制造分析進程數(shù)據(jù)使用快速標記技術(shù)

1、煞樟釹寢崢蛔榆苕殼牝蟆多附錄1 翻譯原文及譯文烷澈胰放呼鷗冕忡茶荇諉滟簦墳倆虬雕弘恕躒躚壩莠貌微捎舂芬溧袍糜映噯氨累喱駕嬌泛且攜敕噥螄魷廒監(jiān)舳醣劭坌躑賺誒鑼喵胳離觸褸酥胝簽羿章邗蔌聊諮鬟瘤陌謖獅諤優(yōu)嚷埝堅郯踺猛櫻敵此遘鋼備樗生土蔭憎唯婁焯館狳醐咬陬罐潤沼悸寄蘗促蝕拷殊潸頌磧替鮑道娼逶戾Doc No: P0193-GP-01-1送茆蓐瑩釵嘲榨某候蕊堆廿Doc Name: Analysis of Manufacturing苒踵畛芪甌證獠酷瓜女酵萍Process Data Using嶝漭需艤逑崾耨犯甕典磚撕 QUICK TechnologyTM映止貌迤臌啵喝窀嗌蘋怯樞爺?shù)锓柒蒗主告妮涆庾鬓＿八判栏?/p>

2、嗍拋屐侯眉閾洹鄉(xiāng)空簍汀烀堊腧炻錙櫝惲裴溻Issue：1蔚娛怠縛浦蛆鼎謚駿膠柯傲Data：20 April ，2006囁崆易嬰碣墻笤怔鵬錙涉頎緗諸席濯鐒囟尾苊紀檬箏粘啖嚎罹篩傖柴扦攘輪盧嘟奏珊腴燠譬蘩版擴茶鰣曲鄂蛞奴騷韞氐沔療婦蹌款錠跗鄣嚳碰煸肯捭鬣艄陣森糯楞梧菹膚銬鰷選貘強縲乾陜狹擂Name(Print)材搶賒咀射姍諜暄豇鰾托爬Signature早螞鰈軍悒橢鹽學魃講廂迎Author：鍆卻飴故摜曜管乳煮鞅狩埸D.Clifton理釓廝帑準蝎撥廝袒變鷦奸登螟婷艫梵溆頜雅籩謖炊侏Reviewer：工遇崦厄鶉掭拯潿敏碉佴慰S.Turner季輩殼捩獰鞭媳忄魷才蘊融肆腴澈恭尉氅頜沼悌傷绱咣駝僻椴蹌肛旦穹氵舷

3、冰涫牧Table of Contents柔簪茭懷鯡袍鱒蕹尻孩譏椅1Executive Summary嘭豳孺錁暮蛆娶贄致婊槁縊6疤蛙香宿蠅漳昴垡藏倘纂岔61.1Introduction閭徘藩矢嬉毅苫狷輞戡峴沅61.2Techniques Employed滋鮚宕瀉僑貂叢拚帝蹴徑汴61.3Summary of Results粼吱管蒿帽特相矚崴姹稹斥61.4Observations峻膈宛鵲醍卡幃蠲櫞米陌亥7灌穢評刻汽淙鋱柃駱漕菥懇72Introduction丈吻氏錄绔髻瞧擯城帙淺局82.1Oxford BioSignals Limited織兀潁嗖鋦竟板抱壕狀茯猴8芊藿薊諦礬粥族嚌粱鷚啡楱83Exter

4、nal References碘膣蠟獯膩醅改慮砰品倌潢9火緲磐魍瘕徹辱白矩粒蠣垓94Glossary盯霖驛衲呱蜃郟抻晷尿偃耽95 Data Description搓榜版嫉觚拼亥咝脒危漱躚94.1Data types扈孱窺幟鮮螢斗菇驄籍揩鉦9霸蕩番溟幄孫謄樸殃頦魷鰩104.2Prior Experiment Knowledge邁畫檢杈糝窶游蕕烽蜮塄料104.3Test Description嶧硐燕僉裎尖屨縛捆隧窄嗣10朋禾鄴抉魁艮齪僨凡甘恣渚116.Pre-processing訃岡霉夾篁確緗優(yōu)腭凇晡磷114.4Removal of Start/Stop Transients壤戢播芄揮搬諑報咨棺憧繽

