Abstract
Identifying voltage variations has an important role to mitigate degradation of power quality disturbances due to their negative impact on the equipment. Therefore, fast and accurate detection and identification are required in order to evaluate and activate the protection. In this paper, Root Mean Square (RMS) algorithm combined with artificial neural network is presented in order to detect and identify voltage variation such as sag, swell, undervoltage and overvoltage. Artificial neural network to identification of voltage variation using feed forward neural network because of its simplicity and accurate. In order to model voltage variations required, a generator is required to emulate of voltage variation waveform. It consists of transformer single input multi output, SSR Switch and microcontroller. Moreover, Identification is implemented with microcontroller STM32F47. Varying neuron under hidden layer is tested to validate the proposed algorithm. The results show that the proposed algorithm is effective to identify each variation with the average percent of error 2.3749% for 8 neurons in the first layer and 0.0658% for 10 neurons in the first layer.
Original language | English |
---|---|
Pages (from-to) | 224-230 |
Number of pages | 7 |
Journal | International Review of Automatic Control |
Volume | 13 |
Issue number | 5 |
DOIs | |
Publication status | Published - Sept 2020 |
Keywords
- Microcontroller
- Neural Network
- Overvoltage
- Sag
- Swell
- Undervoltage