Scholarlink

<p>The uniform distribution of U<sub>3</sub>O<sub>8</sub> particles within an aluminum matrix is crucial for the optimal performance and safety of plate-type nuclear fuel produced via the dispersion technique. X-ray radiography is among the primary methods used to assess the homogeneity of U₃O₈-Al dispersion fuel plates. This non-destructive inspection technique involves an X-ray source emitting a beam that passes through the fuel plates, with absorption varying with the material’s density. The resulting image reveals the internal distribution of U<sub>3</sub>O<sub>8</sub> particles, allowing inspectors to identify any inhomogeneities. This study presents an automated inspection system that combines deep learning and optimization algorithms to classify fuel plates as “Homogeneous” (regular) or “non-homogeneous” (observed/defective). Multiple optimization techniques are evaluated to identify the most effective approach, using the Chimp Optimization Algorithm (CHOA) for its superior performance. Leveraging a pre-trained EfficientNet for feature extraction and CHOA for feature selection, the system identifies optimal discriminative features. At the same time, a support vector machine (SVM) classifier, optimized via CHOA, achieves classification accuracies of 99.77% (binary) and 99.72% (ternary) classification. This approach significantly improves inspection reliability and efficiency, ensuring robust nuclear fuel quality control.</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

Automated homogeneity inspection of U3O8-Al dispersion fuel plates using X-ray radiography, deep learning and Chimp Optimization Algorithms

  • 期刊论文

  • 发表日期：

    2026年7月18日

查看全文
  • Zeinab F. Elsharkawy,
  • Refaat M. Fikry,
  • Nadia M. Nawwar,
  • A. G. Ahmed,
  • Mohammed E. El-Bedawy

摘要

The uniform distribution of U3O8 particles within an aluminum matrix is crucial for the optimal performance and safety of plate-type nuclear fuel produced via the dispersion technique. X-ray radiography is among the primary methods used to assess the homogeneity of U₃O₈-Al dispersion fuel plates. This non-destructive inspection technique involves an X-ray source emitting a beam that passes through the fuel plates, with absorption varying with the material’s density. The resulting image reveals the internal distribution of U3O8 particles, allowing inspectors to identify any inhomogeneities. This study presents an automated inspection system that combines deep learning and optimization algorithms to classify fuel plates as “Homogeneous” (regular) or “non-homogeneous” (observed/defective). Multiple optimization techniques are evaluated to identify the most effective approach, using the Chimp Optimization Algorithm (CHOA) for its superior performance. Leveraging a pre-trained EfficientNet for feature extraction and CHOA for feature selection, the system identifies optimal discriminative features. At the same time, a support vector machine (SVM) classifier, optimized via CHOA, achieves classification accuracies of 99.77% (binary) and 99.72% (ternary) classification. This approach significantly improves inspection reliability and efficiency, ensuring robust nuclear fuel quality control.