为有效控制海运环境下易流态固体散货因液化而导致的船舶倾覆事故风险，基于船舶运动学及波能理论，采用AWQA水动力计算软件，模拟分析了全球固体散货海运主力船型在各典型海域航行过程中所遭遇的外界环境载荷，并根据土动力学理论及三轴试验结果对相关外界环境载荷进行了货物液化风险的等效分析，在此基础上构建了货物液化风险评估模型。基于风险评估模型及其数值计算样本，采用BP神经网络算法完成了货物液化风险快速预测模型，实现了货物海运液化风险的快速智能预测，并通过模型的权值矩阵及阈值矩阵得到BP快速预测模型的显性表述。运用该预测模型，对4种典型易流态固体散货海运环境下的液化风险进行了快速预测，预测结果和评估模型给出的结果具有很好一致性。该预测模型可基于船型和海况等参数快速、有效预测易流态固体散货海运液化风险，为一线海事监管提供了有效安全监管方法，同时对IMSBC规则的发展也提供了良好的促进作用。

In order to effectively manage the ship capsizing risk caused by liquefaction of liquefiable solid bulk cargo during sea transport, based on ship kinematics and wave energy theory, AWQA hydrodynamic analysis software was used to simulate and analyze the external environmental loads encountered by ships while sailing in various typical sea areas. According to the theory of soil dynamics and results of triaxial tests, the equivalent analysis of the liquefaction risk of the cargo and the related assessment model were carried out. On account of the assessment model and numerical calculation samples, BP propagation neural network algorithm was applied to construct the predict model, which realized the quick intelligent assessment of the liquefaction risk of cargo shipping. The dominant expression of the BP quick predict model was obtained through the weight matrix and threshold matrix of the model, and the predict model was used to quick assess the liquefaction risk of four typical liquefiable solid bulk cargos under shipping environment. The predict results were in good agreement with the assessment results. This risk predict model can quickly and effectively assess the liquefaction risks of solid bulk cargo according to the parameters such as ship type and sea conditions, which provides an effective method for the front-line maritime supervision, and a good promotion for the development of IMSBC Code.