Advanced Optimization of Artificial Neural Networks for a Syngas Power Plant Using Bio-Inspired Optimization Algorithms: : Comparison Between PSO, GA and a Novel Mosquito Inspired Optimization Algorithm

The integration of renewable energies sources into hybrid distributed generation systems presents significant challenges, particularly when predicting critical. Syngas power plants are often integrated into hybrid microgrids to produce electricity instead of fossil fuel-based backup generators. This work proposes an advanced optimization of artificial neural network models applied to the prediction of key parameters for a syngas-based power plant. As part of the proposed approach, a novel Mosquito Mating Swarm Optimization (MMSO) algorithm is presented. The MMSO is compared with Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) in terms statistical performance indexes and linear regression. The results show that MMSO outperforms PSO and GA in biomass and air flow parameter prediction for the syngas power plant in accuracy and stability for variable operating conditions. The improvement in parameter prediction through the MMSO contributes to the development of robust and efficient tools for the adaptable management of microgrids and hybrid distributed generation systems.