基于遗传算法的叶盘压电控制回路优化

针对简单遗传算法收敛速度慢及易产生"早熟"的问题,对遗传算法选择算子和变异算子进行了改进。基于改进遗传算法,对随机失谐叶盘结构的压电控制回路进行了二模式和四模式优化,对比分析了两种优化电路对叶盘结构振动的控制效果及成本。研究结果表明,相比谐调压电控制回路,优化后的二模式和四模式主动失谐控制回路都有更好的振动控制效果,且二模式回路成本更低;改进遗传算法收敛速度快,优化结果具有全局性;叶片刚度失谐强度对主动失谐压电控制回路的振动控制效果有明显的影响。

Optimization on the piezoelectric network of bladed disk based on genetic algorithm

In order to avoid the defect of slow convergence and "immature convergence" existed in the traditional genetic algorithm,the simple genetic algorithm’s selection operator and mutation operator were improved.Then piezoelectric network was optimized based on improved genetic algorithm,the ratio of control effect to cost was separately analyzed when each circuit parameter has two values or four values.The results show that compared with the tuned piezoelectric network,bladed disk vibration is suppressed better under the control of a mistuned network which each circuit parameter has two values or four values,and the network which each circuit parameter has two values is more economical.The convergence of improved genetic algorithm is faster and the optimizing result is global.The vibration suppressed effect of the mistuned piezoelectric network is affected by the mistuned value of the blade stiffness.