Research And Application Of Automatic Diagnosis And Identification Algorithms For Meter Faults In New Meter Reading System

In the power supply management of power enterprises,traditional point-to-point manual meter reading methods usually cause errors,falsifications,misses and calculation errors,which seriously reduces the level of power service,and no longer meet the needs of modern power enterprise management.The remote centralized meter reading system can not only satisfy the demand of power users for acquiring their household appliances in an all-round way,and achieve the goal of "one household,one meter",but also avoid a large number of errors caused by human reading meters and operation process through remote recruitment of data.But when it comes to the phenomena of changing,disassembling and changing the power point of the meter,the software platform and hardware system of the existing centralized reading system can not identify the above phenomena on their own,and the above problems can only be found by the on-site investigation of the staff.At the same time,when the user's meter files are changed,the information recorded by the concentrator is inconsistent with the user's files and appears offline,which not only reduces the success rate of collecting and copying,but also significantly increases the workload of later maintenance.In addition,the data can not be copied because of the damage of field equipment or the interruption of communication channels.Aiming at the above problems,this paper mainly studies the automatic fault diagnosis algorithm and change recognition algorithm of meter,and applies them to the development of a new meter reading system for experimental testing.The specific research work is as follows:Firstly,this paper analyses the functional requirements and basic data requirements of the new centralized transcription system,and designs the overall framework of the system,including the system logical framework,physical framework,software framework,communication mode and so on.Secondly,this paper analyses the common fault situations and common fault diagnosis methods of ammeters,and designs the fault diagnosis methods of ammeters adopted by the new centralized reading system in this scheme,including the fault diagnosis methods for ammeters themselves and the diagnosis methods for the abnormalities of ammeters caused by electricity theft.Then,based on the clustering algorithm,this paper proposes a method of identifying the change of ammeter,and designs a concentrator with the function of automatic table searching.By searching the logical address of ammeter,the problem of identifying the change of ammeter is solved.Finally,this paper develops the software system according to the system requirements and two functional algorithms,and verifies the practicability of the system through the actual pilot application.