A New Fault Location Identification Method for Transmission Line Using Machine Learning Algorithm
| Autor: | Jaedeok Park, Gihun Park, Taesik Park, Jun-Soo Che |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Power transmission
Computer science business.industry GRASP Hardware_PERFORMANCEANDRELIABILITY Fault (power engineering) Machine learning computer.software_genre Identification (information) Transmission line Artificial intelligence Line (text file) MATLAB business computer Algorithm computer.programming_language Voltage |
| Zdroj: | 2019 3rd International Conference on Smart Grid and Smart Cities (ICSGSC). |
| DOI: | 10.1109/icsgsc.2019.00-14 |
| Popis: | Conventionally, the fault types and locations in a power grid are detected based on the voltage and current wave forms. Fault types and locations varies slightly depending on the location of the accident, which is not easy to grasp with the human eye. Therefore, many sensors are needed to diagnose faults, and it is very difficult for an administrator to determine the type and location of faults using voltage and current waveforms. In this paper, a new fault location identification method for power transmission system is proposed. By using machine learning, a model is created to learn the transients of voltage and current data and outputs the location in case of a new accident. The method presents a new classification system of faults based on faults data from simulations and artificial intelligent algorithms. Also, if each company has bus data, classification system can be added without additional devices. The diagnostic performance of the proposed method is verified by MATLAB simulations, It has taught line to ground and line to line fault data in various models, and in this paper, three models with high accuracy are described. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|