基于深度学习的电力工作票生成方法

Generation Method of Electric Work Ticket Based on Deep Learning

随着电网基层员工减负工作的逐步推进,简化电力工作票填写的需求日益增加。为了解决电力工作票填写中存在的安全措施填写不规范、审查耗时长的问题,提出一种基于深度学习的电力工作票生成方法。该方法采用长短期记忆网络学习历史的电力工作票开票信息,获得电力工作票安全措施经验库;在此基础上,对当前的电力工作票内容进行语义分析,自动生成满足工作条件的安全措施。该方法在仿真环境中进行了验证,其安全措施生成准确率为90%。该方法可提高电力工作票的自动生成能力,提升电网基层精益化管理水平。

With the gradual advancement of the work of reducing the burden of grass-roots employees in the power grid,the demand for simplifying the filling of electric work tickets is increasing.In order to solve the problems of irregular filling of safety measures and long review time in the filling of electric work tickets,a deep learning-based electric work ticket generation method is proposed.The method uses long short-term memory network to learn the historical power work ticket billing information,and obtains the power work ticket safety measure experience database.The method is validated in a simulation environment,and its security measure generation accuracy is 90%.This method can improve the ability of automatic generation of electric work tickets and improve the level of lean management at the grassroots level of the power grid.