A Hybrid Fitness‐Guided Starfish Optimiser for Stochastic Allocation of Photovoltaic Energy Sources in Distribution Systems

ABSTRACT This paper develops a hybrid fitness‐guided starfish optimiser (HF‐SFO) for stochastic photovoltaic‐distributed generation (PV‐DG) integration in radial distribution networks. Hybridised with the point estimate method (PEM), the proposed HF‐SFO models load demand and PV uncertainty using a reduced scenario matrix of five cases, each with assigned probabilistic weights. It significantly lowers computational complexity compared to Monte Carlo simulation while preserving accuracy, enabling efficient integration of stochastic uncertainties of PV‐DG and load into backward–forward sweep load flow and HF‐SFO optimisation processes. For each PEM scenario, the expected power loss is formulated as a weighted aggregation of the active power losses and evaluated while their associated stochastic power flows and voltage profiles are maintained. It is implemented on the 15 kV Portuguese feeder of 94 nodes and the 11 kV radial feeder of an urban area in India of 108 nodes. The proposed HF‐SFO achieves up to 17.7% reduction in total power losses and improves the minimum bus voltage by up to 0.74% compared with the others under high PV‐DG penetration scenarios. Statistical validation using Friedman and Wilcoxon tests confirms the superiority of HF‐SFO with mean rank improvements exceeding 70% over others and consistently lower confidence interval bounds across both test systems.