Research On Detection And Identification Of Bad Data In Power System State Estimation

With the development of the electric power industry and the improvement of the people's living standard.the power users have put forward higher requirements for the reliability of power supply and the quality of power.In order to improve the efficiency and reliability of the power grid,a complete and reliable real-time database is needed for modern dispatching system,so as to facilitate the online analysis and decision control of advanced application software.Power system state estimation is the core of power system on-line monitoring,analysis and control function,playing an important role in the intelligent analysis and decision-making of power grid dispatching.The estimation results directly affect the correctness and effectiveness of the operation analysis and decision system,how to improve the accuracy of state estimation is an important content of the research.However,due to unreasonable layout,poor channel transmission and poor management operation,the measurement system must detect and identify the bad data of the measured data.Detection and identification of bad data is one of the important function of state estimation in power system,its purpose is to eliminate a few bad data by the measurement data and improve the reliability of state estimation,which is of great significance to the safe operation of power system.In this paper.the following research is carried out on the detection and identification of bad data.(1)For the first time,external studentized residuals are used as criterion to detect and identify bad data in power system state estimation,improveing the shortcomings of standardized residuals in traditional bad data detection and identification.simulation is carried out to verify its performance.In addition,two new methods for measure mutation detection method are proposed.(2)Aiming at the limitations of the identification of bad data using a single detection criterion,FCM algorithm based on subtractive clustering is proposed to comprehensively analysis external studentized residuals,quantitative measurement of mutation detection and measurement error for bad data detection and identification,and simulation is carried out to verify its performance.(3)Aiming at the limitations of the FCM algorithm based on subtractive clustering in complex situation,the exponential function weighted least squares method based on external studentized residuals is proposed.The form of the original algorithm is modified using the principle of equivalent weight.The external studentized residuals are used instead of the original residual.Simulation is carried out according to the different situation of bad data to verify its performance.The performance is compared with the traditional robust estimation method and basic weighted least squares method.