论文部分内容阅读
利用基于代理模型的整体敏感度分析方法,对考虑模型参数和物质属性的低温介质质量传输空化模型进行了评价,从而获得影响预测精确度的主要因素,并对空化模型参数进行了校准。结果表明:凝结系数对压力和温度预测精度的影响较大,整体敏感度分别为44%和29%;蒸发系数的影响较小,整体敏感度均小于5%。通过代理模型优化分析获得了Merkle空化模型对压力和温度的预测精度随凝结系数的变化趋势:当凝结系数小于20时,随凝结系数的增大,压力预测精度提高,温度预测精度降低;当凝结系数大于20时,压力和温度的预测精度均随凝结系数的增大而降低。优化后的蒸发系数和凝结系数分别为3.8和19.43,满足Pareto最优解,模型预测能力得到提高。
Using the method of holistic sensitivity analysis based on agent model, the cavitation model of low temperature medium mass transfer considering the parameters of the model and the material properties was evaluated, and the main factors affecting the prediction accuracy were obtained. The cavitation model parameters were calibrated. The results show that the coagulation coefficient has great influence on the prediction accuracy of pressure and temperature, the overall sensitivities are 44% and 29% respectively. The influence of evaporation coefficient is small, and the overall sensitivity is less than 5%. Through the optimization of the proxy model, the predicting accuracy of pressure and temperature of Merkle cavitation model with the trend of the condensing coefficient is obtained: when the coagulation coefficient is less than 20, with the increase of the coagulation coefficient, the pressure prediction accuracy increases and the temperature prediction accuracy decreases; When the coefficient of coagulation is more than 20, the prediction accuracy of pressure and temperature will decrease with the increase of coagulation coefficient. The optimized evaporation coefficient and coagulation coefficient are 3.8 and 19.43, respectively. The Pareto optimal solution is satisfied and the model predictive ability is improved.