Optimal Economic Modelling of Hybrid Combined Cooling, Heating, and Energy Storage System Based on Gravitational Search Algorithm-Random Forest Regression

The hybridization of two or more energy sources into a single power station is one of the widely discussed solutions to address the demand and supply havoc generated by renewable production (wind-solar/photovoltaic (PV), heating power, and cooling power) and its energy storage issues. Hybrid energy sources work based on the complementary existence of renewable sources. The combined cooling, heating, and power (CCHP) is one of the significant systems and shows a profit from its low environmental impact, high energy efficiency, low economic investment, and sustainability in the industry. This paper presents an economic model of a microgrid (MG) system containing the CCHP system and energy storage considering the energy coupling and conversion characteristics, the effective characteristics of each microsource, and energy storage unit is proposed. The random forest regression (RFR) model was optimized by the gravitational search algorithm (GSA). The test results show that the GSA-RFR model improves prediction accuracy and reduces the generalization error. The detail of the MG network and the energy storage architecture connected to the other renewable energy sources is discussed. The mathematical formulation of energy coupling and energy flow of the MG network including wind turbines, photovoltaic (PV), CCHP system, fuel cell, and energy storage devices (batteries, cold storage, hot water tanks, and so on) are presented. The testing system has been analysed under load peak cutting and valley filling of energy utilization index, energy utilization rate, the heat pump, the natural gas consumption of the microgas turbine, and the energy storage unit. The energy efficiency costs were observed as 88.2% and 86.9% with heat pump and energy storage operation comparing with GSA-RFR-based operation costs as 93.2% and 93% in summer and winter season, respectively. The simulation results extended the rationality and economy of the proposed model.

Automated summary

This paper discusses the economic model of a hybrid energy system, focusing on the combined cooling, heating, and power (CCHP) system within a microgrid (MG). By optimizing the random forest regression (RFR) model using the gravitational search algorithm (GSA), prediction accuracy is improved and generalization error is reduced. The analysis includes energy utilization indices, cost-effectiveness, and energy efficiency comparisons under different operational conditions.