Operation Mode Identification for High Renewable Energy Penetration Power Systems Based on Planning–Operation Coupling and SHAP Interpretability

Under high renewable energy penetration, power system operation modes exhibit significant diversity and structural complexity, posing challenges to traditional scenario‐based analysis methods in capturing their formation mechanisms. This paper proposes an interpretable analytical framework for operation mode identification and driving mechanism analysis. Based on year‐round operational data, clustering is employed to label similar operation states, while the core analysis focuses on a decision tree model combined with the SHapley Additive Explanations (SHAP) method to automatically identify key driving factors and quantify their contributions. A robustness verification strategy is further developed through clustering parameter perturbation and model simplification comparison, ensuring the stability and explanatory adequacy of the identified factors. Case studies on both the HRP‐38 benchmark test system and a real‐world 154 node power system demonstrate that, as renewable energy penetration increases, operation modes become more numerous and structurally complex, with their formation mechanism shifting from load‐ and thermal‐dominated patterns to renewable–storage coordinated structures. The proposed framework provides accurate mode identification while enabling clear and physically interpretable insights into their formation mechanisms, offering practical support for operation analysis and planning‐oriented decision‐making in high‐renewable power systems.