Transient Stability-Constrained Optimal Power Flow Model Considering Wind–Solar Output Correlation

To address the challenges of wind–solar output correlation, renewable-output uncertainty, transient stability, and economic optimization, this paper proposes a transient stability-constrained optimal power flow (TSCOPF) model considering wind–solar correlation. First, kernel density estimation (KDE) is employed to establish the marginal probability density functions of wind and photovoltaic outputs, and a Frank-Copula function is used to characterize the wind–solar correlation and construct a joint probability distribution model. A Wasserstein generative adversarial network with gradient penalty (WGAN-GP) is then used to generate wind–solar output scenarios, which are further reduced by K-means++ clustering. Second, a transient stability assessment method combining a graph convolutional network with attention mechanism (GCN-Attention) and conditional mutual information (CMI)-based feature selection is developed to extract key stability features, and a TSCOPF model considering renewable-energy integration is constructed. Third, an improved Coati Optimization Algorithm (ICOA) integrating refraction-based opposition learning, Levy flight, and spiral search strategies is proposed to enhance global optimization performance. Simulations on the modified Institute of Electrical and Electronics Engineers (IEEE) 39-bus system and the IEEE 118-bus system demonstrate the accuracy, effectiveness, and scalability of the proposed method.