5、114.5Removal of Power Supply Signal芑鋤圄又裳盾果切嚏膨蛞壚114.6Frequency Transformation薤憨鬏卸壽夫桫蟒欽郄嫻霉11磁鼎鐋踹銫淶丕矍罐鈕韭啃13吮糖內(nèi)鈹?shù)抗∝淂栺~灌莠爍14Analysis I-Visualisation孌橋琮線壁誰霜紋冉逶詮蒔144.7Visualisation of High-Dimensional Data佳苔葦羨斌贅器日萃咬汊隼14The use of large numbers of measured variables introduces problems in the visualization o

6、f the resulting data. A collection of temperatures, pressures, etc. forms a high-dimensional representation of the state of a system, but this is not readily interpreted by an operator. 萆瘀篼偏疑鞒倮蹦況蓑尬噴14Neuroscale allows the visualization of systems that have high-dimensionality by mapping data to lowe

7、r numbers of dimensions(typically two,for visual inspection). It attempts to preserve the inter-pattern distances in the high-dimensional data. Data which are close together in high-dimensional space are typically kept close together in 2-D space, and data that are originally far apart remain well s

8、eparated after projection.罡嫩字元跋崽鏹危遴咦嘣14The projection is performed using a non-linear function from the datas k dimensional space down to 2-D for visualization purposes. In this investigation, k is 5:Ax, Ay, Az, AE, SP are the high-dimensional sample vectors. 堤德椒斜夼挺盎征琳念婢蠱14The creation of a non-line

9、ar mapping from 5-D space to 2-D requires sample data from across the range of tests. In order to reduce the large number of available sample data to a quantity suitable for constructing the mapping, a summary of the data-set is required. Each test was summarized by a number of prototype 5-D vectors

10、 using the k-means clustering algorithm(in which a large number of data are represented by a smaller number of prototype vectors). The non-linear mapping was trained using the prototype 5-D vectors from all tests.爆氳锪蟯畛罹巽稼癮螫脾蝴14To allow the examination of the 5-D data using visualization, it is conve

11、nient to divide the drilling process in to three stages, corresponding to the typical behaviour of the process described in Section 5.3.睞徂稷嬡砸筱僉翅賭咄河濺14A heuristic algorithm was produced to perform automatic segmentation into three episodes using the SP channel, as illustrated in Figure 6(which shows

12、a low-pass filtered version of SP superimposed on the original signal as a red line). The three states identified correspond to :猗恝鷙舵迷攥嵌蛤聘似談霆14Note that this segmentation is only the identification of the times of onset and offset of each of the three described states, for the purposes of graphical

13、display as described in the next sub-section.蕙縟淋民硪隔獪罹逋舔捶碘14孛橘捎號諗翕幅譫唱窗熹佰Table of Figures 漆瀧卩去踹唆闖檣轔睨煞舟Figure 1- Test 90. From top to bottom: Ax, Ay, Az, AE, SP against time t(s)倮歿噶迓灑襯夕解醪嗎嗒笸Figure 2- Power spectra for Test 19 after removal of 50Hz power supply contribution. The top plot shows a 3-D “l(fā)a

14、ndspace” plot of each spectrum. The bottom plot shows a “contour” plot of the same information, with increasing signal power shown as increasing colour from black to red 岢弘伊詠鉻樁量獺夯蕊鯨崾Figure 3- Power spectra for Test 19 after removal of all spectral components beneath power threshold役粢嚴彐褰翥大戢拋歪雞考Figure

15、 4- Az against time (in seconds) for Test 19,before removal of low-power frequency components 拘腸強汀鏍考酉肋片所氤醌Figure 5- Az against time (in seconds) for Test 19, after removal of low-power frequency components榭泫膨顰犭美瘺嫻馗聒頦翟Figure 6- SP for an example test, showing three automatically-detecrmined states:S1

16、-drilling in (shown in green); S2-drill-bit break-through and removal (shown in red); S3-retraction (shown in blue)頸惕妒紡櫪待店庹茈防納矩Figure 7- Example signature of variable plotted against operating-point悱濟繾確煜吃攛哐嗦融唳怒Figure 8- Power spectra for test 51, frequency (Hz) on the x-axis between 0 fs/2勁掃岣挖桅嵫悝敝疲殺

17、髑暑Figure 9- Average significant frequency 蕺鍛浙狩漂額妒淠冠璋澌粳F(xiàn)igure 10- Visualisation of AE signatures for all tests跤共我昴滬邶唧孫郄莓字企Figure 11- Visualisation of Ax broadband signatures for all tests鍍洲琮荻葙逖澗胬脹宗貪烊Figure 12- Visualisation of Ax average-frequency signatures for all tests炳噢竭箅褐癆咐氕斕妖潭芴Figure 13- Novelt

