摘要
本研究旨在提出一种新型的铁路应急处置决策支持系统,该系统基于语义理解与生成模型,能够有效生成应急处置策略、验证并更新处置流程,以及智能分发处置任务。首先,通过融合语义理解与生成模型及铁路应急处置专业知识库,处置策略动态生成模块能够自主生成应急处置策略,并提供决策支持。接着,处置策略验证与更新模块利用Petri网模型的多方面可量化特性,实现对生成策略的验证与更新,确保策略的实时性和有效性。最后,应急处置任务智能分发模块采用基于BERT的深度学习模型进行命名实体识别和关系抽取,构建出详细的〈处置任务—归属—部门〉知识图谱,并通过Neo4j图数据库及Cypher查询语言实现任务的智能分发。通过这一综合性决策支持系统,本研究为铁路应急处置领域的智能化调度提供了有效的解决方案,并开拓了大语言模型在该领域应用的新方向。
This study aims to propose a novel railway emergency response decision support system based on semantic understanding and generation models,which is capable of efficiently generating emergency response strategies,validating and updating response processes,and intelligently distributing response tasks.First,the system integrates semantic understanding and generation models with a professional railway emergency response knowledge base for dynamic strategy generation,providing decision support;then the strategy validation and update module leverages the multi-faceted quantifiable features of Petri nets to verify and update generated strategies,ensuring their timeliness and effectiveness.Finally,the emergency response task distribution module employs a BERT-based deep learning model for named entity recognition and relation extraction,constructing a detailed"task-ownership-department"knowledge graph.Through the Neo4j graph database and Cypher query language,it achieves intelligent task distribution.This comprehensive decision support system offers an effective solution for the intelligent scheduling in the railway emergency response domain and new directions for the application of large language models in this field.
作者
王祥昊
杨怀志
王莉
胡恒闯
Wang Xianghao;Yang Huaizhi;Wang Li;Hu Hengchuang(Key Laboratory of Railway Industry for Operational Active Safety Assurance and Risk Prevention and Control,Beijing Jiaotong University,Beijing 100044,China;School of Traffic and Transportation,Beijing Jiaotong University,Beijing 100044,China;Beijing-Fuzhou Railway Passenger Dedicated Line Anhui Co.,Ltd.,Anhui,Hefei 230061,China)
出处
《铁道技术标准（中英文）》
2024年第6期44-51,共8页
Railway Technical Standard(Chinese & English)
基金
国家重点研发计划(2022YFB4300603)
中央高校基本科研业务费(2023JBZY005)。
