Font Size: a A A

Study Of Power System State Estimation

Posted on:2013-07-26Degree:MasterType:Thesis
Country:ChinaCandidate:N L CaiFull Text:PDF
GTID:2252330392469996Subject:Electrical engineering
Abstract/Summary:PDF Full Text Request
Power system state estimation (SE) is one of the most basic EMS soft wares, whose performance has a direct effect on the reliability and credibility of all the other application soft wares based on it. Nowadays there are a number of issues on the study of power system state estimation. First of all, distributed network has to use three-phase SE because of its asymmetrical parameter and asymmetrical load, and it costs long calculate time. Secondly, the bad data limits the improvement of the SE accuracy. The traditional robust estimation algorithms face the problem of convergence and robustness for the sake of the objective function. How to improve the robustness of the SE has been an issue. What’s more, the transformer tap observation studies, mainly using algebraic methods of decomposing information matrix or the correlation matrix costs long computation time, and is difficult to apply to real-time online analysis.Facing to the problems above, the thesis does research on it, and gets achievements as follows:(1) In order to solve the problems of the long time of the three-phase SE, this thesis does research on single-phase to three-phases hybrid SE based on the equivalent current measurement transform. This thesis set up a П-type equivalent mode to deal with the interface between single phase side and three-phase, raising a hybrid method using this mode and another algorithm based on a single point of alternating iteration. Multiple tests show that the hybrid SE algorithm dealing with a single-phase to three-phase SE algorithm greatly improved computing speed, while still guaranteed the calculation accuracy.(2) This thesis proposes a new robust power system SE algorithm based on the adaptive kernel density estimation to improve the ability of identifying bad data. Based on adaptive kernel density estimation theory, the thesis first provides the mathematical model based on the measurement of adaptive kernel density power system SE, and then derives the approximate optimal bandwidth expression compatible with the state variable and the bandwidth constraints, and thus gives the bandwidth selection method. The mathematical model gives the solution of the objective function. Tests has proved the robustness、high accuracy and fast computing speed. (3) To get a fast and accurate method of online tap observation, this thesis paper presents a practical observability criteria for the transformer tap observability. According to the obserbility of the island and the measurement on the transformer, the obserbility of the transformer tap can be derived quickly. Tests have proved that observability test criterion can achieve fast, reliable transformer tap observable inspection; each algorithm has a good engineering application prospects.
Keywords/Search Tags:power system, state estimation, single-phase to three-phases, baddata detection, robust estimation, adaptive kernel density estimation, tap observation
PDF Full Text Request
Related items