Filtered Leakage Current Measurement for Various Loads

• Main Authors: Erwin Sutanto, Silvi Nurwahyuni, Riky Tri Yunardi, Guillermo Escrivá-Escrivá
• Format: Article
• Language: English
• Published: Universitas Indonesia 2021-04-01
• Series: International Journal of Technology
• Subjects: circuit breaker; electrical loads; filtered signal; leakage current; machine learning
• Online Access: https://ijtech.eng.ui.ac.id/article/view/4302

Thepurpose of this study was to determine the link between induction voltages from various electrical loads. We used a residual current device (RCD) circuit that operates with a capacitor as a DC voltage reading tool. The circuit reads the value of the leakage current generated by the sensing coil from the RCD. It also uses the Blynk framework as an online monitoring system and a WeMos D1-R2 microcontroller to connect to the server using Wi-Fi. Using this system, the dataset was collected in a Python server and utilized with a machine learning technique to draw a correlation between the load power and reading voltage. This will help improve the mistakes of a common RCD cut-off point, which is usually defined only at a specific induced voltage. For the different types, an LED lamp and typical electric fan were used as loads in the experiment. Meanwhile, for a similar type of load, three different LED lamps were characterized using machine learning to show the correlation. From the comparison, a threshold voltage of around 1V and three different gradients with increases of more than 10% are found for LED lamps with loads of 3W, 5W, and 9W.The results show that the relationship depends on the type of its power supply.