Rural Renewable Energy Resources Assessment and Electricity Development Scenario Simulation Based on the LEAP Model

This study combines convolutional neural network (CNN) recognition technology, Greenwich engineering software, and statistical yearbook methods to evaluate rural solar, wind, and biomass energy resources in pilot cities in China, respectively. The CNN method enables the rapid identification of the available roof area, and Greenwich software provides wind resource simulation with local terrain adaptability. The results show that the capacity of photovoltaic power generation reaches approximately 15.63 GW, the potential of wind power is 458.3 MW, and the equivalent of agricultural waste is 433,900 tons of standard coal. The city is rich in wind, solar, and biomass resources. By optimizing the hybrid power generation system through genetic algorithms, wind energy, solar energy, biomass energy, and coal power are combined to balance the annual electricity demand in rural areas. The energy trends under different demand growth rates were predicted through the LEAP model, revealing that in the clean coal scenario of carbon capture (WSBC-CCS), clean coal power and renewable energy will dominate by 2030. Carbon dioxide emissions will peak in 2024 and return to the 2020 level between 2028 and 2029. Under the scenario of pure renewable energy (H_WSB), SO2/NOx will be reduced by 23–25%, and carbon dioxide emissions will approach zero. This study evaluates the renewable energy potential, power system capacity optimization, and carbon emission characteristics of pilot cities at a macro scale. Future work should further analyze the impact mechanisms of data sensitivity on these assessment results.