The Study Of Thermodynamic Dynamic Model And Simulation Of Split-stirling Cryocooler

Numerical calculative simulation could factually reflect the dynamic characteristics of the whole equipment and inner flow and diathermancy of split-Stirling crycooler. Through the calculation, the velocity, temperature, pressure and other detailed information of airflow in any position inner the crycooler could be gained; the distribution of each parameter could be confirmed and reasonable explanation for the experiment result could be made. All of these had established the solid foundation for optimizing design, shorting the period of development and experiment of crycooler.Based on the characteristics of thermodynamics, the physical model of split-Stirling crycooler was set up and artificial neural networks were imported into the model. And also, the dynamic simulative models of compression chamber, hot chamber, cold chamber, connecting hose and regenerator were built and calculated using the node analytical method. Introducing neural networks into models and altering the input parameters of neural networks could improve the efficiency of neural networks. Adopting BP model calculative method, that is trainbpx, could shorten the time of model exercitation and study and realize the amalgamation between neural networks and thermodynamic models of crycooler. The learned neural networks model could be used in the thermodynamic calculation of crycooler.Compared with simulative value and experimental value of models, it indicated that the models were reliable and proved the emulational feasibility of artificial neural networks models in split-Stirling crycooler, which could improve the precision of crycooler model greatly and provide strong theory for the research and development of actual crycooler.