Identification And Research Of Data Center Inspection Robot Based On Game Theory

Since the data center computer room still uses dedicated personnel and manual inspections for operation and maintenance,the inspectors need to inspect the relevant equipment in the data center on time every day,and conduct inspection,inspection,analysis and research on the data of the equipment.Manual inspections are inefficient and costly,and are no longer suitable for current data center inspection requirements.Therefore,the research objectives are determined according to actual needs,in the specific research of this article.Use the inspection robot in the data center to perform daily inspection work,liberate manpower,reduce costs,and realize intelligent and informatized inspections.This article mainly focuses on the research of the inspection robot system in the entire data center.In order to be effective and convenient for research,separate research is carried out on all the computer rooms,each computer room,and related equipment in the computer room on each floor.Through separate research methods,the research goals of each sub-system can be better achieved.After each sub-system is implemented in order,the entire inspection system can be realized and the overall experimental goal can be completed.First,assign and plan tasks for the entire computer room of the data center.In the research,the Pareto theory is used to carry out the research of task allocation.After initial allocation,task redistribution and task set allocation,the optimal task allocation method is realized,and the optimal order of computer room inspection is obtained.Afterwards,the ant colony algorithm is used to realize obstacle avoidance and specific path planning problems for the inspection robot to reach each task point.After completing the above research,the navigation and positioning of each computer room in the data center during the inspection.In the research,firstly,the visual navigation method is used to make the inspection robot move forward according to the preset guiding trajectory to ensure that it passes through all the equipment that needs to be inspected.Then use the combined positioning method of two-dimensional code recognition and monocular distance measurement to achieve precise visual positioning and ensure that the inspection robot can park in front of each equipment that needs to be inspected.Finally,research on the realization of specific inspection tasks for each device in the data center computer room.In the specific implementation,firstly,the recognition of temperature and humidity indications is realized by using the digital character recognition method of sample matching,and then the color recognition technology is used to realize the determination of the equipment alarm phenomenon.Through all the above-mentioned researches,the operation of the entire inspection robot system can be completed.Through the design and research of the integrity of the inspection robot system in the entire large-scale data center,this article can effectively achieve the goal of the inspection robot in the large-scale data center that can complete the tasks autonomously,ensuring the executable of the inspection robot system.Reduced the time and manpower invested in the inspection process,improved the intelligence and information level of the data center,and successfully completed the research goals in accordance with the design requirements.