Research on integrated energy system planning based on the correlation between wind power and photovoltaic output

Abstract

The research on the randomness and volatility of wind power (WP) and photovoltaic (PV) output of the integrated energy system (IES) has emerged as a pivotal concern, commonly dealt with by clustering techniques. However, traditional clustering techniques often fall short of capturing the comprehensive characteristics of the original scenery data. This paper presents an enhanced clustering logic to improve WP and PV output correlation scenarios. Utilizing the Frank-Copula function, the complex relationship between WP and PV is accurately described. The Kendall rank correlation coefficient (Kendall′s tau) is used as a metric to study the correlation between these two renewable energy sources. The original scenes' correlation coefficients are then clustered using the K-means algorithm, which forms the basis for scene reduction. On the basis, a two-phase robust optimization method is employed to conduct a planning study for a real IES, with the total cost of the system as the optimal objective. This approach ensures that the planning results are better aligned with the specific local conditions during the operation cycle.