Scholarlink

<p>Driven by new quality productive forces, communication networks and energy transmission systems are increasingly integrated, creating environments with high loads, real-time constraints, and multiple interdependencies. This paper proposes a Pareto-based bi-objective path optimization framework—minimizing path length while maximizing reliability—combined with a dynamic failure-response mechanism to support industrial control (IC) instruction flows in the Energy Industrial Internet. Transmission dependability is modeled using a multiplicative reliability accumulation approach, while predefined failure-node sets and topology-pruning strategies mitigate disruptions from failures in critical infrastructure such as converter stations and substations, ensuring dynamic reconfiguration and uninterrupted flow delivery under abnormal conditions. The study advances in three dimensions: (1) constructing an instruction semantic labeling and reliability mapping mechanism that builds a semantically trusted knowledge graph; (2) enabling dynamic restoration of sessions under faults by using the Length-Reliability (LR) index to predict instruction execution performance, guiding scheduling decisions; and (3) recognizing instruction-flow patterns to identify both robust and risk-prone paths, forming a closed-loop “perception-decision-execution” optimization framework. Overall, the research achieves cross-layer fusion from path-level to instruction-level reliability, enhancing the stability, efficiency, and resilience of industrial control flows in complex energy Internet environments. Unlike conventional Pareto-based routing methods that employ additive cost accumulation, this study introduces a novel multiplicative reliability accumulation mechanism combined with depth-dependent pruning, forming the first integrated LR–Dijkstra framework capable of real-time self-reconfiguration under node and edge failures. This innovation provides a unified path search and fault recovery paradigm for reliability-critical industrial Internet applications.</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

A bi-objective path optimization for highly reliable industrial control flows in the energy internet

  • 期刊论文

  • 发表日期：

    2026年3月10日

查看全文
  • Xuan Zhang,
  • Hao-Hui Su,
  • LvJun Zheng,
  • XiaoJie Shen,
  • Hua Liao

摘要

Driven by new quality productive forces, communication networks and energy transmission systems are increasingly integrated, creating environments with high loads, real-time constraints, and multiple interdependencies. This paper proposes a Pareto-based bi-objective path optimization framework—minimizing path length while maximizing reliability—combined with a dynamic failure-response mechanism to support industrial control (IC) instruction flows in the Energy Industrial Internet. Transmission dependability is modeled using a multiplicative reliability accumulation approach, while predefined failure-node sets and topology-pruning strategies mitigate disruptions from failures in critical infrastructure such as converter stations and substations, ensuring dynamic reconfiguration and uninterrupted flow delivery under abnormal conditions. The study advances in three dimensions: (1) constructing an instruction semantic labeling and reliability mapping mechanism that builds a semantically trusted knowledge graph; (2) enabling dynamic restoration of sessions under faults by using the Length-Reliability (LR) index to predict instruction execution performance, guiding scheduling decisions; and (3) recognizing instruction-flow patterns to identify both robust and risk-prone paths, forming a closed-loop “perception-decision-execution” optimization framework. Overall, the research achieves cross-layer fusion from path-level to instruction-level reliability, enhancing the stability, efficiency, and resilience of industrial control flows in complex energy Internet environments. Unlike conventional Pareto-based routing methods that employ additive cost accumulation, this study introduces a novel multiplicative reliability accumulation mechanism combined with depth-dependent pruning, forming the first integrated LR–Dijkstra framework capable of real-time self-reconfiguration under node and edge failures. This innovation provides a unified path search and fault recovery paradigm for reliability-critical industrial Internet applications.