本文選題：數(shù)控車床載荷譜　切入點：多小波降噪　出處：《吉林大學》2017年博士論文　論文類型：學位論文

【摘要】：產(chǎn)品的固有可靠性首先是由設(shè)計決定的,對數(shù)控機床整機及關(guān)鍵功能部件進行可靠性設(shè)計是提高機床可靠性水平的主要途徑之一,而數(shù)控機床載荷譜是可靠性設(shè)計的客觀依據(jù);數(shù)控機床關(guān)鍵功能部件可靠性試驗是獲取產(chǎn)品故障數(shù)據(jù)、進行故障分析和建立可靠性模型的主要技術(shù)手段,數(shù)控機床載荷譜是開展模擬實際工況的功能部件可靠性臺架試驗的前提條件。因此,研究科學、合理的數(shù)控機床載荷譜的編制方法,并編制載荷譜,具有重要的理論意義和工程應用價值。近十多年來,隨著直驅(qū)技術(shù)、伺服技術(shù)和數(shù)字控制技術(shù)的發(fā)展和廣泛應用,現(xiàn)代數(shù)控機床在結(jié)構(gòu)、功能、性能指標、加工范圍等方面發(fā)生了巨大變化,依據(jù)經(jīng)驗公式編制的靜態(tài)切削力載荷譜不能準確反映真實工況,已經(jīng)不再適用于數(shù)控機床可靠性設(shè)計與功能部件可靠性試驗的發(fā)展需求;另外,編制反映動態(tài)切削載荷特征的數(shù)控機床載荷譜需要獲取大量的機床工藝載荷數(shù)據(jù),在此基礎(chǔ)上,確定典型工況并進行長時間的實際切削試驗測取動態(tài)載荷,該項研究工作耗時、費力,使很多研究者望而卻步。本文依托吉林大學機械工業(yè)數(shù)控裝備可靠性技術(shù)重點實驗室,結(jié)合國家科技重大專項課題,系統(tǒng)的研究了數(shù)控車床載荷譜的編制方法和流程,主要包括:典型測試工況的選擇、載荷數(shù)據(jù)的測取、載荷信號的處理、樣本長度的確定、載荷譜的外推、切削力譜的編制等內(nèi)容,并利用編制的載荷譜對機床的數(shù)控刀架進行了可靠性臺架試驗。論文主要研究內(nèi)容如下:(1)為確保載荷譜數(shù)據(jù)來源的真實性和有效性,本文在確定機床典型用戶的基礎(chǔ)上,研究制定了機床用戶現(xiàn)場工藝數(shù)據(jù)采集記錄表和機床運行記錄表,并對抽樣的數(shù)控車床進行了現(xiàn)場切削工藝數(shù)據(jù)和機床運行數(shù)據(jù)的采集;根據(jù)數(shù)控車床加工工況及工藝的特點,制定了車床切削力實測典型測試工況及工藝的選擇原則,提出了基于切削參數(shù)區(qū)間的典型測試工藝參數(shù)的確定方法;搭建了數(shù)控車床切削力測試系統(tǒng),并完成了典型測試工藝的切削力載荷實測試驗。(2)分析了切削力測試過程中的噪聲來源,針對低信噪比的切削力載荷信號,提出了改進的多小波相鄰系數(shù)降噪法,采用統(tǒng)計方差平滑法去除載荷信號的奇異點,通過引入關(guān)聯(lián)維數(shù)和BDS統(tǒng)計值等指標識別載荷信號中的噪聲,從而選出合理的多小波分解層數(shù)、小波函數(shù)和相鄰系數(shù)塊的長度;利用對不同信噪比的四種模擬信號的降噪處理,驗證了多小波相鄰系數(shù)法優(yōu)于經(jīng)典的小波閾值降噪法,并對實測載荷信號的降噪從時域波形、頻域能量和雨流域損傷三方面進行對比分析;為了保證采集的載荷數(shù)據(jù)樣本能夠代替總體的信息,提出了基于切削載荷特性的載荷樣本長度的確定方法。(3)根據(jù)數(shù)控車床恒速切削工況的載荷特性,提出了基于改進的時域載荷擴展的POT外推方法。由于傳統(tǒng)時域載荷擴展未考慮極值載荷間的時間間隔,利用改進的MCMC時域載荷擴展法,將每組載荷數(shù)據(jù)擴展到指定的載荷作用時間;然后根據(jù)MEF圖法確定載荷極值閾值的范圍,針對每個閾值對應的超越量進行GPD擬合,利用灰關(guān)聯(lián)分析法選擇最優(yōu)的閾值和載荷極值分布;合并恒速切削工況下的33組載荷的均幅值與頻次的矩陣,編制了數(shù)控車床恒速切削工況下的切削力譜。(4)根據(jù)數(shù)控車床變速切削工況的載荷特性,提出了基于混合威布爾分布的參數(shù)外推方法。利用改進的MCMC時域載荷擴展法將30組載荷數(shù)據(jù)擴展到指定的載荷作用時間;根據(jù)30組載荷數(shù)據(jù)合并后的均幅值分布特性,建立了載荷均幅值的混合威布爾分布模型;利用灰關(guān)聯(lián)分析法確定了混合威布爾分布的基本函數(shù)個數(shù),并利用改進的EM算法進行了參數(shù)估計;通過載荷的均幅值相關(guān)性分析,確定備選Copula函數(shù),采用平方歐式距離選擇最優(yōu)的Copula函數(shù),建立了載荷均幅值的聯(lián)合分布函數(shù),并編制了數(shù)控車床變速切削工況下的切削力譜。(5)以數(shù)控刀架可靠性臺架為例,編制了“切-轉(zhuǎn)-切”的八級切削力程序加載譜,確定了可靠性試驗過程中每步程序加載譜對應的動態(tài)切削力的加載角度、加載頻率以及刀架的換刀頻率;制定了基于載荷譜的數(shù)控刀架可靠性試驗方案,在實驗室開展了數(shù)控刀架可靠性臺架試驗,并針對試驗中出現(xiàn)的小樣本數(shù)據(jù)的情況,提出了基于融合多源先驗信息的貝葉斯可靠性建模與評估方法,依據(jù)數(shù)控刀架實驗室可靠性試驗數(shù)據(jù),建立了受試數(shù)控刀架的可靠性評估模型。通過本文的研究,提出了一套適用于數(shù)控車床切削力譜的測取、編制和應用的方法,為后續(xù)開展數(shù)控車床及關(guān)鍵功能部件的可靠性試驗、加速試驗、壽命預測和可靠性設(shè)計等提供理論依據(jù)。
[Abstract]:The inherent reliability of the product is the first by design, the machine and key part of NC machine reliability design is one of the main ways to improve the reliability level of machine tools, CNC machine tools and the load spectrum is the objective basis of the reliability design; key functional parts of NC machine tools reliability test is to obtain product failure data, and to establish the reliability model of the main technology of NC machine tool fault analysis, load spectrum is the precondition of simulation components reliability bench test of actual operating conditions. Therefore, the scientific research, the reasonable numerical control preparation method of machine load spectrum, and the preparation of load spectrum, which has important theoretical significance and engineering application value. In the past more than 10 years, with direct drive technology development and wide application of servo technology and digital control technology, modern CNC machine tools in the structure, function, performance index, processing range Great changes have taken place in other aspects, on the basis of experience formulas, the static cutting force spectrum can not accurately reflect the real conditions, development demand is no longer applicable to the reliability of CNC machine tool design and functional components of the reliability test; in addition, preparation of NC machine tool dynamic cutting load reflects the load spectrum of the need to get the machine sign treeing process load data, this on the basis of determining the typical working conditions and cutting tests of long time measurement of dynamic load, the work is time-consuming, laborious, so many researchers from. This paper rely on Jilin University Machinery Industry Key Laboratory of CNC equipment reliability technology, combined with the national science and technology major project, systematic study of the preparation method and process of NC lathe load the spectrum mainly includes: the typical test conditions, measured load data, the load signal processing, sample length The degree of determination, extrapolation of the load spectrum, such as the preparation of the contents of the cutting force spectrum, and the reliability test