利用遗传算法进行振动吸振器的优化设计

为了解决振动吸振器优化设计的非线性多维离散问题，引入了全局搜索遗传算法，以频率范围内的声功率峰值最小作为目标函数，对吸振器的位置及结构参数进行了优化设计。采用了混合编码与ｓｔｅａｄｙｓｔａｔｅ算法操作，以提高遗传算法的性能，用位移响应基础法替代传统的有限元方法计算了声功率值，节约了遗传算法的计算时间。最后以悬臂板梁为例，对算法进行了验证。

Vibration Absorber Design for Reducing Structural Sound Level Using Genetic Algorithms

Optimal design of vibration absorbers for reducing structural noise level often requires finding the minima of highly nonlinear multi dimensional functions. Genetic algorithms are introduced as a new promising tool for numerical optimization of such problems. The application presented is on the control of the acoustic response of a cantilever plate beam carrying vibration absorbers. Genetic algorithms have been used to determine the discrete locations and physical parameters of the absorbers. A new mixed coding method is adopted to improve the ability of genetic algorithms. Several cases in different frequency bands show that the improved algorithms are efficient.