18、y detection using a template signature迅菩挫誹呲蘿俊十闋兀淦盍Figure 14- 畚受滋護娓榻詞夙油瀾猓票1 Executive Summary嘭豳孺錁暮蛆娶贄致婊槁縊疤蛙香宿蠅漳昴垡藏倘纂岔1.1 Introduction閭徘藩矢嬉毅苫狷輞戡峴沅The purpose of this investigation conducted by Oxford BioSignals was to examine and determine the suitability of its techniques in analyzing data from an ex

19、ample manufacturing process. This report has been submitted to Rolls-Royce for the expressed of assessing Oxford BioSignals techniques with respect to monitoring the example process. 忖褪婆樸槨翰糕澌棺丸界蓐The analysis conducted by Oxford BioSignals (OBS) was limited to a fixed timescale固定的時間標度( 時間間隔)改悖涔擲俗支輝沮穢

20、碡嬡趲, a fixed set of challenge data for a single process (as provided by Rolls-Royce and Aachen university of Technology), with no prior domain knowledge 相關(guān)先驗知識述煮仃堇庫俚髡蹭墨洮逯蒗, nor information of system failure系統(tǒng)崩潰的征兆半皤銀可淖削鵲悅礻竽颮茂 .穢妣癜沾碾蠻滯臠郛蕩兼燼1.2 Techniques Employed引用的技術(shù)悉裰黜駛抨濂袋囤丬境尢蕓滋鮚宕瀉僑貂叢拚帝蹴徑汴OBS used

21、a number of analysis techniques given the limited timescales:輥洼仆耶削崩痞譴琿蕕榆踵I-Visualisation, and Cluster Analysis聚類分析菜裱頦蹌鉦圳俗歟熠鈾擼岸 榀踹盎蒙顢戥靂痄球槐吧怪This powerful method allowed the evolution of the system state (fusing all available data types提煉所有的現(xiàn)有數(shù)據(jù)類型副畏陸匱閎揭朕蹭縈燥嗯毿) to be visualised throughout the series of

22、 tests. This showed several distinct modes of operation幾個截然不同的運作模式箅矽潦萇珞西堀吭蒙握評丁 during the series, highlighting major events observed within the data, later correlated with actual changes to the systems operation by domain experts.默暮機撓荏洮鲞叟閃敏菝簽Cluster analysis automatically detects which of these even

23、ts may be considered to be “abnormal”, with respect to previously observed system behavior 聚類分析能夠綜合考慮系統(tǒng)的特性習慣，自動從這些系統(tǒng)事件中識別出異常事件鶴歆蓮對狄晌驥彰汀樣鞠陀.趴黯策聶塬棱肴災葉找彡墀II-Signature represents信號再現(xiàn)鲇亠鵲髏稈綁顓報嗓盞拐燾 each test as a single point on a plot, allowing changes between tests to be easily identified. Abnormal tests

24、 are shown as outlying points, with normal tests forming a cluster每一次實驗將一一個單獨的點的形式出現(xiàn)在圖表中，允許個實驗點之間存在一定的誤差，超出誤差范圍的實驗點就可以很容易的識別出來。在正常的一系列實驗點圖表中，異常實驗點將被高亮顯示出來。柳貓早螂苑鈄頜鱈甬壁戎賾.傳薰忱縞埯抬欲狄院悶獻太Modeling the normal behavior of several features selected from the provided data分析已經(jīng)獲取的數(shù)據(jù)的各種特性，并依據(jù)這些行為特性建立數(shù)學模型攻被莧扶睽才封茜娓衲

25、繇撩, this method showed that advance warning of system failure could be automatically detected using these features, as well as highlighting significant events within the life of the system.該方法證明：根據(jù)這些行為特性，加工失敗的預警系統(tǒng)可以自動針對檢測加工狀態(tài)判斷是否報警，并高亮提醒重要事件。拭瘋盹倏蠕訟目嶁聆萘泠棗裴鷹取狗錘驍莎慘孔嶙臘疚III-Template Analysis模板分析壢猊埤蕎姜柑模休褡

26、礬北媸 史野擂唆祿囪映撫舴卉愣慍This method allows instantaneous sample-by sample novelty detection, suitable for on-line implementation.模板分析法可以進行逐點瞬時采樣分析，這種方法適合于在線分析。虜裔索愈纖挲訊裾迪襖尖踴貿(mào)墑醉鄴見架祿虛耿瘓捂唼Using a complementary approach to Signature Analysis, this method also models normal system behavior. Results confirmed the o

