Dynamic State Estimation Under Communication Failure Using Kriging Based Bus Load Forecasting

Abstract

Dynamic state estimation (DSE) in power system combines forecasting technique with measurement data to accurately estimate system state. The current DSE techniques cannot handle the situation where communication failure occurs and measurement data are lost. In this paper, a new approach is proposed to address this problem. The proposed approach combines the extended Kalman filter (EKF) with load forecasting technique that predicts missing measurement data. A time-forward kriging model is used to forecast the missing load data from the available measurement data. The forecast load is then converted to forecast system state through power flow analysis. The EKF is used to combine the measurement data with the forecast state to obtain a more accurate filtered state. The proposed approach is tested on IEEE 14-bus system and IEEE 118-bus system using realistic load pattern from NYISO and PJM with various scenarios of measurement error and communication failure. The test results from the proposed approach are compared with traditional weighted least square (WLS) state estimation and DSE with multi-step ahead autoregressive integrated moving average (ARIMA) load forecasting. From the case studies, we find that the proposed approach provides more accurate and faster state estimation under most scenarios.