bench of machine tool CNC turret spectrum using the load. The main contents are as follows: (1) to ensure the authenticity and validity of the load spectrum data sources, this paper is based on determining the typical user machine on the study and formulate the machine tool user process data in the field acquisition table and machine operation records, and conducted on-site sampling of CNC lathe cutting process data and machine operation data acquisition; according to the characteristics of CNC lathe processing conditions and process, formulated the selection principle of lathe cutting force data of typical test conditions and process. Put forward the method of determining the typical test parameters of cutting parameters based on interval; build a CNC lathe cutting force testing system, and completed the typical cutting force test process load. Test. (2) analyzed the noise sources of the cutting force in the process of testing, according to the cutting force signal with low SNR, the proposed denoising method improved wavelet coefficients of adjacent singular points, removal of load signals using statistical variance smoothing method, using correlation dimension and BDS statistical indexes such as load identification signal in the noise, to choose the more reasonable wavelet decomposition level, wavelet function and adjacent coefficient blocks using length; noise reduction of four different signal-to-noise ratio of the analog signal, verify the wavelet threshold denoising method outperforms the classical multiwavelet neighboring coefficient method, and the measured load signal denoising from time domain waveform, frequency domain energy and rain damage basin three aspects of comparative analysis; in order to ensure the load data can replace the general information, put forward the method of determining load sample size based on the characteristics of cutting load. (3) according to the load characteristics of CNC lathe with constant speed cutting conditions, put forward the POT extrapolation method in time domain based on improved extended load. Because of the traditional time domain load expansion without considering the extreme load of the time interval between the expansion by using the method of MCMC time domain load improved, each load data extended to load the specified time; and determine the scope of the extreme load value threshold according to the MEF diagram, GPD fitting for beyond the amount of each threshold corresponding to the use of grey relational analysis method to select the optimal threshold and load extreme value distribution; 33 groups with constant speed load matrix cutting under the condition of both amplitude and frequency, the cutting force under the condition of constant speed cutting CNC lathe the spectrum is compiled. (4) according to the load characteristics of variable speed cutting condition of CNC lathe, puts forward parameter extrapolation method based on mixed Weibull distribution. Using MCMC time-domain expansion method improved 30 Group load data extended to load the specified time; according to the 30 group load data after the merger are amplitude distribution characteristics, the mixed Weibull distribution model of load amplitude; analysis method to determine the number of mixed Weibull distribution is the basic function of the grey correlation, and parameter estimation using EM algorithm improved by; the correlation analysis of the load amplitude were determined, alternative Copula function, using the squared Euclidean distance Copula optimal function, to establish a joint distribution function of load amplitude, and the cutting force of CNC lathe cutting speed under the condition of the spectrum. (5) the reliability of CNC turret rig as an example, eight level cutting force program preparation of loading "cut to cut" spectrum to determine the dynamic cutting force of each program load spectrum corresponding to the reliability test in the process of loading angle, loading frequency and frequency change knife knife system; The test scheme of CNC turret reliability based on load spectrum, in the laboratory to carry out CNC turret reliability bench test, and based on small sample data in the test case, Bias proposed reliability modeling and evaluation method for multi-source information fusion based on the basis of the reliability test data of CNC turret laboratory, established reliability evaluation model try CNC turret. Through this research, put forward a set of suitable for CNC lathe cutting force spectrum measurement, method of preparation and application of the reliability test, the component for further research on the CNC lathe and the key to successful accelerated test, to provide a theoretical basis for life prediction and reliability design.

