Abstract: With the large-scale integration of electric vehicles (EVs) into the power grid, the stochasticity of their charging behaviors poses formidable challenges to power system stability. Although EVs possess considerable regulatory potential, existing evaluation methods lack dynamic characterization of user choices and group evolution processes. This limitation makes it difficult to effectively evaluate the boundary domains and evolution trends of cluster schedulable potential. To address this issue, an EV user response potential evaluation method is proposed, which integrates a multidimensional semi-cloud model and evolutionary game theory. Firstly, user behavior characteristics are analyzed using clustering algorithms. Subjective factors of users are then quantified using the Decision Laboratory-Webber-Fechner algorithm. Secondly, a user response willingness model is constructed based on the semi-cloud model. Furthermore, user behavior classification is achieved through two-dimensional mapping. Finally, the dynamic impacts of compensation electricity price mechanisms are analyzed. These impacts are explicitly revealed through an evolutionary game model, focusing on user response behaviors and the schedulable potential of EV clusters. Simulation analysis is conducted based on actual charging pile data. The effectiveness of the proposed evaluation method is verified by considering both user behavior characteristics and incentives of compensation electricity prices.