Hybrid State Estimator with SCADA and Phasor Measurements

This paper investigates the problem of combining both the Supervisory Control And Data Acquisition (SCADA) measurements and phasor measurements for state estimation. The SCADA measurements are slow dynamic in nature than phasor measurements. The latency between those measurements is compensated using Auto Regressive and Neural Networks. Weighted Least Square estimator technique is used for estimating states of the power system. A Traditional State Estimator (TSE) solution is obtained using SCADA measurements which already exist in energy management system. The proposed work is performed on the TSE estimated states for future state prediction. Future TSE state estimates are predicted at the instant of Phasor Measurement Unit (PMU) measurements received at the energy management systems. Auto Regressive, Feed Forward Neural Network and Nonlinear Auto Regression exogenous models are used for state prediction and the effectiveness of these models are evaluated. Then, the state vectors of Hybrid State Estimation (HSE) are calculated using predicted state vector and selected PMU measurements. The Neural Network based SE provides best state estimates over AR based SE. The working of the proposed process is validated through the simulations carried out on standard IEEE 14 bus, 30 bus, 57 bus and 118 bus systems.