基于人工神经网络的LPG公路运输的风险评价模型及其应用

为使LPG公路运输的风险评价更加客观、合理,遵循区域性、动态性和可量化原则,建立了LPG公路运输的风险评价指标,并将个人风险和社会风险作为评价结果。利用人工神经网络的自组织、自学习和自适应功能,提出基于人工神经网络理论的LPG公路运输的风险评价方法,通过Matlab软件中的神经网络工具箱对评价模型进行训练,并通过实践验证该模型的可行性。结果表明,该方法能较好地克服评价深度、评价指标间复杂关系和评价指标作用模糊的限制,且评价结果直观的反映了评价对象的个人风险和社会风险的风险程度。

Risk Assessment Model for Road Transportation of LPG Base on Artificial Neural Network and Its Application

In order to make the risk assessment of road-transportation of LPG(liquefied petroleum gas)more objective and rational, the risk assessment indicators of LPG are built with the individual risk and social risk as the assessment result according to regional, dynamic and quantifiable principles. Then, a risk assessment method for road-transportation of LPG is put forward by use of the self-organization, self-learning and self-adaptive functions of artificial neural network. The neural network toolbox in Matlab software is used to train this assessment model, and the model is practically demonstrated to be feasible. The result shows that the proposed method overcomes many limitations from assessment depth, complex relationships between indicators and fuzziness of assessment indicators. Meanwhile, the assessment result can also intuitively reflect the level of both individual risk and social risk of assessment objects.