The Application And Research Of Neural Network In Power System Harmonics Analysis

In recent years, as a result of a wide application of capacitors in the industrial and a large amount of nonlinear loads and a variety of sensitive equipment application in the general population, the harmonic distortion level of distribution network is increasing fast. The harmonic distortion issue become increasingly serious. Harmonic has become an important power quality impact factors of the power system, this paper primarily studies the neural network applications in harmonic analysis area.Firstly, this paper describes the causes of harmonic and harmonic harm, and the current situation and development trend of harmonic treatment. Also introduces the development history of artificial neural network theory and its current application status.In order to solve the current problems existed in harmonic analysis method, proposes a kind of high precision analysis harmonics method based on neural networks for power system. By training the sampling signals using neural network method, access to high-precision power system fundamental frequency, amplitude and phase. This paper describes the analysis of neural network theory and algorithm implementation process.A neural network method based on simulated annealing is proposed for harmonic analysis algorithm, this method is aimed at the system in which the sampling frequency cannot be locked on the actual fundamental frequency. By updating the relevant parameters of angular frequency study rate, phase and amplitude harmonic and etc, the accuracy of the estimates provided by the proposed approach in the asynchronous case is relatively better than that of the estimates obtained with the conventional harmonic analysis methods. Compared to an ordinary neural network algorithm, the method could reduce the training frequency significantly, and reduce the concussion of estimate.In the process of harmonic analysis for the fundamental frequency of mutations appears in the situation, proposes a frequency tracking algorithm based on BP neural network. Every time, it reserve a part of the sampling points, and update the other part of the value of sampling points constantly. Simulation results shows that at the situation of the power system frequency change, it can be completed tracking the frequency of high-precision in a relatively short period of time, it could solve theproblems which caused of the frequency mutation.