火焰筒浮动瓦块的壁温-结构一体化优化

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.

Optimal design for temperature and structure of the float-wall of flame tube

Taking the float-wall structure as the object of study,neural network and genetic algorithm were used to settle synchronistically the optimization of cooling structures and install position.Using genetic algorithm to perform cooling structures and install position optimization,and taking the result of install position optimization as the foundation of cooling structures optimization,thus finally realized the synchronized optimizations of cooling structures and install position.In order to enhance the efficiency of algorithm,neural network was used to perform mapping on the result of install position optimization so as to replace its optimization processes.The result of calculating example shows that the method being mentioned is highly active,accurate and possesses good application value to engineering.