论文部分内容阅读
采用人工神经网络结合遗传算法,利用周向涡量,对一单级低速压气机转子空间弯掠积叠曲线进行了气动数值优化.优化目标为小流量工况下绝热效率最大.结果表明,优化后转子出口的周向涡量分布改善,压气机级在小流量区域的效率得到很大提高,稳定工作裕度得以拓宽.该方法不仅在数学和物理上使得优化进程高效的收敛于最优解,而且能给后续的压气机级提供一个组织有序的流场.
Using artificial neural network and genetic algorithm, the circumferential vorticity is used to optimize the aerodynamic numerical value of the swept over lag curve of a single-stage low-speed compressor rotor. The optimization objective is to maximize the adiabatic efficiency under the conditions of low flow rate. The distribution of the circumferential vorticity at the outlet of the rotor is improved, the efficiency of the compressor stage in the small flow area is greatly improved, and the stability of the working margin is broadened. This method not only makes the optimization process converge optimally in the optimal solution both mathematically and physically , But also provide a well-organized flow field for subsequent compressor stages.