Scholarlink

<p>The growing use of sensors in water management generates large, heterogeneous data volumes, making automated quality assurance essential. Machine learning (ML) methods can detect and correct errors but require context-aware design, robust evaluation, and often lack access to high-quality reference data. Sensor-specific preprocessing further complicates data use. Combining ML with physics-based models enhances transparency and generalization. This hybrid approach supports trustworthy digital twins and enables reliable, scalable data-driven monitoring in complex environmental systems.</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

Perspektiven einer KI-basierten Daten-Qualitätssicherung

  • 期刊论文

  • 发表日期：

    2025年8月5日

查看全文
  • Karen Schulz,
  • André Niemann,
  • Thorsten Mietzel

摘要

The growing use of sensors in water management generates large, heterogeneous data volumes, making automated quality assurance essential. Machine learning (ML) methods can detect and correct errors but require context-aware design, robust evaluation, and often lack access to high-quality reference data. Sensor-specific preprocessing further complicates data use. Combining ML with physics-based models enhances transparency and generalization. This hybrid approach supports trustworthy digital twins and enables reliable, scalable data-driven monitoring in complex environmental systems.