[1]杨雅棠,陈富民,何帅,等.叶片生产过程的非线性轮廓控制方法研究[J].西安交通大学学报,2020,54(08):091-98+162.[doi:10.7652/xjtuxb202008012]
 YANG Yatang,CHEN Fumin,HE Shuai,et al.Nonlinear Profile Control Method for Blade Production Process[J].Journal of Xi'an Jiaotong University,2020,54(08):091-98+162.[doi:10.7652/xjtuxb202008012]
点击复制

叶片生产过程的非线性轮廓控制方法研究
分享到:

《西安交通大学学报》[ISSN:0253-987X/CN:61-1069/T]

卷:
54
期数:
2020年第08期
页码:
091-98+162
栏目:
出版日期:
2020-08-10

文章信息/Info

Title:
Nonlinear Profile Control Method for Blade Production Process
文章编号:
0253-987X(2020)08-0091-09
作者:
杨雅棠 陈富民 何帅 李建华
西安交通大学机械制造系统工程国家重点实验室, 710049, 西安
Author(s):
YANG Yatang CHEN Fumin HE Shuai LI Jianhua
State Key Laboratory for Manufacturing Systems Engineering, Xi'an Jiaotong University, Xi'an 710049, China
关键词:
非线性轮廓控制 差异度量指标 自适应粒子群算法 支持向量数据描述
Keywords:
nonlinear profile control difference metric adaptive particle swarm optimization support vector data description
分类号:
TH16
DOI:
10.7652/xjtuxb202008012
文献标志码:
A
摘要:
针对叶身型线轮廓特征参数间相关关系复杂、历史批次数据不足等原因造成的生产过程型面几何轮廓质量控制困难的问题,提出一种叶片生产过程非线性轮廓控制方法。该方法以叶身型线三坐标值为基础数据进行分析,首先构建叶身型线轮廓差异度量指标,并基于测量数据进行指标参数提取及标准化处理,作为非线性轮廓控制对象; 其次以受控状态下的差异度量指标分类准确率为优化目标,通过自适应粒子群算法(APSO)优化高斯核函数及惩罚系数; 然后联合优化参数及受控指标数据训练支持向量数据描述(SVDD)模型,获得超球体半径作为控制限,构建基于超球体核距离的非线性轮廓控制图; 最后计算待测轮廓指标数据点到超球体中心的内核距离,得到控制图标绘点,进而判断叶片生产过程是否异常。仿真结果表明,相比于传统方法,该方法可有效表征型线不同区域质量特征对轮廓形状的综合影响,且对不同均值偏移都具有更强的异常波动检测力,并能有效解决中小批量生产过程控制中数据量不足的问题。
Abstract:
A nonlinear profile control method for blade production process is proposed to deal with the problem that the quality control of the surface contour of production process is difficult due to the complex relationship between the profile parameters of blade body and lack of historical data. According to three-coordinate data of blade profile, the blade body profile difference metrics are constructed, and the index parameter extraction and standardization processing are performed based on the measured data, which are taken as the nonlinear contour control objects. Then the classification accuracy of the difference metrics under controlled state is considered as the optimization goal, and the parameters of Gaussian kernel function are optimized by the adaptive particle swarm optimization. The SVDD model is trained by combining the optimized parameters and controlled index data to obtain the radius of the hypersphere as the control limit, and the nonlinear profile control chart based on the hypersphere kernel-distance is constructed. The distance between the measured contour index data point and the center of the hypersphere is calculated to obtain the statistical point of the control chart, and then it is determine whether the blade production process is abnormal. Compared with the traditional methods, the proposed method can effectively characterize the comprehensive influence of the quality characteristics of different regions of the profile on the contour shape, and has a stronger detection ability for abnormal fluctuation of different mean deviation to solve the difficulty of insufficient data in the process control for small and medium-sized batch production.

参考文献/References:

