【摘要】 以某T-5200型LNG双圆筒储罐为研究对象,考虑罐体长度、罐体球心距、气室高度、气室半径、垫木厚度、垫木与罐体夹角、绝热层导热系数和垫木导热系数等影响因素,基于有限元法对其进行参数化建模,利用MATLAB与ANSYS进行数值实验模拟出各因子在多种组合下的储罐漏热量。基于数值结果,用均值分析法对各因子的效应进行定量分析。用MATLAB建立基于BP算法的人工神经网络预测模型并进行改进,测试表明,建立的模型预测罐体蒸发率误差在[-5%,5%]的比率为100%,可用于船型及储罐方案设计和论证阶段的蒸发率快速预报。更多还原

【Abstract】 Taking the T-5200 type double cylinder LNG storage tank as the research object,considering the length of the tank,the distance between tank center,gas chamber height,gas chamber radius,pad thickness,the angle between pad and tank,thermal conductivity of wood pad and insulation layer,the parametric FE model was established. The heat leakage of LNG storage tank was calculated under various combinations by MATLAB and ANSYS. The effect of each factor was quantitatively analyzed falling back on the analysis of means(ANOM). The artificial neural network(ANN) improved prediction model was established based on the BP algorithm. Test showed that the error of the evaporation rate of the tank predicted by the established model is within 5%,which can be used for rapid prediction of evaporation rate in the design and demonstration phase of ship type and cargo tank.更多还原