CNG站模糊综合和神经网络的安全评价

影响CNG站安全的因素和CNG站安全状态之间存在非线性关系,利用神经网络的非线性逼近等特点,可以对CNG站安全状态作出准确可靠的评价.从CNG站的人员、设施、管理和环境4方面考虑建立了CNG站安全评价体系,利用模糊层次分析法确定各因素的权重,将岭型函数作为隶属度函数对CNG站进行模糊综合评价,从而求出神经网络所需的训练样本,最后用训练好的BP神经网络对未知CNG站的安全作出评价.用该模型对3座CNG站进行安全评价,评价结果与模糊综合评价得到的结果完全一致,因此该方法可以用于评价CNG站的安全状态,对CNG站的安全评价和管理提供了一种新的方法.

Fuzzy comprehensive evaluation and neural network safety assessment for CNG station

The factor affecting the security of CNG station and its safe condition have nonlinear nexus and the nonlinear approximation of neural networks can be used to give an accurate and reliable assessment of CNG station. The system of security assessment was established in terms of person,facilities,management and environment. The weight of each factor was determined through fuzzy hierarchy analysis and mountains function was used as subjection degree function in the process. Consequently the training sample which is necessary in neural network was obtained and the trained BP neural network was utilized to evaluate the safety status of CNG station. This model was applied to the security assessment of 3 CNG stations and the result accorded with that obtained through fuzzy integration analysis completely. Therefore this model can be used to evaluate the safety status of CNG station and provides a new method for security assessment and management.