Copula-Based Scenario Generation for Future Annual Load Time Series

To address the limitation of existing annual load time series scenario modeling methods in characterizing overall load variations and supporting new energy accommodation assessment, this paper proposes a Copula-based scenario generation approach for future annual load time series. First, considering the inherent correlations among daily load factor, daily peak-valley difference rate, and daily maximum load, a three-dimensional joint distribution model is constructed via Copula theory, from which future daily load characteristic indices are generated through random sampling and inverse CDF transformation. Subsequently, with typical daily load profiles extracted by K-means clustering as morphological benchmarks, the generated indices are incorporated as equality constraints into a linear programming model to minimize morphological deviations. The optimized daily sequences are then concatenated chronologically to form a complete annual scenario. Simulation results based on actual load data from a provincial power grid demonstrate that the scenarios generated by the proposed method outperform those from the benchmark method in probability distribution, autocorrelation, and monthly/annual indices, validating the effectiveness and practicality of the proposed approach.