Data Mining-Based Performance Evaluation of DC-DC Power Converter Test Bench Using IoT Monitoring

Lovera-Garcia, Jesus Adan, De la Garza-Macedo, Jose Angel, Mastache-Mastache, Jorge Edmundo, Balderas-Perez Karina

Abstract: This study presents the implementation of data mining techniques in a DC-DC power converter test bench integrated with IoT systems. By using decision trees, the efficiency of the converters under various duty cycles is evaluated in real time. The system utilizes ESP32 and Arduino MEGA boards alongside multiple voltage and current sensors to collect data, which is stored in the THINGSPEAK cloud. RAPIDMINER software is used for mining and visualization. The findings highlight the conditions under which converters perform optimally, making this a promising approach for future control system designs.

Keywords: DC-DC converters, IoT monitoring, data mining, decision trees, efficiency analysis