基于BP神经网络的核安全文化星级评价体系

Nuclear Safety Culture Star-Class Assessment System Based BP Neural Network

为了加强核电安全文化建设,本文提出了一种对核电安全文化进行科学、全面评价的方法。并根据核电安全文化的特点,以SMART准则为依据,从安全意识、安全价值观、安全行为、安全现状等方面确立了24项安全文化评价指标,提出了安全文化星级划分标准,并在Visual Basic 6.0平台上建立了基于BP神经网络的安全文化星级评价体系。通过泛化能力测试,该体系能准确地评价出核电安全文化发展到了什么阶段,具有良好的可行性和有效性,操作简便,易于推广。

In order to build the safety culture for nuclear power industry, tt is important to evaluate the safety culture scientifically. Considering the traits of safety culture in the nuclear power industry, 24 safety culture assessment indexes are established from 4 aspects such as Safety consciousness, Safety attitude, Safety action and Safety actuality by using the SMART criteria. Safety culture star-class assessment criterion is presented and safety culture star-class assessment system is developed by using Visual Basic 6.0 and BP neural network. The system has a better generalization ability, and it can show exactly which phase the safety culture is in. Experimental results show that safety culture star-class assessment is practical and easy to perform.