Against the backdrop of striving to achieve “carbon peak” and “carbon neutrality”, virtual power plants (VPPs) have become an important direction for the development of the energy system due to their advantages in increasing the penetration rate of renewable clean energy. The power generation of renewable energy systems has strong volatility and intermittency, leading to uncertainty in their power prediction. Based on the above background, this article proposes a VPP operation optimization model that adapts to climate change using the Long Short Term Memory (LSTM) neural network prediction model and interval optimization algorithm. Coupling interval optimization algorithm with power generation prediction results to reduce the impact of PV power generation uncertainty on the optimization model results. The results show that the model can generate the optimal operating strategy for VPP that adapts to climate change, reduce system operating costs, and improve VPP operating efficiency.