面向关键结构参数的复杂多质量特性稳健设计
发布时间：2018-11-24 08:46
【摘要】：复杂多质量特性稳健设计是工程优化领域的难点,而关键结构参数的复杂多质量特性的稳健设计问题更为复杂。解决此项工程难题需要完成三项工作:关键结构参数的提取、挖掘出复杂多质量特性与关键结构参数之间的关系式、具备最佳关键结构参数条件下保证质量特性达到高稳健水平。在传统工程优化的基础之上,考虑外界因素导致复杂多质量特性产生波动,将稳健设计理念融入优化过程中,展开复杂多质量特性的稳健设计并且获得最佳关键结构参数优化值。以微动平台为研究实例来贯穿全文,提出面向关键结构参数的复杂多质量特性稳健设计,主要研究内容如下:首先,展开微动平台关键结构参数的提取工作。确定各参数属性并构建微动平台模型,剖析微动平台运动原理。为获取设计过程中所需要的试验数据,采用有限元仿真技术手段对微动平台进行模拟仿真。对微动平台所有的结构参数进行正交试验设计制定试验方案,结合微动平台具有尺寸数量多且质量特性复杂的特点,运用灰色关联法对微动平台质量特性和结构参数的关系进行分析,根据结构参数与质量特性的关联程度的大小来确定关键结构参数,完成微动平台关键结构参数的提取工作。其次,在提取微动平台关键结构参数的基础上,构建复杂多质量特性与关键结构参数的数学模型关系式。采用拉丁超立方试验设计方法获取试验方案,通过模拟仿真得到试验方案中复杂多质量特性的数值,根据径向基函数代理模型构建思路搭建微动平台复杂多质量特性与关键结构参数之间的数学模型关系式。为验证该模型关系式的精确度,采用相对平均绝对误差、均方根误差和相对最大绝对误差三项评价标准进行精确度的评估,结果表明构建的数学模型关系式精确度高,能够用于进行后续的稳健设计工作。最后,针对复杂多质量特性进行稳健设计并确定关键结构参数优化值。传统确定性优化可以提高微动平台综合性能,但未考虑随机变动载荷、材料特性等不确定性因素,无法解决由外界因素影响带来的波动性问题。为提高微动平台质量特性稳健水平,在确定性优化决策模型的基础上构建6σ设计数学模型,采用蒙特卡洛法对关键结构参数进行抽样,将抽样结果代入6σ设计数学模型中并运用多岛遗传算法求解,最终得到关键结构参数优化值和各质量特性的均值与标准差,完成微动平台复杂多质量特性稳健设计。为了验证设计结果,进行确定性优化,将确定性优化结果与稳健设计结果进行对比,同时分析质量特性的平均值、标准差以及正态分布情况,结果表明本文提出面向关键结构参数的复杂多质量特性稳健设计方法的正确性和有效性。
[Abstract]:Robust design of complex multi-mass characteristics is a difficult problem in the field of engineering optimization, and the robust design problem of complex multi-mass characteristics of key structural parameters is more complex. To solve this problem, we need to complete three tasks: extracting the key structural parameters, mining out the relationship between the complex multi-quality characteristics and the key structural parameters. The quality characteristics are guaranteed to reach a high robust level under the condition of the best key structural parameters. On the basis of the traditional engineering optimization, considering the external factors, the complex multi-quality characteristics are fluctuated, and the robust design concept is integrated into the optimization process. The robust design of complex multi-mass characteristics is developed and the optimal key structural parameters are obtained. Taking the fretting platform as an example, the robust design of complex multi-quality characteristics for key structural parameters is proposed. The main research contents are as follows: firstly, the extraction of key structural parameters of fretting platform is carried out. The parameters are determined and the fretting platform model is constructed, and the motion principle of the fretting platform is analyzed. In order to obtain the experimental data needed in the design process, the finite element simulation technique was used to simulate the fretting platform. All the structural parameters of the fretting platform are designed by orthogonal test and the test scheme is made. The fretting platform has the characteristics of large quantity and complex quality characteristics. The grey correlation method is used to analyze the relationship between the quality characteristics and structural parameters of the fretting platform. The key structural parameters are determined according to the degree of correlation between the structural parameters and the quality characteristics, and the extraction of the key structural parameters of the fretting platform is completed. Secondly, on the basis of extracting the key structural parameters of the fretting platform, a mathematical model of complex multi-quality characteristics and key structural parameters is constructed. The Latin hypercube experimental design method is used to obtain the test scheme, and the numerical value of the complex and multi-mass characteristics of the test scheme is obtained by simulation. According to the idea of constructing radial basis function agent model, the mathematical model relationship between complex multi-mass characteristics and key structural parameters of fretting platform is established. In order to verify the accuracy of the model, the relative mean absolute error, root mean square error and relative maximum absolute error are used to evaluate the accuracy. The results show that the established mathematical model has high accuracy. Can be used for subsequent robust design work. Finally, robust design and optimization of key structural parameters are carried out for complex multi-mass characteristics. Traditional deterministic optimization can improve the comprehensive performance of fretting platform, but it can not solve the volatility problem caused by external factors without considering the uncertain factors such as random variable load and material characteristics. In order to improve the robustness level of the fretting platform's quality characteristics, a 6 蟽 design mathematical model is constructed on the basis of deterministic optimization decision model. The Monte Carlo method is used to sample the key structural parameters. The sampling results are substituted into the 6 蟽 design mathematical model and solved by multi-island genetic algorithm. Finally, the optimization values of key structural parameters and the mean and standard deviation of each quality characteristic are obtained, and the robust design of complex multi-mass characteristics of fretting platform is completed. In order to verify the design results and carry out deterministic optimization, the deterministic optimization results are compared with the robust design results. At the same time, the average value, standard deviation and normal distribution of the quality characteristics are analyzed. The results show that the robust design method for complex multi-mass characteristics is correct and effective for key structural parameters.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F273.2
