The Study Of Dection And Classification Of The Power Quality Disturbance Based On Atomic Rapid Decomposition

At present, the extensive application of intelligent precision meter puts forward more strict requirements on the quality of power supply. At the same time, due to the large disturbance load connected to the grid or other disturbance source, the power quality problems have apperaed to be increasingly prominent. Therefore, research on and analysis of disturbance signal of power quality is very important. This paper analyses various disturbance signal of power quality by using atomic decomposition technology.The shortcoming of the traditional atomic decomposition technique is large amount of calculation. In this paper, in order to overcome the shortcomings, discrete parameters of atoms are transformed to consecutive parameters of atoms, which can reduce atomic quantity of the reconstructed signal and make the result more accurate; if the range of frequency of the signal spans widely, such as harmonics and damping oscillation signals,the fast Fourier transform is adoped to solve the frequency of the optimal atoms in advance, which can reduce the scale of the dictionary; the method applies the particle swarm algorithm and genetic algorithm to optimize the process of matching pursuit. The examples indicate that the performance of atomic decompoaition based on the particle swarm algorithm is better than the performance of atomic decompoaition based on the Genetic algorithm.Atomic decompoaition based on the particle swarm algorithm and the continuous coherent dictionary is used to study the power quality. In this paper, we will analyse six kinds of power quality disturbance, include voltage sag, swell, impulse, damping oscillation, harmonics, flicker. The simulation examples indicate the method can rapidly and accurately extract the characteristics of power quality disturbance, and have better antinoise abilities.Beacuse continuous coherent dictionary has pertinence, so the article uses the continuous coherent dictionary and matching pursuit based on the particle swarm algorithm to classify power quality disturbance. Through the simulation examples of multiple disturbance signal indicate that the classified method of the power quality disturbance can be done well, and may get all the parameters of the power quality disturbance.