27、bservation made using previous methods.嵊渥診迥軼它這蕪淀婦鬮浣IV-Neural network Predictor 葫燠嗾哚楠圍登酌乞庥癖紊Similarly useful for on-line analysis, this method uses an automated predictor of system behaviour(a neural network predictor), in which previously identified events were confirmed, and further significant epi

28、sodes were detected.頰偌埡緇笠條喘悅朵粉士討鱒婧者兔騰獎卡憲簧囪榧贛1.3 Summary of Results粼吱管蒿帽特相矚崴姹稹斥Early warning of system failure was independently identified by the various analysis methods employed. 軍居殍紅佛親羔觸佾戔鰒泊Several significant events during the life of the process were correlated with actual known events later re

29、vealed by system experts.姝某銚蛩舉荔試袖練皓齄颮Changes in sensor configurations are identified, and periods of system stability (in which tests are similar to one another) are highlighted.害蝕睢鷹埝灄燴墓繞鱗哼皤This report shall be used as the basis for further correlation of detected events against actual occurrences w

30、ithin the life of the system, to be performed by Aachen University of Technology.鱈豢財滿煒迫掂喃壹虞踉腠1.4 Observations峻膈宛鵲醍卡幃蠲櫞米陌亥Based on this limited study, OBS are confident that their techniques are applicable to condition monitoring of the example manufacturing process as follows:爨吹曰聰撈逍揶算裝刮瀵錄Evidence sh

31、ows that automated detection of system novelty is possible, compared to its “normal” operation.贓磊閉樅送摔啦溧淋腆張厚Early warning of system distress may be provided, giving adequate time to take preventative maintenance actions such that system failure may be avoided.諢恍葦蝕退諂戴刀蜜候葦蔦Provision “fleet-wide” analys

32、is is possible using the techniques considered within this investigation.棘晡痂訂褙付燙夷渝忉囂氧The involvement of domain knowledge from system experts alongside OBS engineers will be crucial in developing future implementations. While this “blind” analysis showed that OBS modelling techniques are appropriate

33、for process monitoring, it is the coupling of domain knowledge with OBS modelling techniques that may provide optimal diagnostic and prognostic analysis.攙旦鈽繪緙蠹鍪斬薤狼燧錒灌穢評刻汽淙鋱柃駱漕菥懇2 Introduction丈吻氏錄绔髻瞧擯城帙淺局2.1 Oxford BioSignals Limited織兀潁嗖鋦竟板抱壕狀茯猴This document reports on the initial analysis conducted

34、by Oxford BioSignals of manufacturing process challenge data provided by Rolls-Royce, in conjunction with Aachen University of Technology(AUT).普哉躉娘裒鰱討?zhàn)T椴胤婿滁Oxford BioSignals Limited(OBS) is a world-class provider of Acquisition, Data Fusion, Neural Networks and other Advanced Signal Processing techni

35、ques and solutions branded under the collective name QUICK Technology. This technology not only provides for health and quality assurance monitoring of the operational performance of equipment and plant.圖蚯鄧誦靖胼笱縹卦搟哨訌QUICK Technology has been extensively proven in the field of gas turbine monitoring w

36、ith both on-line and off-line implementations at multiple levels: as a research tool, a test bed system, a ground support tool, an on-board monitoring system, an off-line analysis tool and a “fleet” manager.埯爻崍議喪躍訊裉戡跳瀋富Many of the techniques employed by OBS may be described as novelty detection meth

37、ods. This approach has a significant advantage over many traditional classification techniques in that it is not necessary to provide fault data to the system during development. Instead, providing a sufficiently comprehensive model of the condition can be identified automatically. As information is

38、 discovered regarding the causes of these deviations it is then possible to move from novelty detection to diagnosis, but the ability to identify previously unseen abnormalities is retained at all stages.沒蜿隆詐湫劭淑覓汰財律鎳芊藿薊諦礬粥族嚌粱鷚啡楱3 External References碘膣蠟獯膩醅改慮砰品倌潢Accompanying documentation providing fu

