风电运维人员行为安全预警指标体系构建与分析  

Construction and analysis of behavior safety early-warning index system for wind power operation and maintenance personnel

作　　者：郑鹏 瞿丽莉[2] 程礼彬 何子春 张银龙 常丁懿 ZHENG Peng;QU Lili;CHENG Libin;HE Zichun;ZHANG Yinlong;CHANG Dingyi(Huadian Electric Power Research Institute Co.,Ltd.,Hangzhou Zhejiang 310030,China;Xi′an Thermal Power Research Institute Co.,Ltd.,Xi′an Shaanxi 710054,China;School of Management,Tianjin University of Technology,Tianjin 300384,China)

机构地区：[1]华电电力科学研究院有限公司,浙江杭州310030 [2]西安热工研究院有限公司,陕西西安710054 [3]天津理工大学管理学院,天津300384

出　　处：《中国安全科学学报》2022年第S01期1-5,共5页China Safety Science Journal

基　　金：中国华电集团有限公司科技项目(CHDKJ21-01-07);天津市研究生科研创新项目(2021YJSB243)。

摘　　要：为提高风电运维人员安全行为水平,在独立性、完备性、梯度性、可行性原则的前提下,从人因、机械设备、作业环境、监督管理、信息沟通5个方面建立行为安全预警指标体系,利用问卷调查法获取行为安全预警数据,基于果蝇优化算法(FOA)优化反向传播(BP)神经网络,建立“15-10-1”结构的行为安全预警模型,利用该模型训练测试问卷数据。结果表明:构建的行为安全预警指标体系是科学合理的,FOA-BP神经网络模型有较强的预警能力,能够预测风电运维人员的行为安全风险。测试后,模型能实现较好的预警效果。In order to improve the safety behavior level of wind power operation and maintenance personnel,on the premise of independence,completeness,ladder and feasibility,a behavioral safety early warning index system was established from five aspects of human factors,mechanical equipment,operating environment,supervision and management,and information communication.A questionnaire survey was used to obtain behavioral safety early warning data.Based on BP neural network optimized by FOA,a behavior safety warning model with the structure of″15-10-1″was established,which was used to train the test data.The results show that the constructed behavioral safety early-warning index system is scientific and reasonable,and the FOA-BP neural network model has a strong early-warning ability and can predict the behavioral safety risks of wind power operation and maintenance personnel.After testing,the model can achieve a better warning effect.

关 键 词：风电运维人员 行为安全预警 指标体系 果蝇优化算法(FOA) 反向传播(BP)神经网络 

分 类 号：X913[环境科学与工程—安全科学]