Development of an Automated Testing System for Alternators with Real-Time Monitoring Based on IoT and Electrical Analysis

• Autores: Maravi Obispo, Kevin Gliserio, Gago Salcedo, Jeancarlos, Huamán Rojas, Jezzy James
• Formato: tesis de grado
• Fecha de Publicación: 2025
• Institución: Universidad Continental
• Repositorio: CONTINENTAL-Institucional
• Lenguaje: inglés
• OAI Identifier: oai:repositorio.continental.edu.pe:20.500.12394/18229
• Enlace del recurso: https://hdl.handle.net/20.500.12394/18229; https://doi.org/10.14445/23488379/IJEEE-V12I6P102
• Nivel de acceso: acceso abierto
• Materia: Búsquedas en Internet; Internet searches; Inteligencia artificial; Artificial intelligence; Sistema eléctrico; Electrical system; https://purl.org/pe-repo/ocde/ford#2.02.01

This research focuses on designing an automated test bench for alternators, which incorporates real - time monitoring using Internet of Things (IoT) technologies, advanced electrical analysis, and intelligent diagnostics based on a Multilayer Perceptron (MLP ) artificial intelligence model. The system allows the testing of automotive alternators under various operating conditions, measuring the RPM, voltage, and current generated and sending the data to a web platform via an ESP32 microcontroller. Multiple tes ts were performed during the experiment at various load levels and speeds, demonstrating a direct relationship between voltage and RPM. Additionally, the PZEM - 003/017 achieved a measurement margin of error of less than 1%, and the AI model's fault detectio n accuracy exceeded 90%. Likewise, a finite element analysis (FEA) of the system's structural framework was performed, validating the rigidity and safety of the structure under specified rigid loads through simulations of tension, displacement, and safety factors. The developed system provides accurate, cost - effective, and scalable diagnostic tools for alternators in industrial maintenance, technical training, and testing environments. The modular architecture, incorporation of dynamic speed control, and re al - time predictive analytics capabilities represent a significant improvement over traditional methods.