工矿自动化2025，Vol.51Issue(9)：1-9,9.DOI:10.13272/j.issn.1671-251x.2025080041
基于数字孪生的煤矿工作面历史场景回溯与故障追溯技术
Coal mine working face historical scene reconstruction and fault tracing technology based on digital twin
摘要
Abstract
To address the problems of fragmented multi-source data,spatiotemporal misalignment,and low fault diagnosis efficiency in coal mine working face historical scene reconstruction and fault tracing,this study investigated a coal mine working face historical scene reconstruction and fault tracing technology based on digital twin.Six key technologies were used to realize complex scene reconstruction and fault tracing:① a high-precision 3D model at the scale of equipment such as the"three machines"was constructed using Physically-Based Rendering(PBR).② A hierarchical data preloading mechanism was proposed to establish a three-level loading system consisting of a dynamic data layer,a static data layer,and an event data layer.③ Full-element state synchronization of equipment such as shearers and hydraulic supports was carried out,focusing on geometric position and orientation,physical parameters,behavior logic,and rule-based judgment.④A multi-host time unification algorithm based on shearer position data was proposed,which performed feature extraction and time alignment on position data recorded by each host,thereby realizing precise time synchronization of multi-source heterogeneous data.⑤ A playback consistency verification mechanism was established,and process parameter similarity was used to quantify playback accuracy.⑥ A support loss fault tracing algorithm was designed,and a multi-variable coupled support-shifting dynamics model was established.A three-level early warning mechanism was implemented based on the matching degree between theoretical and actual stroke.Experimental results showed that the hierarchical data preloading algorithm maintained stable running time,with an average data loading time of 35.5 s,improving efficiency by 29.3％compared with the full backup algorithm.The average similarity of the middle-section follow-up parameters was relatively low,at 93.73％,while the average similarity of the follow-up parameters in the triangular coal area was relatively high,at 98.56％.Through historical scene reconstruction of the working face,changes in the appearance color of the support with support loss could be observed,which verified the correctness of the support loss fault tracing algorithm.关键词
煤矿工作面/数字孪生/历史场景回溯/丢架故障追溯/分层预加载/全要素状态同步/多主机时间统一算法/回放一致性验证Key words
coal mine working face/digital twin/historical scene reconstruction/support loss fault tracing/hierarchical preloading/full-element state synchronization/multi-host time unification algorithm/playback consistency verification分类
矿业与冶金引用本文复制引用
李天越,刘清,董启凡,冯沛..基于数字孪生的煤矿工作面历史场景回溯与故障追溯技术[J].工矿自动化,2025,51(9):1-9,9.基金项目
国家重点研发计划项目(2023YFC2907504). （2023YFC2907504）
