Application Research Of Fault Diagnosis Method Based On Apriori Algorithms In Dezhou Power Grid

Dezhou Power Grid is currently responsible for 12 counties and cities including Shengli(Linpan)Oilfield,but its transmission line environment is complex,and severe weather such as strong convection occurs.Ensuring the stability of transmission lines,responding quickly and locating faults in transmission lines is an important issue facing the current Dezhou power grid.In this context,this paper studies the existing information management system of Dezhou Power Grid,combines big data processing technology,Web technology and other functions to reconstruct and upgrade the power information management system function,and realizes a transmission line based on big data analysis.The fault diagnosis system can effectively improve the fault diagnosis and analysis efficiency of the current power system transmission line.Firstly,based on the theory of over-limit driving and event-driven detection of power grid fault analysis,the event-driven module for evaluating the operation status,energy supply,secondary system faults,transmission system equipment,management and meteorological environment faults in each self-control area of power grid is constructed by using the over-limit index design.Voltage safety,overload safety and system transient instability are set up on these event-driven modules,and real-time monitoring and early warning of transmission network operation status and fault are realized through the above indicators.Secondly,on the basis of index data,data feature extraction is carried out,and a data mining and analysis scheme based on Apriori algorithm is constructed to realize automatic fault diagnosis and analysis.In addition,in order to ensure the efficiency of Apriori algorithm in data mining of transmission network indicators to meet the requirements of real-time analysis,the frequent item pruning method is adopted to reduce the final reduction of combinations caused by this item by deleting unnecessary items in generating frequent item sets from two aspects of generating huge candiTime sets and scanning multiple item data.The traditional Apriori algorithm is optimized to reduce the number of candiTime items by about 52%.From the analysis of the processing efficiency of different algorithms for the same number of transactions,the performance of the optimized algorithm is improved by 15%.Through the optimized M-Apriori algorithm,the correlation between event index data and faults can be analyzed in real time,and the automatic diagnosis and analysis of faults can be realized.Finally,the design and implementation of the fault diagnosis and analysis platform for transmission line based on data mining is completed.The design of the database conceptual model and the design of the entity table for the storage of fault information data such as indicators and analysis results are given,and t_trouble_vector and t trouble vector are constructed.T trouble record two tables to store grid power system fault vector and fault data.Based on the business data of the existing information management system of Dezhou Power Grid,the fault data warehouse is formed through data extraction and preprocessing,and based on the data warehouse,the transmission line index management,fault indicator data management,fault information mining analysis and transmission line are completed.The design and implementation of the four core functions of fault prediction.Taking the typical fault analysis of the Dezhou power system as an example,an example analysis of the common voltage transformer grounding and overhead Integererval tripping faults is carried out,which can accurately lock the fault location and area information when the fault occurs,and the fault analysis response time.Can be controlled within 3s,the accuracy of multiple test failure analysis can reach 95%.