| 摘要 |
An Adaptive Kalman Filtering scheme for statistically predicting the occurence and type of a fault on a three phase power transmission line. Additionally, estimations of the steady-state postfault phasor quantities, distance protection and fault location information is provided. Current and voltage data for each phase is processed in two separate Adaptive Kalman Filtering models simultaneously. One model assumes that the phase is unfaulted, while the other model assumes the features of a faulted phase. The condition of the phase, faulted or unfaulted, is then decided from the computed a posteriori probabilities. Upon the secure identification of the condition of the phase, faulted or unfaulted, the corresponding Adaptive Kalman Filtering model continues to obtain the best estimates of the current or voltage state variables. Thus, the Adaptive Kalman Filtering model having the correct initial assumptions adapts itself to the actual condition of the phase faulted or unfaulted. Upon convergence of the computed a posteriori probabilities indicative of a faulted phase to highly accurate values, the type of fault is classified and the appropriate current and voltage pairs are selected to compute fault location and to provide distance protection. The voltage models are two state variable Adaptive Kalman Filtering schemes. The model for the current with no fault condition is two state variable, while the model that assumes that the phase is faulted is a three state variable model. Estimation convergence reached exact values within half a cycle and consequently, in the same time fault location was determined.
|
