Scholarlink

<p>The transition from hardware-based Continuous Emissions Monitoring Systems (CEMS) to software-driven Predictive Emissions Monitoring Systems (PEMS) is driven by the need for cost-effective and efficient emissions monitoring. PEMS leverages equation, statistical learning and machine learning to predict emissions in real-time, reducing capital costs by 50% and operational costs by 90% while minimizing maintenance and safety risks and ensuring continuous data availability compared to traditional hardware-based solutions. This review examines current emissions monitoring technologies, regulatory frameworks for emissions monitoring, and PEMS model development, with a focus on machine learning techniques such as LSTM, TCN and stacked model architecture, which are recommended for enhancing PEMS accuracy and predictive capabilities. Machine learning-based PEMS has significant advantages in handling complex, non-linear problems where simpler models may struggle. By synthesizing recent advancements, this study underscores AI-driven emissions management as a crucial step in digital transformation, optimizing efficiency, ensuring compliance, and promoting sustainability.</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

Researching the landscape of predictive emissions monitoring system: a review of literature and technology trends

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年6月4日

查看全文
  • Fanxing Wang,
  • Yu Pang,
  • Ling Bai,
  • Marc Godin

摘要

The transition from hardware-based Continuous Emissions Monitoring Systems (CEMS) to software-driven Predictive Emissions Monitoring Systems (PEMS) is driven by the need for cost-effective and efficient emissions monitoring. PEMS leverages equation, statistical learning and machine learning to predict emissions in real-time, reducing capital costs by 50% and operational costs by 90% while minimizing maintenance and safety risks and ensuring continuous data availability compared to traditional hardware-based solutions. This review examines current emissions monitoring technologies, regulatory frameworks for emissions monitoring, and PEMS model development, with a focus on machine learning techniques such as LSTM, TCN and stacked model architecture, which are recommended for enhancing PEMS accuracy and predictive capabilities. Machine learning-based PEMS has significant advantages in handling complex, non-linear problems where simpler models may struggle. By synthesizing recent advancements, this study underscores AI-driven emissions management as a crucial step in digital transformation, optimizing efficiency, ensuring compliance, and promoting sustainability.