The Research And Design Of Substation Intelligent Patrolling Platform

Since the concept of smart grid was proposed,intensive research has been conducted on the related area.As the central part of the smart grid,any malfunction of the substation may lead to harmful impact on the people's life and the society.Thus,it's of vital importance to maintain the substation's secure operation,while a regular inspection is one of the key methods.However,traditional inspection on the substation is performed by humans,which suffers from problems such as abnormal inspection procedures,instability to human factors and insufficient utilization.To address these problems,an intelligent patrolling platform is introduced in this thesis,which effectively formalizes the patrolling procedure,gives fault diagnosis to electronic equipment,and helps the regular inspection on the substation.The main contributions of this thesis are listed as following.1)Investigated recent intelligent patrolling techniques,clarified the motivation and target of this thesis,pointed out that most inspection systems focus on the data collection,while fault diagnosis is performed afterwards at the main station with insufficient communication with the patrolling process.2)Proposed the overall system design and operation procedure of the intelligent patrolling platform.Furthermore,through a thorough investigation on the communication standards between the platform and the mainstream equipment,the AMQP message queue protocol is adopted in this thesis.3)Categorized the common faults of the equipment at the substation,the partial discharge test and dissolved gas test in oil were studied emphatically.Firstly,the fault diagnosis algorithm of partial discharge test is designed on the basis of referring to the inspection protocol of power industry.Secondly,based on the traditional DGA method,the improved GABP neural network algorithm is proposed and implemented in Matlab.4)Utilized “Webservice” to transfer information between the platform and the backend management system,designed the functional modules and the corresponding databases for the management system.The management system is then used to verify the platform's effectiveness.In summary,we proposed an intelligent patrolling platform,which is capable of analyzing the inspection data,giving fault diagnoses,effectively communicating with the backend system and uploading the diagnosis results.The proposed patrolling platform improves the utilization of the inspection data,formalizes the inspection procedure and gives actual reference value to the secure operation of the electrical system.