39、rther information on the data sets is available in unnumbered documents.膪繃煸巰忒澮杉汽舳柘瘳俾火緲磐魍瘕徹辱白矩粒蠣垓4 Glossary盯霖驛衲呱蜃郟抻晷尿偃耽AUT- Aachen University of Technology 藝宰崇迓褙鯫即咯銃傘蚊菘GMM- Gaussian Mixture Model 杵幫蘼眠鲴覡冶魃哇銹訓滟MLP- Multi-Layer Perception袋菲黍戍賦洲隘步術(shù)糕誄恒OBS- Oxford BioSignals Ltd.盲帆儻磋雹集崽利饌螅榭夂5 Data Descript

40、ion搓榜版嫉觚拼亥咝脒危漱躚The following sections give a brief overview of the data set obtained by visual inspection of the data. 槲儡杠靶圈渤溥起士暉溆藎4.1 Data types扈孱窺幟鮮螢斗菇驄籍揩鉦The data provided were recorded over a number of tests. Each test consisted of a similar procedure, in which an automated drill unit moved towa

41、rds a static metallic disk at a fixed velocity (“feed”), a hole was drilled in the disk at that same feed-rate.學雍坳沿裕艚蓀饋浮氙甾蹺The following data streams were recorded during each test, each sampled at a rate of 20 KHz:惺帝擒硝丁否卓鳊蹈疵課趟 Ax acceleration of the disk-mounting unit in the x-plane1 , 榨愀奮浦暌胛倪案綻后狽迪

42、 Ay- acceleration of the disk-mounting unit in the y-plane1 ,斐翕黟嗎混逶瞰抗袂鱗截攜 Az- acceleration of the disk-mounting unit in the z-plane1 ,癱狼媧抒囪刷塏士密很砹擄 AE-RMS acoustic emission, 50-400 KHz2,鮮梳詰莨藶邃醚榆擬踮庚赳 SP-power delivered to the drill spindle3.那還鞒杖宸喚螢口婭凸賜絞Tests considered in this investigation used three

43、 drill-prices (of identical product specification) as shown in Table 1.等婪鮭秸珈惕釕沿洱岑懋怯Table 1-Experiment Parameters by Test末諏遲躚組褪筲仲喧即漂蚊Drill Number卻梢晶媼檐暾涔副呷眠賄倍Test Numbers饗黌必怙濁科壘回焦臼胙魔Drill Rotation Rate煒紛瓢笄悻監(jiān)我踴撓雞薏蚪Feed Rate硤痤瘭讓蒯較妝劂呻坤托炳1兩龐揠釜斯俎覺哦籜顫薛肴12謚蚍蔥弧樾跬僮縲薟墟埭粹1700RPM绱毖尥搟裴鰩慕縣臭菔萜矛80 mm/min春蕃合菜傣豐喊鄖鼠乒孳樽2

44、尉僻椎啼拇清僧奎酢駛麝邛3127衛(wèi)瀝濱備姥跛閫腮曬彳冤勢1700RPM浜坦桶最須櫻窿下靖鉑槊鵠80 mm/min莛鏌害饒獨粱柿誠料濾袖朋3鶉顙獨鋸裳笛踢麈睚猛鯛熬冬涿競噌岑斡蜿肪倥粗凱笤1700RPM箅?yún)闻箨飞觅邤P郫戕緘雹升120mm/min忤怫澤蚰羰色姆潭詰撬巳鐘Note that tests 16,54,128,129 were not provided, thus a series of 190 tests are analysed in this investigation. These 190 tests are labeled as shown in Table 2.碲茗廾焐叛槽鍍

45、仂淥桷啕噪Table 2 Test indices used in this report against actual test numbers邃晴旄淳敏怒揍隅劐航軍吉Test Indices繁適柚延吹羌茅霄篼竄嗶皿Actual Test Number釣硫侄夏羅剞汛宰東戶將轆115酒初張奉鑰嫌檬跡妲蛙汪檢115嘜瑣婁翎弟溜狺賄極劈鴆揉1652垂看沒嫌堯鎦酚剔謫潰襝1753沱绔殺紋包芫矽氬腸徨每芻53125漚隼避梁泡鈺鯖閃脎拚卿秋55127比哆恣罡斜蝰嵋繯蹭衽薪恬綴郭閃碡謦鮑潮曜嗆哏覲吁郎暴氚胰縑珩憲瀲羧叱謝碚霸蕩番溟幄孫謄樸殃頦魷鰩4.2 Prior Experiment Knowledge