[Abstract]:Robust design of complex multi-mass characteristics is a difficult problem in the field of engineering optimization, and the robust design problem of complex multi-mass characteristics of key structural parameters is more complex. To solve this problem, we need to complete three tasks: extracting the key structural parameters, mining out the relationship between the complex multi-quality characteristics and the key structural parameters. The quality characteristics are guaranteed to reach a high robust level under the condition of the best key structural parameters. On the basis of the traditional engineering optimization, considering the external factors, the complex multi-quality characteristics are fluctuated, and the robust design concept is integrated into the optimization process. The robust design of complex multi-mass characteristics is developed and the optimal key structural parameters are obtained. Taking the fretting platform as an example, the robust design of complex multi-quality characteristics for key structural parameters is proposed. The main research contents are as follows: firstly, the extraction of key structural parameters of fretting platform is carried out. The parameters are determined and the fretting platform model is constructed, and the motion principle of the fretting platform is analyzed. In order to obtain the experimental data needed in the design process, the finite element simulation technique was used to simulate the fretting platform. All the structural parameters of the fretting platform are designed by orthogonal test and the test scheme is made. The fretting platform has the characteristics of large quantity and complex quality characteristics. The grey correlation method is used to analyze the relationship between the quality characteristics and structural parameters of the fretting platform. The key structural parameters are determined according to the degree of correlation between the structural parameters and the quality characteristics, and the extraction of the key structural parameters of the fretting platform is completed. Secondly, on the basis of extracting the key structural parameters of the fretting platform, a mathematical model of complex multi-quality characteristics and key structural parameters is constructed. The Latin hypercube experimental design method is used to obtain the test scheme, and the numerical value of the complex and multi-mass characteristics of the test scheme is obtained by simulation. According to the idea of constructing radial basis function agent model, the mathematical model relationship between complex multi-mass characteristics and key structural parameters of fretting platform is established. In order to verify the accuracy of the model, the relative mean absolute error, root mean square error and relative maximum absolute error are used to evaluate the accuracy. The results show that the established mathematical model has high accuracy. Can be used for subsequent robust design work. Finally, robust design and optimization of key structural parameters are carried out for complex multi-mass characteristics. Traditional deterministic optimization can improve the comprehensive performance of fretting platform, but it can not solve the volatility problem caused by external factors without considering the uncertain factors such as random variable load and material characteristics. In order to improve the robustness level of the fretting platform's quality characteristics, a 6 蟽 design mathematical model is constructed on the basis of deterministic optimization decision model. The Monte Carlo method is used to sample the key structural parameters. The sampling results are substituted into the 6 蟽 design mathematical model and solved by multi-island genetic algorithm. Finally, the optimization values of key structural parameters and the mean and standard deviation of each quality characteristic are obtained, and the robust design of complex multi-mass characteristics of fretting platform is completed. In order to verify the design results and carry out deterministic optimization, the deterministic optimization results are compared with the robust design results. At the same time, the average value, standard deviation and normal distribution of the quality characteristics are analyzed. The results show that the robust design method for complex multi-mass characteristics is correct and effective for key structural parameters.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F273.2
【参考文献】
相关期刊论文 前10条
1 黄焕军;张博文;吴光强;李凡;;基于组合代理模型的车身多学科设计优化[J];汽车工程;2016年09期
2 侯雨伸;王秀丽;张s，
本文编号：2353020
本文链接：https://www.wllwen.com/jingjilunwen/xmjj/2353020.html
