The traditional power system monitoring and control methods still have some problems, such as difficult data processing, insufficient robustness, excessive manual intervention and insufficient adaptability. By introducing artificial intelligence (AI) algorithm, this paper aimed to improve the robustness and data processing ability of power system and realize automatic intelligent control to adapt to the modern challenges of power system. In order to meet the need of large quantity and high quality marked image data during the actual training of power system intelligent monitoring method (IMM), this paper constructed the power system intelligent monitoring image data set. In order to analyze the IMM of power system based on AI technology, the E-R Model (Entity-Relationship Model) database was established, and the variational mode decomposition method was used to extract the abnormal state characteristics of power system equipment in the image. The feature was input into support vector machine (SVM) to realize real-time recognition of abnormal state of power system equipment. This paper also analyzed an efficient single-agent reinforcement learning algorithm, Q-learning, to achieve the optimal coordinated control of power system. In this paper, the power system equipment of a power company was simulated. The results show that the false recognition rate and missing recognition rate of the proposed method were both below 0.4%, while some of the false recognition rate and missing recognition rate of the traditional power system monitoring and control method were over 0.4%. Compared with the traditional method, the identification performance of this method has been effectively improved, and the accuracy rate and recall rate of this method have reached more than 80%. The maximum response time of the proposed method was 1477.18 ms and the maximum processing time was 2452.34 ms, both of which were superior to the traditional power system monitoring and control methods. The proposed method has better system performance. Therefore, the IMM of power system based on AI technology is expected to solve traditional problems such as difficult data processing, insufficient robustness, excessive manual intervention, and insufficient adaptability, which can make substantial contributions to the development of intelligent monitoring and control of power system.
