Data rectification algorithm and its application in the supervisory information system in power plant were studied in this paper. Fundamental concepts of measurement error and data rectification were introduced. Based on the principle of multivariate statistical analysis, the theoretical model of data rectification was developed. Data reconciliation algorithm based on Kalman digital filtering was analyzed in detail. Combined with the characteristic of supervisory information system in power plant, by grading the measured parameters, a new measurement error detection and amendment model was established. Based on this model and the techniques of redundancy check and digital filtering, a new data rectification algorithm has been proposed, which unifies the data reconciliation and the gross error detection and identification. Results show that the model has clear physical meaning and can meet the needs of performance analysis.
Journal of Southeast University(Natural Science Edition)