摘要
空调配管系统的减振降噪是空调结构开发中的一个设计难点,是制约空调整体品质的一个关键参数。针对空调仿真优化设计中计算成本和计算精度之间的矛盾,将统计学中的近似模型技术应用到空调配管系统的阻尼优化。采用正交试验设计与均匀试验设计相结合的试验设计方法,建立了多项式响应面、Kriging最优内插、BP神经网络近似模型,研究了阻尼层位置、宽度等参数与结构响应频率、阻尼比之间的近似映射关系。最后采用多目标遗传算法分析了结构阻尼比与结构质量之间的Pareto曲线,并选择最优结果进行了试验验证。研究表明,采用近似模型的阻尼层优化方法可以有效地提高设计效率,降低成本,为空调系统的仿真提供了一种可行的方法。
An approximation model method, based on the statistical theory, was applied to optimal damping design of pipe system in air conditioner. A hybrid experimental design approach, including the orthogonal and uniform experimental design, was introduced to produce the training data. The approximate model reflecting the mapping relation between the damping parameters, such as layers＇ positions, widths, etc. And the responses＇ frequency and damping ratio was established by using response surface method, Kriging optimal interpolation and neural network. The Pareto curve between damping ratio and structural mass was given by using NSGA-Ⅱ algorithm. Tests were carried out to validate the optimization result. The approximation model method provided enables to shorten the design process and reduces the computational cost, which gives a feasible method for air conditioner simulation.
出处
《振动与冲击》
EI
CSCD
北大核心
2013年第6期185-189,194,共6页
Journal of Vibration and Shock
关键词
空调配管
近似模型
多目标优化
遗传算法
pipe system in air conditioner
approximation model
mulit-objective optimization
genetic algorithm
