No abstract

[Background/Reason] The operation ticket system is an important measure to ensure the safety of substation operation. At present, the issuance and review of substation operation tickets rely on manual operation knowledge and experience, and there are problems such as low work efficiency, lack of standardization, and human error. The risk of grid operation is reduced. [Method/Process] This paper proposes a substation operation ticket system based on natural language analysis and intelligent reasoning. It uses typical/standard operation tickets as the basic corpus and builds a natural language analysis model based on manual labeling to achieve compliance with the operation tickets. Sexual inspection, deduction and comparison of equipment status. Finally, combining the primary wiring diagram of the substation and the operation ticket to construct the equipment relationship knowledge map and the equipment operation knowledge map, use the Neo4j graph database for storage, and realize the intelligent drafting of operation tickets based on the graph search method. [Results/Conclusions] The research results have been implemented and applied in three substations in a certain area, verifying that the system has a high degree of intelligence and versatility, and can effectively solve the existing problems of standard invoicing and audit errors, and improve operation The work efficiency of personnel and the safety control level of power grid substation operation.

In order to improve the autonomous digging ability of the inspection robot, an auto disturbance rejection control method based on monocular vision is designed for the indoor inspection robot in the intelligent substation. Through tracking and detecting the end position and posture changes of the robot in the x-z plane during the operation of the robot in the intelligent substation, and mining the abnormal data, and combining with the PID control principle, the inspection robot single view automatic disturbance rejection control method is designed The visual interference information is cleaned and denoised to realize the auto disturbance rejection and autonomy of the indoor inspection robot monocular vision pan-tilt. Due to the adjustment of the operating parameters of the pan-tilt active disturbance rejection controller based on monocular vision of the indoor inspection robot, it is difficult to set the parameters due to the interaction between the parameters. Furthermore, the parameters of the ADRC are adjusted and optimized to meet the research requirements of the effective control of the pan-tilt active disturbance rejection of the indoor inspection robot in the substation. Finally, the experimental results show that compared with the traditional methods, the control accuracy and strong robustness of the indoor inspection robot in the substation are significantly improved, which fully meets the research requirements.

No abstract

No abstract