1、 ISO 2014 Applications of statistical and related methods to new technology and product development process Robust parameter design (RPD) Application de mthodologies statistiques et connexes pour le dveloppement de nouvelles technologies et de nouveaux produits Modle paramtrique robuste INTERNATIONA
2、L STANDARD ISO 16336 First edition 2014-07-01 Reference number ISO 16336:2014(E) ISO 16336:2014(E)ii ISO 2014 All rights reserved COPYRIGHT PROTECTED DOCUMENT ISO 2014 All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized otherwise in any form or
3、by any means, electronic or mechanical, including photocopying, or posting on the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address below or ISOs member body in the country of the requester. ISO copyright office Case postale 56 CH-1
4、211 Geneva 20 Tel. + 41 22 749 01 11 Fax + 41 22 749 09 47 E-mail copyrightiso.org Web www.iso.org Published in Switzerland ISO 16336:2014(E) ISO 2014 All rights reserved iii Contents Page Foreword iv Introduction v 1 Scope . 1 2 Normative references 1 3 T erms and definitions and s ymbols 1 3.1 Ter
5、m and definitions 1 3.2 Symbols . 3 4 Robust parameter design Overview 4 4.1 Requirements . 4 4.2 Assessing the robustness of a system . 4 4.3 Robustness assessment through SN ratio . 6 4.4 An efficient method for assessing technical ideas Parameter design . 7 4.5 Two-step optimization (Strategy of
6、parameter design) 8 4.6 Determination of the optimum design 10 5 Assessment of robustness by SN ratio .10 5.1 Concepts of SN ratio 10 5.2 Types of SN ratio 11 5.3 Procedure of the quantification of robustness .11 5.4 Formulation of SN ratio: Calculation using decomposition of total sum of squares 13
7、 5.5 Some topics of SN ratio .19 6 Procedure of a parameter design experiment 20 6.1 General 20 6.2 (Step 1) Clarify the systems ideal function .20 6.3 (Step 2) Select a signal factor and its range 21 6.4 (Step 3) Select measurement method of output response.21 6.5 (Step 4) Develop noise strategy an
8、d select noise factors and their levels 21 6.6 (Step 5) Select control factors and their levels from design parameters 22 6.7 (Step 6) Assign experimental factors to inner or outer array .22 6.8 (Step 7) Conduct experiment and collect data .23 6.9 (Step 8) Calculate SN ratio, , and sensitivity, S .
9、23 6.10 (Step 9) Generate factorial effect diagrams on SN ratio and sensitivity .26 6.11 (Step 10) Select the optimum condition 28 6.12 (Step 11) Estimate the improvement in robustness by the gain 28 6.13 (Step 12) Conduct a confirmation experiment and check the gain and “reproducibility” .29 7 Case
10、 study Parameter design of a lamp cooling system 30 Annex A (informative) Comparison of a systems robustness using SN ratio 40 Annex B (informative) Case studies and SN r atio in v arious t echnical fields 47 Bibliography .72 ISO 16336:2014(E) Foreword ISO (the International Organization for Standar
11、dization) is a worldwide federation of national standards bodies (ISO member bodies). The work of preparing International Standards is normally carried out through ISO technical committees. Each member body interested in a subject for which a technical committee has been established has the right to
12、 be represented on that committee. International organizations, governmental and non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization. The procedures use
13、d to develop this document and those intended for its further maintenance are described in the ISO/IEC Directives, Part 1. In particular the different approval criteria needed for the different types of ISO documents should be noted. This document was drafted in accordance with the editorial rules o
14、f the ISO/IEC Directives, Part 2 (see www.iso.org/directives). Attention is drawn to the possibility that some of the elements of this document may be the subject of patent rights. ISO shall not be held responsible for identifying any or all such patent rights. Details of any patent rights identifie
15、d during the development of the document will be in the Introduction and/or on the ISO list of patent declarations received (see www.iso.org/patents). Any trade name used in this document is information given for the convenience of users and does not constitute an endorsement. For an explanation on
16、the meaning of ISO specific terms and expressions related to conformity assessment, as well as information about ISOs adherence to the WTO principles in the Technical Barriers to Trade (TBT) see the following URL: Foreword - Supplementary information The committee responsible for this document is IS
17、O/TC 69, Applications of statistical methods, Subcommittee SC 8, Application of statistical and related methodology for new technology and product development.iv ISO 2014 All rights reserved ISO 16336:2014(E) Introduction Robust parameter design, also called parameter design, can be applied in produ