【學位授予單位】：吉林大學
【學位級別】：博士
【學位授予年份】：2017
【分類號】：TG519.1

【相似文獻】

相關(guān)期刊論文 前10條

1 賈志新,艾冬梅,賈亞洲,邱成;數(shù)控車床致命性分析[J];機床與液壓;2000年04期

2 朱福敏;數(shù)控車床故障分析與檢修[J];中國設(shè)備管理;2000年06期

3 胡正隆;車輛輪對落輪數(shù)控車床研制成功[J];機床與液壓;2001年04期

4 趙鍵;如何選擇數(shù)控車床[J];機械工程師;2001年03期

5 ;車輛輪對落輪數(shù)控車床[J];機電工程技術(shù);2001年03期

6 梁焱;數(shù)控車床實習探討[J];中山大學學報論叢;2002年05期

7 王繼興;數(shù)控車床上巧用U鉆加工小孔[J];機械工人.冷加工;2002年03期

8 ;精誠恒泰科技公司推出高精密復合式數(shù)控車床[J];模具制造;2002年09期

9 鐘友志;數(shù)控車床維修兩例[J];機床電器;2003年05期

10 宗國成;數(shù)控車床維修一例[J];制造技術(shù)與機床;2004年01期

相關(guān)會議論文 前10條

1 費金紅;;西德數(shù)控車床的系統(tǒng)改造[A];設(shè)備維修與改造技術(shù)論文集[C];2000年

2 沈興全;;數(shù)控車床編程模擬加工系統(tǒng)開發(fā)與研制[A];面向制造業(yè)的自動化與信息化技術(shù)創(chuàng)新設(shè)計的基礎(chǔ)技術(shù)——2001年中國機械工程學會年會暨第九屆全國特種加工學術(shù)年會論文集[C];2001年

3 沈興全;宋文超;;數(shù)控車床編程模擬加工系統(tǒng)開發(fā)與研制[A];第二屆全國信息獲取與處理學術(shù)會議論文集[C];2004年

4 宋現(xiàn)春;王兆坦;劉現(xiàn)銀;;三米數(shù)控車床實現(xiàn)螺補功能的原理及方法[A];機床與液壓學術(shù)研討會論文集[C];2004年

