Research On Fluid Temperature Inversion Algorithms For One-dimensional Unsteady Convection Heat Transfer

Thermal stratification and thermal fluctuation are easy to occur in hot and cold fluid intersection,and long term fluctuations lead to thermal fatigue due to random thermal stresses.In the nuclear power plant regulator wave tube and safety injection pipeline,thermal fatigue often occurs due to the thermal stratification of hot and cold fluid,even causing a major accident by fatigue cracks,seriously.Therefore,it is important to know the fluid temperature in the pipeline in time.Without destroying the whole structure of the pipeline,it is one of the focuses of the scientific and technical research to obtain the fluid temperature accurately.Based on the inverse heat conduction problem,in this dissertation,a one-dimensional unsteady heat conduction model is constructed.Using simple measured external surface temperature,this dissertation plans to inverse fluid temperature of internal convective heat transfer.The internal temperature is obtained under the condition of ensuring the integrity of the structure.Accurate solution of heat conduction forward problem is the precondition of solving inverse problem.So,under the premise that all the parameters are known and based on the finite difference method,the temperature field of the unsteady heat conduction model is solved accurately and then the exact value obtained from the positive problem is used as a measure of the inverse problem.The influence of sampling time and mesh size on the accuracy of numerical solution is analyzed in this dissertation.Based on the QPSO algorithm,an inverse problem model is constructed using external surface measurements to estimate the internal fluid temperature.The accuracy of the inverse problem model is verified by the exact temperature value of the positive problem.Based on the research of the above algorithms,a hybrid algorithm combining conjugate gradient method to reduce the time cost is designed.The characteristics of the two algorithms are compared and analyzed,and the influence of various parameters on the performance of the algorithm is analyzed.Then,the engineering of the algorithm is completed,the algorithm is encapsulated by MFC,and the platform of human-computer interaction is built.Finally,a forward process generation device is designed,which lays the foundation for the development of the following device.Experimental results show that the QPSO algorithm has the advantages of strong global search ability,high inversion accuracy and good robustness.The disadvantage of this algorithm is that it has a long computing time.The hybrid algorithm has fast convergence and good stability,but the inversion accuracy is slightly lower than that of the QPSO algorithm.The hybrid algorithm takes advantage of the fast convergence speed of the CG method.At the same time,the QPSO algorithm is used to preliminarily estimate the amount to be retrieved,thus avoiding the shortcoming of the initial value of the CG method.Although the hybrid algorithm sacrifices some accuracy,it reduces the computation time greatly.