[1] CHEN Lei, LI Bing, JIANG Zhuangde, et al. Parameter extraction of featured section in turbine blade inspection [C]∥2010 IEEE International Conference on Automation and Logistics. Piscataway, NJ, USA: IEEE, 2010: 501-505.
[2] YANG Xu, LIU Xiaolong, LIU Jiming, et al. Matching algorithm for profile error calculation of blade surface [J]. Advanced Materials Research, 2016, 1136: 630-635.
[3] KHEDMATI M, NIAKI S T A. Monitoring simple linear profiles in multistage processes by a MaxEWMA control chart [J]. Computers & Industrial Engineering, 2016, 98: 125-143.
[4] 何帅, 陈富民, 杨雅棠, 等. 基于点云数据的自由曲面加工误差评定 [J]. 西安交通大学学报, 2019, 53(10): 135-142.
HE Shuai, CHEN Fumin, YANG Yatang, et al. Machining error evaluation of free-form surface based on point cloudy data [J]. Journal of Xi'an Jiaotong University, 2019, 53(10): 135-142.
[5] 余忠华, 吴昭同. 面向小批量制造过程的质量控制方法研究 [J]. 机械工程学报, 2001, 37(8): 60-64.
YU Zhonghua, WU Zhaotong. Algorithm of quality control in small batch manufacturing process [J]. Chinese Journal of Mechanical Engineering, 2001, 37(8): 60-64.
[6] 陈鑫, 陈富民. 面向多品种小批量生产的贝叶斯动态质量控制方法 [J]. 西安交通大学学报, 2019, 53(6): 17-22.
CHEN Xin, CHEN Fumin. Bayesian dynamic quality control method for multi-variety and small batch production [J]. Journal of Xi'an Jiaotong University, 2019, 53(6): 17-22.
[7] VAGHEFI A, TAJBAKHSH S D, NOOROSSANA R. Phase II monitoring of nonlinear profiles [J]. Communications in Statistics-theory and Methods, 2009, 38(11): 1834-1851.
[8] ZHANG Y, HE Z, FANG J T, et al. Nonparametric control scheme for monitoring phase II nonlinear profiles with varied argument values [J]. Chinese Journal of Mechanical Engineering, 2012, 25(3): 587-597.
[9] 张阳. 复杂产品制造过程中轮廓控制方法研究 [D]. 天津: 天津大学, 2013: 45-49.
[10] 唐浩, 屈梁生. 基于支持向量机的发动机故障诊断 [J]. 西安交通大学学报, 2007, 41(9): 1124-1126.
TANG Hao, QU Liangsheng. Fault diagnosis of engine based on support vector machine [J]. Journal of Xi'an Jiaotong University, 2007, 41(9): 1124-1126.
[11] HOSSEINIFARD S Z, ABDOLLAHIAN M, ZEEPHONGSEKUL P. Application of artificial neural networks in linear profile monitoring [J]. Expert Systems with Applications, 2011, 38(5): 4920-4928.
[12] MOGUERZA J M, MUÑOZ A, PSARAKIS S. Monitoring nonlinear profiles using support vector machines [M]∥Lecture Notes in Computer Science: Volume 4756. Berlin, Germany: Springer, 2007: 574-583.
[13] PAPOUTSIDAKIS M G, PIPE A G, CHAMILOTHORIS G E. A supervised learning method of neural networks in a non-linear and time depended control process [C]∥21st Mediterranean Conference on Control and Automation. Piscataway, NJ, USA: IEEE, 2013: 1526-1531.
[14] 虞跨海, 李立州, 岳珠峰. 基于解析及特征造型的涡轮冷却叶片参数化设计 [J]. 推进技术, 2007, 28(6): 637-640, 656.
YU Kuahai, LI Lizhou, YUE Zhufeng. Parametric design for cooling turbine blades based on analytic and feature modeling [J]. Journal of Propulsion Technology, 2007, 28(6): 637-640, 656.
[15] MA Wenqi, LIU Shugui. The research on the methods of the aero engine blades evaluation [J]. Applied Mechanics and Materials, 2013, 333-335(1): 322-326.
[16] 何帅, 陈富民, 李建华, 等. 叶身型线轮廓度评定方法研究 [J]. 西安交通大学学报, 2019, 53(8): 175-182.
HE Shuai, CHEN Fumin, LI Jianhua, et al. Profile evaluation method for blade pattern [J]. Journal of Xi'an Jiaotong University, 2019, 53(8): 175-182.
[17] 贾新章, 游海龙, 顾铠. 统计过程控制理论与实践 [M]. 北京: 电子工业出版社, 2017: 142-153.
[18] TAX D M J, DUIN R P W. Support vector data description [J]. Machine Learning, 2004, 54(1): 45-66.
[19] GANI W, LIMAM M. Performance evaluation of one-class classification-based control charts through an industrial application [J]. Quality and Reliability Engineering International, 2013, 29(6): 841-854.
[20] 武千惠, 黄必清. 基于支持向量数据描述的剩余寿命预测方法 [J]. 计算机集成制造系统, 2018, 24(11): 2725-2733.
WU Qianhui, HUANG Biqing. Remaining useful life prediction based on support vector data description [J]. Computer Integrated Manufacturing Systems, 2018, 24(11): 2725-2733.
[21] LI Mi, CHEN Huan, WANG Xiaodong, et al. An improved particle swarm optimization algorithm with adaptive inertia weights [J]. International Journal of Information Technology and Decision Making, 2019, 18(3): 833-866.
[22] FARAZ A, HEUCHENNE C, SANIGA E. Optimal T2Control chart with a double sampling scheme-an alternative to the MEWMA chart [J]. Quality and Reliability Engineering International, 2012, 28(7): 751-760.
[23] 郑辉, 张伟, 张峻霞, 等. 多元动态控制图的性能对比分析与选择 [J]. 计算机集成制造系统, 2017, 23(11): 2335-2343.
ZHENG Hui, ZHANG Wei, ZHANG Junxia, et al. Performance comparison and selection of multi-variate adaptive control chart [J]. Computer Integrated Manufacturing Systems, 2017, 23(11): 2335-2343.

备注/Memo

备注/Memo:
收稿日期: 2019-11-25。作者简介: 杨雅棠(1995—),女,硕士生; 陈富民(通信作者),男,副教授。
更新日期/Last Update: 2020-08-10