Research On Artificial Intelligence In Power System Dynamaic Equivalence

Dynamic equivalence is a an excellent way of simplifying electricalsystems, reducing the number of equations to solve while retaining themost important characteristics of the full system, which has been drawingincreasing attention throughtout the world with the steady expansion ofthe system, important for real-time online dynamic analysis. The mainworks in this dissertation are as following：1. Study on basic theory of power system dynamic equivalence,including modeling of elementary electrical components and thealgorithms to achieve it.2. A Novel Load Modeling Approach has been proposed, which isbased on asymmetric data recording and solutions to Park equations. It’sapplicable to not only the normal networks, but also the systems with tielines.3. A niche immune algorithm has been proposed and applied to apractical infinite-bus system, which has several advantages, such as highdegree of accuracy, high speed, satisfied astringency and good robustness.In the end, the proposed algorithm and the presented method applyto Tianjin network to test their feasibility, and the static characteristic and dynamic characteristic of the obtained models are checked as well.Simulation results show that the methods proposed in this dissertation isfeasible and applicable, qualified to be used in actual engineeringprojects.