1.Key Laboratory of Power System Intelligent Dispatch and Control of Ministry of Education (Shandong University), Jinan 250061, China;2.College of Electrical Engineering, Sichuan University, Chengdu 610065, China;3.China Electric Power Research Institute Co., Ltd., Beijing 100192, China

The untrustworthiness issues of artificial intelligence (AI) algorithm hinder its practical application in the scenarios such as power grid stability analysis and control. At present, there are no specific trustworthiness evaluation indicators applicable to the intelligent assessment model for power grid stability. Regarding the characteristics of the electric power industry, the trustworthiness evaluation for AI stability assessment models is carried out, and the five sub-indicators of correctness, complexity, robustness, transferability and interpretability are chosen to construct the trustworthiness evaluation indicator system of the intelligent assessment model for power grid stability, and the specific calculation methods for each sub-indicator are summarized. Meanwhile, the fuzzy analytic hierarchy process is introduced and combined with subjective and objective evaluation to determine the weights of each sub-indicator and calculate the comprehensive trustworthiness indicator. Finally, by taking the voltage support strength, broadband oscillation, and frequency stability assessment models as a case, the trustworthiness evaluation and analysis are carried out, and the results verify the effectiveness of the proposed method.