Scholarlink

<p>Diesel generators (DGs) serve as a leading backup power solution for providing an uninterrupted electricity supply to residential and commercial properties. Hybrid machines combine electric generators with diesel engines, produce the needed electrical energy. Optimal operation of DGs depends on continuous evaluation of essential parameters through regular assessments. Advanced Condition Monitoring Systems (CMS) have replaced traditional Remote Monitoring (RM) systems through the evolution of the Internet of Things (IoT). A new framework that combines RS-485 IoT sensors with machine learning (ML) powered parameter selection for performance improvements in real-time monitoring and predictive maintenance of DG units is proposed. A 50kVA Cummins DG receives operational data often through its 4G capable IoT node that dwells on the motherboard to retrieve information about engine speed, power factor, fuel levels, battery health, voltage, current, and coolant temperature. Data transmission takes place through the RS-485 networking protocol for trustworthy and protected multi-hop digital communication. The system implements an adaptive ML-based ranking system, which organizes performance metrics according to their influence on system operational efficiency. Through its adaptable evaluation approach, the proposed system allows decision-makers to detect system weaknesses by reweighting different performance metrics for better system outcomes. This technique establishes a connection between real-time monitoring of conditions alongside predictive maintenance by providing a database, scalable insights to expand the DG system’s operating lifetime while decreasing operational downtime and expenses.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimizing diesel generators in industrial settings with IoT and machine learning

  • 期刊论文

  • 发表日期：

    2025年8月9日

查看全文
  • Anita Mohanty,
  • Ambarish G. Mohapatra,
  • Narayan Nayak,
  • S. Bairavel,
  • Chaithra M.H.,
  • M. Lakshmi,
  • Anitha Govindaram,
  • Jose Anand A.

摘要

Diesel generators (DGs) serve as a leading backup power solution for providing an uninterrupted electricity supply to residential and commercial properties. Hybrid machines combine electric generators with diesel engines, produce the needed electrical energy. Optimal operation of DGs depends on continuous evaluation of essential parameters through regular assessments. Advanced Condition Monitoring Systems (CMS) have replaced traditional Remote Monitoring (RM) systems through the evolution of the Internet of Things (IoT). A new framework that combines RS-485 IoT sensors with machine learning (ML) powered parameter selection for performance improvements in real-time monitoring and predictive maintenance of DG units is proposed. A 50kVA Cummins DG receives operational data often through its 4G capable IoT node that dwells on the motherboard to retrieve information about engine speed, power factor, fuel levels, battery health, voltage, current, and coolant temperature. Data transmission takes place through the RS-485 networking protocol for trustworthy and protected multi-hop digital communication. The system implements an adaptive ML-based ranking system, which organizes performance metrics according to their influence on system operational efficiency. Through its adaptable evaluation approach, the proposed system allows decision-makers to detect system weaknesses by reweighting different performance metrics for better system outcomes. This technique establishes a connection between real-time monitoring of conditions alongside predictive maintenance by providing a database, scalable insights to expand the DG system’s operating lifetime while decreasing operational downtime and expenses.