基于NLP的转向架故障信息处理系统  

Bogie Fault Information Processing System Based on NLP

作　　者：李闻涛[1] 罗敏[1] 黄江山[1] 

机构地区：[1]同济大学铁道与城市轨道交通研究院,上海201804

出　　处：《机电一体化》2017年第4期53-59,共7页Mechatronics

基　　金："十二五"国家科技支撑计划项目(2015BAG19B02);国家科技支撑计划(2015BAG13B01)

摘　　要：搭建了一套基于NLP的转向架故障信息处理系统,针对地铁转向架系统故障信息的实际情形,设计了转向架故障信息处理系统的自然语言处理策略,通过计算句子相似度寻找故障记录在标准故障语句表中对应的表达,并运用该标准故障语句重新形成包含原故障信息内容的标准语句。最后,使用本系统处理了某轨道交通运营单位车辆转向架系统的现场故障信息,验证了算法的有效性。In this paper,we build a bogie fault information processing system based on natural language processing. By researching the real fault information from the bogie system of rail transit,we design a set of processing strategies for bogie fault information. The strategies first look for the standard statement corresponds to the original fault information by computing the sentence similarity,and then use the standard statement to form a new standard statement that contains the message of the original fault information. At last,the system is used to deal with the fault information from the bogie system of a rail transit operation unit,and the effectiveness of the system is verified.

关 键 词：轨道交通 故障信息 自然语言处理 转向架系统 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]