| 摘要:以柴油机缸盖为研究对象，在流固耦合计算缸盖温度场的基础上，探讨了应用导热反问题理论求解缸盖燃气侧传热边界的方法。开发了基于LM算法的反求程序，有效解决了算法的准确性和收敛性等问题。建立了缸盖温度场与其燃气侧传热边界的BP神经网络，利用遗传算法对该神经网络进行了优化。结果表明:优化后的网络测试样本的输出值相对误差较小，且计算得到的缸盖温度值相对误差也较小。 |
| 关键词: 缸盖 温度场 导热反问题 神经网络 优化 |
|
| Research on Inverse Problem Method of Temperature Field of Cylinder Head |
| Zha Changsong,Guo Xu,Zhang Ping |
| Naval Deputy Office of Wuhu District, Anhui Wuhu 241000;Naval University of Engineering, Hubei Wuhan 430033 |
| Abstract:Taking the cylinder head of diesel engine as the research object, the solution to the cylinder head heat transfer condition on fire deck applying the inverse heat transfer theory was discussed, based on the fluid structure coupling calculation methods. The new program based on the LM algorithm was developed, which could effectively solve the accuracy and computational convergence of the LM algorithm. Furthermore, a BP neural network of cylinder head temperature field and heat transfer condition on fire deck was established. Optimized results showed the error of testing samples in testing values and simulation values was small, and the same as the temperature field of cylinder head in testing values and simulation values. |
| Key words: cylinder head temperature field inverse heat transfer neural network optimization |
