Optimization Method for Reserve Capacity in New-Type Power Systems Considering the Uncertainty of New Energy Output Forecasting

With the rapid development and large-scale integration of new energy sources, the power system has shifted from single-sided load fluctuations to dual-sided uncertainties in both generation and load. Traditional deterministic methods for reserve assessment struggle to effectively account for the uncertainty of new energy output. The system needs to allocate reasonable and effective reserve capacity to ensure its safe and stable operation. To address this, this paper proposes a dynamic reserve demand assessment model based on feature clustering, non-parametric estimation, and convolution algorithms. The model uses historical data of wind power and load forecast errors as a basis, and employs an improved clustering algorithm to effectively consider the impacts of seasons, time periods, and new energy output on reserve capacity. Through kernel density estimation and convolution algorithms, the model derives the joint probability distribution function of forecast errors under various influencing factors, enabling dynamic assessment of reserve demand affected by system uncertainties in different scheduling periods. Subsequently, a security-constrained unit commitment model considering wind power uncertainty is established. The paper compares and analyzes the operational impacts of static and dynamic reserve demand assessment methods, and validates the effectiveness of the proposed model. The results show that the model is effective in enhancing the integration of new energy and reducing generation costs.