论文部分内容阅读
为了克服由于实际装置的复杂性及生产工艺的差异对冷凝器稳态仿真精度的影响 ,提高冷凝器仿真模型的通用性和准确性 ,提出了冷凝器基本模型结合人工神经网络的仿真思路 .以相区划分和制冷剂出口焓值迭代为基础 ,提出了一种稳定的逆流型冷凝器仿真分布参数模型和算法 ,建立了冷凝器仿真的基本模型 .其计算结果与实验数据的变化趋势一致 ,能够在定性上反映实际物理过程的基本特性 .通过对部分实验数据的学习 ,进一步建立了与基本模型相结合的人工神经网络 .利用其非线性映射能力进行模型修正 ,显著提高了冷凝器的仿真精度 ,从而为同时提高冷凝器仿真的通用性和准确性提供了一种有效的工程应用方法
In order to overcome the complexity of the actual plant and the difference of the production process on the steady-state simulation accuracy of the condenser and to improve the universality and accuracy of the condenser simulation model, the simulation model of the basic model of the condenser combined with the artificial neural network Based on the iteration of enthalpy value, phase partition and refrigerant outlet, a stable parameter model and algorithm of countercurrent condenser simulation distribution was proposed and the basic model of condenser simulation was established. The calculated results are consistent with the trend of experimental data, Which can qualitatively reflect the basic characteristics of the actual physical process.According to some experimental data, we further establish the artificial neural network combined with the basic model.Using its nonlinear mapping ability to modify the model, the simulation of the condenser is significantly improved Accuracy, thus providing an effective engineering application method for improving the versatility and accuracy of the condenser simulation at the same time