Adaptive Back-Propagation Neural Network and Particle Swarm Optimization-Based Approach for Optimizing the Output Power Biogas Fueled Electric Generator

The utilization of biogas contributes significantly to the reduction of greenhouse gas emissions and diminishes the dependency on traditional fossil fuels. Nonetheless, the consistency and optimization of electricity generation through a biogas generator are hindered by fluctuations in key input variables, namely biogas input pressure, methane (CH4) content, and the flow rate of biogas. Hence, the optimization of these influential input parameters is crucial for enhancing the power output of the electricity generator. The primary objective of this study was to refine the input parameters to maximize the electrical power output generated by the biogas electricity generator. The data collected from the monitoring system were analyzed using an artificial intelligence (AI) model, specifically an Artificial Neural Network (ANN). This study introduces an enhanced ANN model that incorporates an Adaptive Backpropagation Algorithm to ensure best practices in modelling optimization. Furthermore, the utilization of Particle Swarm Optimization (PSO) aids in determining the optimal values for each variable affecting the output power of the electric generator. The configuration of the multilayer perceptron model, combined with the Adaptive Backpropagation Algorithm and PSO, establishes the fundamental framework for the proposed advancements. The findings reveal that under the optimal conditions of biogas input pressure at 34.36 kPa, CH4 content at 90%, and biogas flow rate at 0.22 l/s, the output power of the electric generator can achieve levels of up to 1.42 kW, with a notably high accuracy testing value of 0.98513. This study highlights the potential benefits of employing adaptive neural network models and optimization methodologies to determine the optimized operational parameters and accurately predict the output power of a biogas-fueled electric generator.