Research on PV Electricity Supply and Carbon Emission Forecasting Based on Multidimensional Indicators

Against the background of the global energy transition and the "dual-carbon" goal, this paper investigates the power supply prediction and carbon emission prediction of photovoltaic (PV) power generation. First, the data were preprocessed through outlier processing, missing value supplementation, and data dimensionality reduction according to 18 indicators in four categories: economy, energy, population, and environment. Secondly, based on the cubic polynomial regression model, the electricity production of PV power generation is predicted from 2025 to 2060. The carbon emission prediction model is further constructed to verify the key role of photovoltaic power generation in the "dual-carbon" strategy. By integrating hybrid dimensionality reduction and dynamic optimization algorithms, the efficiency of solving complex nonlinear problems is improved, and the combination of ridge regression and heat map analysis reflects the multivariate interaction mechanism and visualizes the strength of variable association.The photovoltaic power supply and carbon emission forecasts presented in this paper provide some data-driven decision support based on regional energy planning and sustainable development.