<p>The rapid expansion of natural gas pipeline networks in China necessitates robust reliability assessment and optimization frameworks, particularly for large-scale looped configurations where traditional tree-based models fall short. This study proposes an integrated framework combining connectivity reliability evaluation with adaptive topology optimization. First, a minimum path set-based reliability model is developed, leveraging an enhanced depth-first search (DFS) algorithm for efficient path identification and binary decision diagrams (BDD) to eliminate 92% of redundant terms in reliability formulas, reducing computational complexity by 40% compared to Monte Carlo simulations. Second, an adaptive genetic algorithm (AGA) is designed to optimize network topology, dynamically adjusting crossover and mutation rates (0.8≤<InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_98749_Article_IEq1.gif" Format="GIF" Height="12" Rendition="HTML" Resolution="72" Type="Linedraw" Width="17" /> </InlineMediaObject> <EquationSource Format="TEX">\(\:{p}_{c}\)</EquationSource> </InlineEquation>≤0.01, 0.01≤ <InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_98749_Article_IEq2.gif" Format="GIF" Height="12" Rendition="HTML" Resolution="72" Type="Linedraw" Width="23" /> </InlineMediaObject> <EquationSource Format="TEX">\(\:{p}_{m}\)</EquationSource> </InlineEquation>≤ 0.8) based on population diversity, while enforcing constraints through penalty functions (node degree <InlineEquation ID="IEq3"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_98749_Article_IEq3.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="35" /> </InlineMediaObject> <EquationSource Format="TEX">\(\:{k}_{max}\)</EquationSource> </InlineEquation>=4, pipeline length <InlineEquation ID="IEq4"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_98749_Article_IEq4.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="31" /> </InlineMediaObject> <EquationSource Format="TEX">\(\:{l}_{max}\)</EquationSource> </InlineEquation>=120 km). Case studies on a regional pipeline network (89 nodes, 98 segments) demonstrate that loop structures exhibit 25.7% higher average reliability (<InlineEquation ID="IEq5"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_98749_Article_IEq5.gif" Format="GIF" Height="19" Rendition="HTML" Resolution="72" Type="Linedraw" Width="21" /> </InlineMediaObject> <EquationSource Format="TEX">\(\:{R}_{j}\)</EquationSource> </InlineEquation>= 0.87792) than branch nodes (v79: <InlineEquation ID="IEq6"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_98749_Article_IEq6.gif" Format="GIF" Height="19" Rendition="HTML" Resolution="72" Type="Linedraw" Width="21" /> </InlineMediaObject> <EquationSource Format="TEX">\(\:{R}_{j}\)</EquationSource> </InlineEquation>=0.60933). The AGA-driven optimization increases system-wide connectivity reliability (<InlineEquation ID="IEq7"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_98749_Article_IEq7.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="32" /> </InlineMediaObject> <EquationSource Format="TEX">\(\:{R}_{SU}\)</EquationSource> </InlineEquation>) from 0.03 to 0.247 by strategically adding redundant pipelines (v71–v77), outperforming particle swarm optimization (PSO) by 65%. Key findings reveal that centralized gas source layouts and looped configurations significantly enhance redundancy, with critical segments showing 34% higher D-connectivity importance post-optimization. This work provides a scalable, training-free solution for pipeline network design, balancing computational efficiency (68.7s for 200-node networks) with engineering constraints, and offers actionable insights for infrastructure resilience enhancement.</p>

Research on the connectivity reliability analysis and optimization of natural gas pipeline network based on topology

期刊论文
出版模式：

开放获取
发表日期：

2025年4月18日

查看全文

Xiuxuan Yang,
Kun Chen,
Minghui Liu

摘要

The rapid expansion of natural gas pipeline networks in China necessitates robust reliability assessment and optimization frameworks, particularly for large-scale looped configurations where traditional tree-based models fall short. This study proposes an integrated framework combining connectivity reliability evaluation with adaptive topology optimization. First, a minimum path set-based reliability model is developed, leveraging an enhanced depth-first search (DFS) algorithm for efficient path identification and binary decision diagrams (BDD) to eliminate 92% of redundant terms in reliability formulas, reducing computational complexity by 40% compared to Monte Carlo simulations. Second, an adaptive genetic algorithm (AGA) is designed to optimize network topology, dynamically adjusting crossover and mutation rates (0.8≤ \(\:{p}_{c}\) ≤0.01, 0.01≤ \(\:{p}_{m}\) ≤ 0.8) based on population diversity, while enforcing constraints through penalty functions (node degree \(\:{k}_{max}\) =4, pipeline length \(\:{l}_{max}\) =120 km). Case studies on a regional pipeline network (89 nodes, 98 segments) demonstrate that loop structures exhibit 25.7% higher average reliability ( \(\:{R}_{j}\) = 0.87792) than branch nodes (v79: \(\:{R}_{j}\) =0.60933). The AGA-driven optimization increases system-wide connectivity reliability ( \(\:{R}_{SU}\) ) from 0.03 to 0.247 by strategically adding redundant pipelines (v71–v77), outperforming particle swarm optimization (PSO) by 65%. Key findings reveal that centralized gas source layouts and looped configurations significantly enhance redundancy, with critical segments showing 34% higher D-connectivity importance post-optimization. This work provides a scalable, training-free solution for pipeline network design, balancing computational efficiency (68.7s for 200-node networks) with engineering constraints, and offers actionable insights for infrastructure resilience enhancement.