Enhancing economic and performance effectiveness of a geo-solar system using multi aspect artificial neural network

Abstract

Geo-solar systems, comprising ground-source heat pumps (GSHP) and solar thermal collectors, offer promising solutions for sustainable space heating and hot water provision in residential buildings. However, optimizing these systems for maximum energy efficiency and cost-effectiveness remains a challenge. In this context, this study presents a comprehensive approach to enhancing the performance and cost-effectiveness of geo-solar systems. The proposed methodology integrates artificial neural network (ANN) techniques and multi-aspect optimization to optimize system parameters. The detailed of building mathematical model incorporating energy and exergy equations, thermodynamic, and economic relations for all system components. The neural network model inputs to the include five neurons represents Geothermal temperature, Solar radiation levels, System pressure levels, Ambient temperature, and Operational time, and the model outputs represent Energy efficiency and Exergy destruction rate. Through the training of ANN using simulation data and the application of a multi-objective genetic algorithm, the system is optimized for energy efficiency, cost, and environmental impact simultaneously. The results of analysis show that the exergy efficiency of system is 22.02 % and turbine generates 1963 kW. Changes in P10, ranging from 2000 to 3400 kPa leads to the exergy destruction reduction from 8350 kW to 8000 kW, while the other parameters exhibit an increasing trend.