18、ct design stage to identify the optimum nominal values of design parameters based on the assessment of robustness of its function. Robustness assessment is performed as a consideration of overall loss during the products life cycle. The overall loss is composed of costs and losses at each stage of t
19、he products life. It includes all the costs incurred during not only its production stage, but also its disposal stages. When a product is not robust, the product causes many environmental and social economic losses (including losses to the manufacturer and the users) due to its poor quality caused
20、by functional variability throughout its usable lifetime from shipping to final disposal. Product suppliers have responsibilities and obligations to supply robust products to the market to avert losses and damages resulting from defects in the products. The aim of applying parameter design in produc
21、t design is to prevent defects, failures, and quality problems that can occur during the usage of the product. A robust product, an output of parameter design, is a product which is designed in such a way as to minimize users quality losses caused by defects, failures, and quality problems. Note tha
22、t defects, failures, and quality problems are caused by functional variability of a non-robust product. In parameter design, optimum nominal values of a products design parameters can be selected by treating a products design parameters as control factors and by assessing robustness under noise fact
23、ors. The use of parameter design at development and design stages makes it possible to determine the optimum product design and specification so that the product is robust in the market. At manufacturing stage, the product suppliers manufacture their products that meet the product specifications. On
24、e can optimize manufacturing processes to produce the products that meet the specifications. However, robustness against customers environment and products aging can be addressed only by product design. Robust parameter design methodology provides effective methods for achieving robustness through i
25、ts design of specification determination, and it is a preventive countermeasure against various losses in the market. In practice, many products defects and failures occur due to the products response that deviates from or varies around the designed target values by the change in usage environment a
26、nd deterioration, i.e. noise conditions. The variability of products response due to noises can be used as a measure of robustness, because market losses increase in proportion to the magnitude of variability of products response. SN ratio, corresponding to the inverse of the variability measure, is
27、 used as a measure of goodness in robustness. In other words, the higher the SN ratio is, the less the market losses are. For the experimental plan of parameter design, direct product of inner array and outer arrays is proposed. Control factors are assigned to the inner array, and signal and noise f
28、actors are assigned to the outer array. By using a direct product plan, all the first level interactions between control factors and noise factors can be assessed and can be utilized to select the optimum level of control factors from the point of view of robustness. Assessing robustness through SN
29、ratio is a key of parameter design. The outer array is for evaluating SN ratio, robustness, for each combination of levels of control factors indicated by the inner array. The inner array is for comparing SN ratios and selecting the optimum combination of systems design parameters. As for the inner
30、array, an orthogonal array L 18 , is recommended as an efficient plan, and then only the applications of an orthogonal array L 18are discussed in this International Standard. Applications of experimental layout other than orthogonal array L 18can be found in the examples in references in the Bibliog
31、raphy. More detailed discussions on inner array and orthogonal arrays can be found in the references. Robust parameter design (RPD), and thus this International Standard, is directly targeted at the losses incurred at the usage stage. Where possible, losses at other stages are also investigated so t
32、hat the results of parameter design can be applied to perform the optimum product design for the whole stages of the products life cycle. ISO 2014 All rights reserved v Applications of statistical and related methods to new technology and product development process Robust parameter design (RPD) 1 S
33、cope This International Standard gives guidelines for applying the optimization method of robust parameter design, also called as parameter design, an effective methodology for optimization based on Taguchi Methods, to achieve robust products. This International Standard prescribes signal-to-noise r
