behavior, revealing that these behavior depend on the range of operator controlling parameters, type of optimization problem and formulation structure of the optimizers. The reference controlling parameters of each optimizer (which are operator dependent) are defined for the proposed optimization problem. The results reveal that FA performs better – i.e. with the least relative error (0.126) – while PSO outperforms best in terms of least objective function value (0.2435 $/kWh). The mean efficiency of each algorithm in terms of repeated executions (30 times) is ACO = 95.94%, FA = 96.20%, GA = 93.93%, PSO = 96.20%. The proposed study could help decision-makers to choose an optimization method to solve non-linear problems in the context of storage-based, off-grid systems under different scenarios.