Optimasi Dan Implementasi Kapasitas Daya Terpasang Energi Terbarukan Pada Sistem Kelistrikan Sumatera Utara Menggunakan Metode Firefly Algorithm (FA), Particle Swarm Optimization (PSO) dan Artificial Neural Network (ANN)

Abstract

Increasing environmental awareness and the need to reduce dependence on fossil fuels have driven efforts towards a transition to renewable energy (RE). This study aims to develop an optimised energy mix model for North Sumatra by maximising RE contribution through a predictive approach using Artificial Neural Network (ANN) Backpropagation, combined with the Firefly Algorithm (FA), Particle Swarm Optimisation (PSO), and additional validation using the Genetic Algorithm (GA). The hybrid model incorporates micro-hydro, hydro, geothermal, biomass, biogas, and photovoltaic energy sources, with data collected from the State Electricity Company (Perusahaan Listrik Negara/PLN) in North Sumatra, Independent Power Producers (IPPs), and the palm oil industry. Evaluation is conducted based on the Renewable Energy Contribution Ratio (RECR). The novelty of this research lies in the integration of ANN predictive modelling with three metaheuristic optimisation methods for regional energy mix scenarios, and the use of RECR as a holistic performance indicator to assess the effectiveness of RE contribution. The results show that FA outperforms the other methods, followed by GA and PSO, with lower absolute RECR values. The predicted timeframe for a full transition to renewable energy is 2064 (FA), 2065 (PSO), and 2064 (GA). These findings provide a scientific foundation for the development of a sustainable energy transition roadmap and offer policy insights to support regional energy security and climate change mitigation.