Selection method of transition resistance during automatic voltage regulation of distribution transformer

Abstract

The invention discloses a selection method of transition resistance during automatic voltage regulation of a distribution transformer. The selection method comprises the following steps: starting fromvoltage waveforms and the circulation waveform characteristics in the transition process, filtering the output voltage waveforms and then taking the filtered output voltage waveforms as input of a particle swarm PSO algorithm through BP neural network training and approaching, calculating an optimal threshold value suitable for a BP neural network, then comparing the resistance value of the transition resistance calculated by the neural network with the optimal threshold value by using a threshold value comparison function, and if a BP threshold value selection function is met, outputting a final transition resistance value after repeated calculation. According to the selection method, simulating calculation is carried out on the transition resistance of the distribution transformer system by using the particle swarm PSO algorithm and a BP neural network algorithm, the resistance value of the transition resistance can be solved simply and effectively, and the method is more reliable compared with a manner that the resistance value of the transition resistance is obtained through experience, so that the value of the transition resistance is more scientific, the output voltage of the distribution transformer is stabler, the wave disturbance on a power grid is smaller, and little harmonic pollution is also caused.

Description

technical field [0001] The invention belongs to the technical field of power transformers, relates to automatic voltage regulation of distribution transformers, and is a method for selecting transition resistance in automatic voltage adjustment of distribution transformers, and proposes to use Particle Swarm Optimizer (PSO) to calculate BP propagation) neural network threshold method, which improves the ability of the neural network to approximate the function, and realizes the fast and reliable selection of the best transition resistance value. Background technique [0002] With the rapid development of the power system, the quality of power consumption has attracted more and more attention from users, and users have been pursuing a better power consumption experience. Voltage quality is an important indicator of power quality. Excessive voltage deviation will affect the production of enterprises and the improvement of people's living standards, affect the life and efficien...

AI Technical Summary

Problems solved by technology

At present, the method for selecting transition resistance is to select the resistance value by practical experience, and there is not much selection basis with applicability. The present invention proposes a transition resistance selection method at this point, and selects through the neural network optimized by PSO. According to the actual realization The transition resistance selected by data combined with intelligent algorithm makes the transition process more stable, the circulation impact of the transition process is smaller, and the voltage interference is smaller

Images