The Design And Implementation Of Bad Data Identification System Based On Zigong Distribution Network

With the development of information technology,smart grid has became the focus of attention of power grid companies all over the world.Smart grid integrates power generation,transmission and distribution.Through the advantages of automation and informationization,it can realize automatic monitoring,diagnosis and repair,and the work efficiency has been greatly improved.At present,smart grid has to be widely used in China.However,smart grid requires high accuracy of distribution network data.It needs to identify the bad data in order to solve the problem.It is helpful to improve the stability of service and the quality of power supply.This is also the reason why I chose to study the identification system of bad distribution network data.The identification of bad data plays an indispensable role in power system state estimation.Because the measurement data of ammeter may cause data errors or lost when transmitted to the control center through the network,some telemetry results may be far away from the actual value and other objective factors,so the data become inaccurate and unreliable.In practice,the data are transmitted through instruments and other equipment.The collected information can not be used to judge the system state directly.The measured data may contain bad data.The purpose of bad data detection is to exclude a few bad data and improve the reliability of state estimation.The identification of bad data is mostly based on residual method,which has a significant effect on the identification of bad data,but it also has some drawbacks such as residual pollution.In this paper,residual sensitivity matrix is used to distinguish the possibility of interaction between measurements,to avoid misjudgments in multi-interaction bad data scenarios,and to improve the accuracy of identification of multi-interaction bad data in state estimation.Then,the genetic algorithm is used to design the adaptive function according to the optimization problem to obtain the possible solution combination of the optimization problem.Distribution network bad data identification system is developed based on J2 EE,and Oracle is used to store data.At the same time,JXL plug-in is used to import and export data.Distribution network bad data identification system is composed of distribution data management,bad data identification management,identification result management and security management modules.The system has the following characteristics: firstly,the residual sensitivitymatrix and genetic algorithm are used to realize the bad data identification algorithm;secondly,the web service technology is used to realize the bad data identification results sharing;and the parallel database and backup scheme are used to realize the data protection scheme.At present,the bad data identification system of distribution network has been constructed and tested from two aspects of function and performance.The system meets the application requirements and is deployed to the power grid company to deal with the bad data identification business.The introduction of the bad data identification system of distribution network effectively improved the power grid.Bad data identification level and work efficiency.