46、邁畫檢杈糝窶游蕕烽蜮塄料4.2.1 Normal Tests藎福蒙勤蟾閨湞賄肜觳冠綃AUT indicated that tests 10110 could be considered “normal processes”.纊竺擻絳嗝蜓叢諭邃坡哇碚4.2.2 AE Sensor Placement軛軍駁室樞喜宜蘿蝶梅錛攔AUT noted that the position of the acoustic emission sensor was altered prior to test 77, and was adjusted prior to subsequent tests. From

47、inspection of AE data, it appears that AE measurements are consistent after test 84, and so:乍慳皺碣譬耽庶播臣栳扼肫AE is assumed to be unusable for tests 176 the sensor records only white noise;襦貅陂鎢螗喊芤鐐?cè)舌挳a(chǎn)盾AE is assumed to be usable, but possibly abnormal, for tests 7783 the sensor position is being adjusted,

48、resulting in extreme variation in measurements;鍵勃瑯鳊洽牒岍其計輅淇街AE is assumed to be usable for tests 94190 the sensor position is held constant during these tests.骰咸抨全黷烤璜斐儆鸕擱崦Thus, the range of tests assumed to be normal 10110 should be reduced to 84110 when AE is considered.幸事歿慝板蜉暴心橇幃丶蜉4.3 Test Descript

49、ion嶧硐燕僉裎尖屨縛捆隧窄嗣Data recorded for during a typical test are shown in Figure 1. The duration of this test is approximately t=51 seconds. This section uses this test to illustrate a typical process, as described by AUT.蕺寺跤溯朋雙砦鑌實峻朝萑Drill power-on and power-off events may be seen at the start and end of

50、the test as transient spikes in SP.胖熗艄吩腭湫揶遭坎懿躐亍The drill unit is then moved towards the static disk at the constant feed rata specified in Table 1, between t=12 and 27 seconds. This corresponds to approximately constant values of SP during that period, approximately zero AE, and very lowamplitude ac

51、celeration in x-,y-,and z- planes.佑酥鰱醬懶鐵寧羞敕餳漚道At t=27 seconds, the drill makes contact with the static disk and begins to drill into the metal. This corresponds to a step-change in SP to a higher lever, staying approximately constant until t=38 seconds. During this time, AE increases significantly t

52、o a largely constant but non-zero value. The values Ax and Az increase throughout this drilling operation, while the value of Ay remains approximately zero (as it does throughout the test).彭弼杳脫儲營蓿嘴趄萁潭拍At t=38 seconds, the tip of the drill-bit passes through the rear face of the disk. The value of SP

53、 increases until t=44 seconds. During this period, AE reaches correspondingly high values, while Ax and Az decrease in amplitude.瀝浜鐲腕齋儕姜侈傣級騍滴At t=44 seconds, the direction of the drill unit is reversed, and the drill is retracted from the metal disk. Until t=46 seconds, the value of SP and AE decrea

54、se rapidly. A transient is observed in Ax and Az at t =44 seconds, with vibration amplitude decreasing until t=46 seconds.婪醋惘涮既蔦涇淆盼語琺饑At t=46 seconds, the drill-bit has been completely retracted from the metal disk, and the unit continues to be withdrawn at the feed rate until the end of the test. T

55、he value of SP decreases during this period(noting the power-off transient at the very end of the test), while the values of all three acceleration channels and AE are approximately zero.坻掌日鈁寥瑪繡貯睜憨涑拾朋禾鄴抉魁艮齪僨凡甘恣渚6 .Pre-processing訃岡霉夾篁確緗優(yōu)腭凇晡磷4.4 Removal of Start/Stop Transients壤戢播芄揮搬諑報咨棺憧繽Assuming tha

56、t normal and abnormal system behaviour will be evident from data acquired during the drilling process, prior to analysis, each test was shortened by retaining only data between the start and stop events, shown as transients in SP. For example, for the test shown in Figure 1, this corresponds to reta

57、ining the period 1350 seconds. 犀號闥鼾麓屯墑侍靴除遲掄4.5 Removal of Power Supply Signal芑鋤圄又裳盾果切嚏膨蛞壚The 50 Hz power supply appears with in each channel, and was removed prior to analysis by application of a band-stop filter with stop-band 4951 Hz.謫級繳差翩猖嫡圇艟椹亮奏4.6 Frequency Transformation薤憨鬏卸壽夫桫蟒欽郄嫻霉Data for eac

58、h test were divided into windows of 4096 points. A 4096-point FFT for was performed using data within each window, for Ax,Ay and Az channels. This corresponds to approximately 5 FFTs per second of data，similar to the QUICK system used in aerospace analysis， shown to provide sufficient resolution for identifying frequency-based eve