Bad data validation on the basis of a posteriori analysis

Abstract

Reliability of metering information is very important. There are different mathematical methods of measurement data verification. Large quantity of the bad data identification techniques are proposed to identify low precision measurements. These methods are developed as a part of the state estimation problem to calculate the steady state on the basis of telemetry data. A new approach to the calculation of state variables vector is proposed. A posteriori algorithm of bad data validation is described in details. The example of bad data detection for test scheme is considered.

The Universal Mathematical Model of Excitation System of Synchronous Machines

“…The necessity to improve the reliability of mathematical modeling of modes and processes in EPS, dictated by constantly a large number of severe accidents, responsible for approximately 50% of which are directly or indirectly, is incomplete and insufficiently reliable information about these processes used during the design, operation and development of energy systems [6,7]. In particular, the numerous researches and operating experience proved direct dependence of reliability determination of the vibrational modes stability of the EPS from the accuracy of mathematical models.…”

Defining the parameters of a power transmission line equivalent circuit on the basis of phasor measurements

“…We also have to keep in mind that limiting the error by instrument transformer class 0.5 is quite optimistic since the actual errors may excess the rated values [17,18]; in case the error is not limited in that way the accuracy of defining the equivalent circuit parameters reaches 10 %, which indicates the result may be treated as the reference parameters refinement.…”

Improving of Energy Measurements Reliability Using Weighted and Normalized Residual Analysis

“…The presence of nodal and linear measurements is indicated by mark W. Evaluation of all measurements reliability using the balance technique is impossible as there are no linear measurements in all branches near the nodes 2 and 4. Expressions of actual and permissible imbalances can be formed only for nodes with full coverage of measurements (nodes 3 and 5) [8]. Nevertheless, the network has redundant measurement structure provided by the presence of metering complexes on second end of branch.…”

“…The total number of measurements is 11. Five redundant measurements allow to construct five control equations [8]. Further the reliability of energy flows measurements is estimated.…”