| Popis: |
The measurement quality of each node in the active distribution network is quite different and the quantitative interval of the pseudo measurement can be given. The meter accuracy for the real-time measurement can also be expressed by the error interval, so the interval state estimation is an important available choice. Bad data identification is a necessary step to improve the accuracy of state estimation. Because bad data identification for interval state estimation is a mixed integer nonlinear programming problem for which is sensitive to initial value, three-stage interval state estimation is used. In the first stage, linearized model is used to solve the bad data identification problem to provide a better initial value. In the second stage, more accurate bad data identification nonlinear model is used. In the last stage, interval state estimation is performed to estimate intervals of all the variables. In the process of three-phase interval state estimation, the detailed AC-DC hybrid model and distributed generators model are considered. |
