论文部分内容阅读
以火焰筒浮动瓦块结构为研究对象,提出了利用神经网络与遗传算法相结合的方法对浮动瓦块中冷却结构变量和安装位置变量进行同步优化.利用遗传算法分别对冷却结构变量和安装位置变量进行优化,将安装位置优化结果作为冷却结构变量优化中遗传操作的依据,最终实现冷却结构和安装位置的同步优化.为了通过计算效率,利用神经网络对安装位置的优化结果进行映射取代其优化过程.算例结果表明:该方法高效、精确,有很好的工程实用价值.
In this paper, aiming at the structure of the floating tile of flame tube, the method of neural network and genetic algorithm is proposed to synchronously optimize the cooling structure variables and the installation position variables in the floating tile.By using genetic algorithm, the influence of cooling structure variables and installation location Variables are optimized and the result of installation position optimization is taken as the basis of genetic operation in the optimization of cooling structure variables to achieve the synchronization and optimization of the cooling structure and the installation location finally.In order to optimize the installation location by neural network The results show that the method is efficient, accurate and has good practical value.