Scholarlink

In the field of electrical power distribution, a paradigm change has occurred with the development of conventional power grids into complex, intelligent systems known as Smart Grids. Cyberattack susceptibility is a major problem for smart grid systems. Our suggested method Krill Herd assisted Twin Support vector machine (KH-TSVM) requires the installation of strong cybersecurity safeguards to overcome this drawback. We started by gathering a KDD Cup'99 dataset. KDD'99 is a larger data set, including 4,898,430 records. The suggested technique, KH-TSVM, improves the precision and effectiveness of DoS attack detection systems by fusing the benefits of twin support vector machines with krill herd optimization. The goal of the KH-TSVM framework is to enhance the TSVM's generalization and classification skills by optimizing its hyperparameters using the KH algorithm. According to precision (82.49%), recall (83.89%), accuracy (85.48%) and F1-score (85.95%), the model that is suggested works better. According to these results, there is potential for the KH-TSVM strategy to strengthen Smart Grid Infrastructure security against changing cyber threats.

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

Machine Learning-Based Dos Attack Detection Technique for Smart Grid Infrastructure

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Kamalesh Thoppaen Suresh Babu,
  • Mital Patel,
  • Rupali Attarde,
  • R. Mohan Kumar

摘要

In the field of electrical power distribution, a paradigm change has occurred with the development of conventional power grids into complex, intelligent systems known as Smart Grids. Cyberattack susceptibility is a major problem for smart grid systems. Our suggested method Krill Herd assisted Twin Support vector machine (KH-TSVM) requires the installation of strong cybersecurity safeguards to overcome this drawback. We started by gathering a KDD Cup'99 dataset. KDD'99 is a larger data set, including 4,898,430 records. The suggested technique, KH-TSVM, improves the precision and effectiveness of DoS attack detection systems by fusing the benefits of twin support vector machines with krill herd optimization. The goal of the KH-TSVM framework is to enhance the TSVM's generalization and classification skills by optimizing its hyperparameters using the KH algorithm. According to precision (82.49%), recall (83.89%), accuracy (85.48%) and F1-score (85.95%), the model that is suggested works better. According to these results, there is potential for the KH-TSVM strategy to strengthen Smart Grid Infrastructure security against changing cyber threats.