5 宋現(xiàn)春;王兆坦;劉現(xiàn)銀;;三米數(shù)控車床實現(xiàn)螺補功能的原理及方法[A];第三屆全國流體傳動及控制工程學術(shù)會議論文集（第二卷）[C];2004年

6 謝紅;沈斌;程堂華;吳炳超;;數(shù)控車床的虛擬設(shè)計和有限元分析[A];先進制造技術(shù)論壇暨第五屆制造業(yè)自動化與信息化技術(shù)交流會論文集[C];2006年

7 肖俊;胡偉;魯志政;陳志俊;楊建國;;數(shù)控車床故障的可靠性分析[A];2006年全國機械可靠性學術(shù)交流會論文集[C];2006年

8 李倩;;數(shù)控車床精加工復雜表面編程的幾何處理[A];中國當代教育理論文獻——第四屆中國教育家大會成果匯編（上）[C];2007年

9 李耀剛;張文明;;數(shù)控車床與基礎(chǔ)耦合振動數(shù)學模型的研究[A];全國金屬礦山采礦新技術(shù)學術(shù)研討與技術(shù)交流會論文集[C];2007年

10 趙韓;田杰;張棟;徐林森;;一種新型的多平面切削數(shù)控車床[A];2004“安徽制造業(yè)發(fā)展”博士科技論壇論文集[C];2004年

相關(guān)重要報紙文章 前10條

1 李文;普及型數(shù)控車床產(chǎn)業(yè)化工程的開發(fā)研究效果好[N];中國機電日報;2002年

2 佚名;數(shù)控車床工概述[N];就業(yè)時報;2007年

3 ;濟南開發(fā)出四軸數(shù)控車床[N];今日信息報;2005年

4 ;普及型數(shù)控車床產(chǎn)業(yè)化工程的開發(fā)研究[N];中國機電日報;2002年

5 成吉昌;數(shù)控車床節(jié)省大量外匯[N];中國船舶報;2001年

6 本報記者 宋文華;圍繞三個“新”字做文章[N];中國機電日報;2002年

7 吳黎旺;提升數(shù)控車床附加值的新利器[N];中國工業(yè)報;2008年

8 本報記者 司成鋼;愛唱“咱們工人有力量”的青年技工[N];遼寧日報;2008年

9 ;普及型數(shù)控車床產(chǎn)業(yè)化工程的開發(fā)研究[N];中國機電日報;2002年

10 本報記者 趙春偉　通訊員 丁嘉禮 紀惠婷;化蛹成蝶[N];張家口日報;2009年

相關(guān)博士學位論文 前4條

1 何佳龍;數(shù)控車床切削力譜的編制及其在功能部件可靠性試驗中的應用[D];吉林大學;2017年

2 孟書;故障率相關(guān)下數(shù)控車床預防維修策略研究[D];吉林大學;2017年

3 樊少華;基于模糊理論的數(shù)控車床可靠性分配[D];吉林大學;2011年

4 李慶興;異型軋輥數(shù)控車床切削進給系統(tǒng)的設(shè)計理論及其關(guān)鍵技術(shù)研究[D];河北工業(yè)大學;2009年

相關(guān)碩士學位論文 前10條

1 李研;國內(nèi)外數(shù)控車床可靠性對比分析[D];吉林大學;2006年

2 陳云雷;面向并行工程的數(shù)控車床可靠性研究[D];吉林大學;2004年

3 張從從;數(shù)控車床的綠色化再設(shè)計研究[D];大連交通大學;2013年

4 鮑成偉;基于MBD技術(shù)的數(shù)控車床知識庫系統(tǒng)[D];西安工業(yè)大學;2014年

5 陳博;基于人機工程學的數(shù)控車床人機界面的設(shè)計研究[D];西安工程大學;2015年

6 朱陽;雙刀式數(shù)控車床結(jié)構(gòu)設(shè)計、分析及優(yōu)化[D];浙江工業(yè)大學;2011年

7 阮忠誠;數(shù)控車床虛擬樣機的研究[D];東北大學;2013年

8 劉云好;精密數(shù)控車床的顫振分析[D];東北大學;2013年

9 辛燕;數(shù)控車床切削參數(shù)優(yōu)化研究[D];昆明理工大學;2016年

10 柏博;數(shù)控車床低能耗切削工藝參數(shù)優(yōu)化試驗研究[D];大連海事大學;2016年

，

本文編號：1562921