Introduction

Supervisory control and data acquisition (SCADA) systems are widely used to monitor and control large-scale transmission power grids. Monitoring traditionally involves the measurement of voltage magnitudes and power flows; these data are collected by meters located in substations. In order to deliver the measured data from the substations to the control centre, the measurement data measured by meters in the same substation are multiplexed by a remote terminal unit (RTU) [1, 2]. Because electric power transmission systems extend over large geographical areas, typically entire countries, wide-area networks (WANs) are used to deliver the multiplexed measurement data from the substations to the control centre.

For large-scale transmission grids it is often not feasible to measure all power flows and voltages of interest. Furthermore, the measurements are often noisy. Therefore the measurement data are usually fed into a model-based state estimator (SE) at the control centre, which is used to estimate the complete physical state (complex bus voltages) of the power grid. The SE is used to identify faulty equipment and corrupted measurement data through the so-called bad-data detection (BDD) system. Apart from BDD, the state estimate is used by the human operators and by the energy-management systems (EMS) found in modern SCADA systems, such as optimal power flow analysis, and contingency analysis (CA), see for example [1]. Future power grids will be even more dependent on accurate state estimators to fulfil their task of optimally and dynamically routing power flows, because clean renewable power generation tends to be less predictable than nonrenewable power generation.