34、atio (hereafter SN ratio) as a measure of robustness, and the procedures of parameter design to design robust products utilizing this measure. The word “robust” in this International Standard means minimized variability of products function under various noise conditions, that is, insensitivity of t
35、he products function to the changes in the levels of noises. For robust products, their responses are sensitive to signal and insensitive to noises. The approach of this International Standard can be applied to any products that are designed and manufactured, including machines, chemical products, e
36、lectronics, foods, consumer goods, software, new materials, and services. Manufacturing technologies are also regarded as products that are used by manufacturing processes. 2 Normative references The following documents, in whole or in part, are normatively referenced in this document and are indisp
37、ensable for its application. For dated references, only the edition cited applies. For undated references, the latest edition of the referenced document (including any amendments) applies. ISO 3534-1, Statistics Vocabulary and symbols Part 1: General statistical terms and terms used in probability I
38、SO 3534-3, Statistics Vocabulary and symbols Part 3: Design of experiments 3 T erms an d definiti ons and s ymbols 3.1 T erm and definitio ns For the purposes of this document, the terms and definitions given in ISO 3534-1 and ISO 3534-3, and the following apply. 3.1.1 function work which a system p
39、erforms in order to fulfil its objective Note 1 to entry: A function can be expressed by the mathematical form of input-output relation. 3.1.2 robustness degree of smallness in variability of a systems function under various noise conditions Note 1 to entry: Systems performance can be assessed by ro
40、bustness. SN ratio is a quantitative measure of robustness. INTERNATIONAL ST ANDARD ISO 16336:2014(E) ISO 2014 All rights reserved 1 ISO 16336:2014(E) 3.1.3 signal-to-noise ratio SN ratio ratio of useful effects to harmful effects in response variations Note 1 to entry: SN ratio is usually expressed
41、 in db value. The notation of db is used instead of dB for SN ratios of robustness measurements. Note 2 to entry: The anti-logarithm value of an SN ratio, real number, is the inverse of a variation measure such as a variance or a coefficient of variation, and inversely proportional to monetary loss.
42、 Note 3 to entry: The change in response caused by intentional change of input signal value is a useful effect. In case of the ideal function being zero point proportional, the linear slope forced through the zero point is a useful term. Note 4 to entry: The change in response caused by noise factor
43、s is a harmful effect. Effects of noise factors and deviation from the ideal function are examples. Note 5 to entry: SN ratio should contain the variability under noise factors and the discrepancy from the ideal function under average usage condition. 3.1.4 sensitivity amount of change in response c
44、aused by unit change of input Note 1 to entry: Sensitivity is usually expressed in db value. Note 2 to entry: For dynamic characteristic cases, the sensitivity shows the magnitude of linear coefficient due to input signal, 2 , where is a proportional constant. Note 3 to entry: For the nominal-the-be
45、st response, the sensitivity shows the magnitude of mean, m 2 , where m is an average of responses. 3.1.5 noise variable which disturbs a systems function Note 1 to entry: Any variable in the users conditions for operating is either a signal or a noise. Note 2 to entry: Noise is composed of internal
46、 noise and external noise. They are sometimes called as capacity and demand, respectively. Changes of internal constant of the system or its parts over time, such as deterioration, aging. and wear, and manufacturing variations are examples of internal noises. Usage conditions and environment conditi
47、ons of the product are examples of external noises. 3.1.6 signal input variable to the system, which is intentionally changed by the user to get an intended value of response in input-output relation Note 1 to entry: Any variable in the users conditions for operating is either a signal or a noise No
48、te 2 to entry: There are two kinds of signal: active signal and passive signal. Active signal is operated by user to get intended response, for example, rotating angle of a steering wheel to change the vehicles direction. Passive signal is used by user to know the value of input from response readin
49、g, for example, temperature in thermal measurement. In both cases, output will change by changing the value of the signal but the user wants to get response value in the active case, and the user wants to know the value of signal in the passive case. 3.1.7 dynamic characteristics output response which has multiple ideal target values depending on the value of a signal Note 1 to entry: The relation between dynamic characteristics and a signal can be expressed by input-output functional form. The output of a systems f
