State Estimation Research Of Power System Base On Robust Theory

In Power system state estimation, identification of bad data in all situation correctly, to preclude the bad influence of bad data, so the aim that the nicety of the state estimation is guaranteed is hard to get. So to research a suitable way, when the bad data can't be identify entirely, the estimation is still veracity. That may be a good plan to improve nicety of the estimation result. The robustness estimation theory means that, when the gross error is inevitable, choosing a proper estimation method and excluding the influence of the gross error, in order to get the best value of estimation in normal mode. This theory is built upon the factual distributing mode of the measure data.Based on the principle of equivalent weights, the least robustness square method is combined with robust estimation theory and the Weighted Least Square (WLS) form. This method not only has the high efficiency as WLS, but also has favorable outlier resistance. In the iteration, the residual is used to revise weight value. The choice of the Weight function is the key of the robustness of WLS. In this paper, sorting the measure data by statistical analysis, a new choice of the weight function which based on the least robustness square method named"triple segment"is presented. By setting of retaining weight region, the merit of the high efficiency of the least square method is retained. To make the best of the effectual measure data is due to setting of the reduced weight region. And the repellent region is used to resist outlier. The precept was proven feasibility in IEEE14 Node and IEEE30 Node.The application of the robustness estimation theory is a new question for discussion. On account of the huge account work of this theory, it can't satisfy the power system in requirement of the real time performance. Also the problems of dealing with leverage point and robust estimation of associated measurement are worth research.