Prediction and Analysis of Power Flow by Machine Learning for Power System
| Autor: | Chen, Yen-Yu, 陳彥宇 |
|---|---|
| Rok vydání: | 2019 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 107 The main purpose of this paper is to use the neural network-like method of machine learning to determine the maximum injection capacity at the connection point of the power system. The paper also studies of how to increase the accuracy by adjusting the varying capacity of generators that connects to the maximum injection capacity at the connection point of the power system. In the wind farm of power connects to the power system, it should check the rate of current-carrying capacity to meet the spec, following to utility rule to consider the maximum injection capacity. The next step in analysis is to realize generators capacity for increasing power or reducing power in generators about the influence and accuracy. During the system operation, the method of calculating the maximum injection capacity about Busbar loops of the rate of current-carrying capacity is based on the neural network topology. For the accuracy, it relies on the variable capacity between the maximum injection capacity and the variation of increasing and reducing capacity in generators. For this paper study, it increases the number of neuron samples by injection different capacities and different connection positions to analyze the transmission lines that are easily congested under system operation. The purpose is to increase the accuracy up to 87% for the connection point of maximum injection capacity by neural network modeling. After that, using the data about increasing or reducing generator capacity as neuron samples can raise the accuracy up to 83% in maximum injection capacity calculation. The model for this paper study can apply to new wind machines to connect to grid for power flow analysis by using machine learning. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|