中国机械工程

• 机械基础工程 • 上一篇    下一篇

响应面模型与混合优化算法相结合的锯片参数优化设计

田永军;段国林;夏晓光;张萼   

  1. 河北工业大学,天津,300130
  • 出版日期:2016-11-25 发布日期:2016-11-23
  • 基金资助:
    天津市自然科学基金重点资助项目(11JCZDJC23100);河北省自然科学基金资助项目(F2014202241) 

Optimization Design of Low Noise Circular Saw Parameters Combining Response Surface Model with Hybrid Optimization Algorithm

Tian Yongjun;Duan Guolin;Xia Xiaoguang;Zhang E   

  1. Hebei University of Technology,Tianjin,300130
  • Online:2016-11-25 Published:2016-11-23
  • Supported by:
     

摘要: 针对传统算法在锯片声学特征优化中的局限性,提出了一种将二阶响应面模型与混合算法相结合的优化设计方法。在设计区域内应用D-optimal试验设计法抽取样本点,分别通过显式算法获取锯片声学、应力响应以及隐式算法获取锯片变形量,并用试验验证了数值模型的准确性,然后建立了由6个变量参数所决定的锯片的声学、应力以及刚度的二阶响应面模型;利用自适应模拟退火法和蛙跳混合算法对响应面模型进行循环逼近,获得了设计变量影响度以及最优结果。结果表明,在保证刚度和应力许可条件下,通过有限次数值分析,经全局优化后的最佳结构可降低空载噪声4~7dB。数字算例表明,该方法适用于旋转类刀具的声学性能优化设计。

关键词: 锯片, 响应面模型, 混合优化算法, 动态, 噪声

Abstract: Focusing on inherent limitations in the traditional optimization methods for structure parameters of low noise circular saw, an optimization combining response surface model with hybrid algorithm was proposed to optimize the design of dynamic rotating blades. Firstly, a set of experimental design data points were extracted by D-optimal experimental design scheme,then the points were calculated respectively by display algorithm for acoustic, stress response and implicit algorithm for deformation saw blades.Then the corresponding response surface models were set up by the points' responses. Finally, these second-order regression models were optimized by adaptive simulated annealing and leapfrog algorithm. The results show that under the conditions of safety, by a limited number of value analysis after the optimization, the final structure may reduce the idle noise 4~7dB.Numerical example indicates that this method is suitable for the optimization design of acoustic properties of rotating cutting tools.

Key words: circular saw, response surface model, hybrid optimization algorithm, dynamic, noise

中图分类号: