本文關(guān)鍵詞：基于機(jī)器視覺(jué)的共軌針閥尺寸參數(shù)在線測(cè)量監(jiān)控系統(tǒng)的研究與開(kāi)發(fā)　出處：《江蘇大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： 機(jī)器視覺(jué) 在線檢測(cè) 共軌針閥 尺寸檢測(cè) 圖像處理

【摘要】：共軌噴油嘴針閥偶件是柴油機(jī)燃油噴射系統(tǒng)的三大精密偶件之一,在工業(yè)大馬力柴油發(fā)動(dòng)機(jī)以及輕型客、轎車(chē)柴油發(fā)動(dòng)機(jī)上均有使用,是保證柴油發(fā)動(dòng)機(jī)燃油噴射系統(tǒng)穩(wěn)定運(yùn)行的關(guān)鍵部件之一,其精度要求高、產(chǎn)量大,相應(yīng)地測(cè)量精度、檢測(cè)效率等要求也隨之提高。如何實(shí)現(xiàn)共軌針閥零件尾部尺寸參數(shù)的快速、穩(wěn)定、高效、高精和全數(shù)在線測(cè)量,是目前工業(yè)生產(chǎn)中亟待解決的問(wèn)題。本文圍繞某外資企業(yè)噴油嘴共軌針閥零件的尾部尺寸參數(shù)測(cè)量與過(guò)程控制所展開(kāi)。針對(duì)視覺(jué)在線測(cè)量系統(tǒng)的精度、效率、穩(wěn)定性、完整性等難點(diǎn),在系統(tǒng)平臺(tái)的整體設(shè)計(jì)、圖像高精度處理、系統(tǒng)參數(shù)標(biāo)定、光機(jī)電一體化控制及系統(tǒng)開(kāi)發(fā)等方面進(jìn)行了一定的研究。本文的主要研究?jī)?nèi)容和成果如下:首先,進(jìn)行視覺(jué)在線測(cè)量監(jiān)控系統(tǒng)的整體設(shè)計(jì)。具體包括光學(xué)視覺(jué)系統(tǒng)硬件選型匹配和運(yùn)動(dòng)機(jī)構(gòu)方案設(shè)計(jì)。結(jié)合目標(biāo)零件輪廓特征、待測(cè)尺寸、企業(yè)生產(chǎn)線狀況,確定光學(xué)視覺(jué)系統(tǒng)的打光方式,并進(jìn)行相機(jī)、鏡頭及光源的選型;基于光源在視覺(jué)系統(tǒng)中的重要性,定性研究了復(fù)合光和單色光對(duì)測(cè)量系統(tǒng)的影響,得出單色光在背光方式中效果優(yōu)異;同時(shí),研究了光強(qiáng)與輪廓像素定位之間的關(guān)系,得出本文視覺(jué)系統(tǒng)在某一光強(qiáng)范圍內(nèi),效果達(dá)到最優(yōu);根據(jù)待測(cè)零件特征,確定上料輸送、定位及分揀運(yùn)動(dòng)機(jī)構(gòu)方案。其次,對(duì)光學(xué)系統(tǒng)獲取的共軌針閥圖像進(jìn)行處理算法設(shè)計(jì)。在降噪方面,通過(guò)濾波對(duì)比實(shí)驗(yàn),選擇了中值濾波算子,引入基于中值濾波的小區(qū)域?yàn)V波算法,大大縮短了濾波時(shí)間;在輪廓邊緣檢測(cè)方面,通過(guò)對(duì)比實(shí)驗(yàn),引入了基于Sobel的小區(qū)域定向搜索邊緣檢測(cè)算法,提高了邊緣定位效率,并采用小區(qū)域邊緣特征點(diǎn)跟蹤方法,基于學(xué)習(xí)訓(xùn)練邏輯,完成目標(biāo)特征點(diǎn)定位;在邊緣亞像素定位和擬合方面,引入基于穩(wěn)健線性回歸的直線擬合方法,在特定小區(qū)域下完成高精度邊緣擬合,并完成直徑、角度、距離的多參數(shù)同時(shí)計(jì)算。然后,深入地進(jìn)行了視覺(jué)系統(tǒng)標(biāo)定的理論研究與設(shè)計(jì)。對(duì)相機(jī)系統(tǒng)的參數(shù)標(biāo)定、鏡頭的光學(xué)畸變及校正等進(jìn)行了說(shuō)明;分析研究了時(shí)下常用了標(biāo)定方法,基于Microscape軟件系統(tǒng),采用兩步標(biāo)定法,完成相機(jī)內(nèi)外參數(shù)及鏡頭畸變的標(biāo)定與校正;運(yùn)用標(biāo)準(zhǔn)網(wǎng)格及標(biāo)準(zhǔn)件,對(duì)相機(jī)位置及像素當(dāng)量進(jìn)行校準(zhǔn)和計(jì)算。接下來(lái),對(duì)在線視覺(jué)測(cè)量監(jiān)控系統(tǒng)軟件、硬件控制進(jìn)行了具體的設(shè)計(jì)與開(kāi)發(fā)。在Microscape視覺(jué)系統(tǒng)下設(shè)計(jì)圖像算法處理過(guò)程,并基于VB完成了其在視覺(jué)檢測(cè)、控制、統(tǒng)計(jì)計(jì)算、人機(jī)交互等方面的二次開(kāi)發(fā);同時(shí)基于統(tǒng)計(jì)學(xué)理論設(shè)計(jì)了工序質(zhì)量實(shí)時(shí)監(jiān)控系統(tǒng),為生產(chǎn)線的實(shí)時(shí)在線監(jiān)控、統(tǒng)計(jì)和分析提供完整應(yīng)用;運(yùn)用CATIA軟件對(duì)整體自動(dòng)化機(jī)構(gòu)進(jìn)行設(shè)計(jì),完成整體上料輸送、定位夾緊及下料分揀機(jī)構(gòu)的三維圖繪制,并設(shè)計(jì)了系統(tǒng)工作流程、自動(dòng)化設(shè)備的通信與控制,完成整套系統(tǒng)的邏輯與運(yùn)行控制。最后,進(jìn)行了樣機(jī)研制與可靠性評(píng)價(jià),通過(guò)測(cè)試,運(yùn)行情況良好。借助專用的Q-DAS數(shù)據(jù)統(tǒng)計(jì)分析軟件,對(duì)樣機(jī)平臺(tái)系統(tǒng)的穩(wěn)定性、偏倚能力指數(shù)、重復(fù)再現(xiàn)性進(jìn)行實(shí)驗(yàn)及分析評(píng)價(jià),統(tǒng)計(jì)分析表明樣機(jī)對(duì)共軌針閥零件尾部直徑、角度、距離的尺寸測(cè)量,在公差范圍內(nèi)達(dá)到了10倍以上的測(cè)量精度,系統(tǒng)亞像素分辨率達(dá)到了0.0008~0.0003mm,滿足了工業(yè)實(shí)際應(yīng)用需求;并以直徑參數(shù)為例,測(cè)試了平臺(tái)的工序質(zhì)量實(shí)時(shí)監(jiān)控統(tǒng)計(jì)系統(tǒng),驗(yàn)證了本文的理論研究成果的可行性、有效性。本文完成了共軌針閥零件的尾部尺寸參數(shù)在線測(cè)量監(jiān)控系統(tǒng)的研究與開(kāi)發(fā),對(duì)其他同類產(chǎn)品的在線測(cè)量具有重要的借鑒作用。
[Abstract]:The common rail injector needle valve is one of the three big precision coupling of fuel injection system in diesel engine industry, and light passenger car diesel engine are used, is to ensure that the diesel engine fuel injection system is one of the key components of its stable operation, high precision, large output, the corresponding measurement accuracy so, the detection efficiency requirement is increasing. How to realize the common rail valve parts tail size parameters of fast, stable, high efficiency, high precision and full online measurement, is the current industrial production problems to be solved. This paper focuses on a foreign enterprise common rail injector needle valve parts of the tail size parameter measurement and process control. According to the visual online measurement system accuracy, efficiency, stability, integrity and other difficulties, in the overall design of system platform, high precision image processing, system calibration, optical electromechanical The research of integrated control and system development. The main research contents and results are as follows: firstly, the overall design of measuring and monitoring system for visual online. Including the design of optical vision system hardware matching and motion mechanism scheme combining the target contour features, the size to be measured, the enterprise production line status determination of the optical visual system lighting methods, and selection of camera lens and light source; light importance in the vision system based on the qualitative research on the effects of compound light and monochromatic light on the measurement system, the monochromatic light in the backlight in excellent effect; at the same time, study the relation between the light intensity and the contour pixel location. The visual system in a certain intensity range, to achieve the optimal effect; according to the features of parts to be measured to determine material conveying, sorting and positioning movement mechanism second scheme, Processing algorithm design of common rail valve image optical system is obtained. In denoising, filtering through the contrast experiment, choose the median filtering operator, introduce small domain filtering algorithm based on median filtering, greatly shorten the filtering time; in the edge detection, through the contrast experiment, introduces the search algorithm of edge detection area orientation based on Sobel, and improves the edge localization efficiency, and the tracking feature point area edge training logic based on target feature points, positioning; in sub-pixel edge locating and fitting, introducing the linear fitting method of linear regression based on the completion of high precision edge fitting in specific area, and complete the diameter at the same time, angle, distance calculation of multi parameter. Then, in-depth theoretical research and design of the vision system calibration. The calibration parameters of the camera lens system. The head of the optical distortion and correction are described; analysis of the nowadays commonly used calibration method, based on Microscape software, using two step calibration method, calibration and calibration of camera parameters and lens distortion; using standard grid and standard parts, calibration and calculation of the camera position and the pixel equivalent. Next, the online visual measurement monitoring system software, hardware control is designed and developed. The design of specific image processing algorithms in the process of Microscape vision system based on VB, and completed its control in visual inspection, statistical calculation, two times the development of human-computer interaction and so on; at the same time, based on statistical theory and design of real-time monitoring system of process the quality of real-time online monitoring for production line, to provide a complete application of statistics and analysis; application of integrated automation mechanism design CATIA software, complete the whole conveying, set A three-dimensional map of clamping and cutting the sorting mechanism is drawn, and the design of the workflow system, communication and control automation equipment, complete set of logic and operation control system. Finally, the prototype development and reliability evaluation, through the test, running in good condition. With the help of Q-DAS data for statistical analysis software, the stability of the prototype the platform system, the bias capability index, repeatability of experimental evaluation and analysis, statistical analysis showed that the prototype of common rail valve parts, tail diameter, distance measurement, the measurement accuracy reached more than 10 times in the tolerance range, sub-pixel resolution system has reached 0.0008~0.0003mm, and meets the requirements of industrial application; and the diameter of parameters for example, process quality real-time monitoring and statistical system platform test, verify the feasibility of this paper, the theoretical research results. The validity of this paper. It has become the research and development of on-line measurement and monitoring system for tail size parameters of common rail needle valve parts, which has important reference for other similar products online measurement.

【學(xué)位授予單位】：江蘇大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP391.41;TP277

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 于善虎;龔堤;葉鳴;張威;溫必劍;;柴油機(jī)噴油器針閥密封錐面磨損試驗(yàn)研究[J];潤(rùn)滑與密封;2016年02期

2 李旭;王俊元;曾志強(qiáng);張川川;林娣;;色差對(duì)機(jī)器視覺(jué)尺寸測(cè)量精度的影響研究[J];包裝工程;2015年19期

3 葉雯;;基于MATLAB的圖像去噪研究[J];計(jì)算機(jī)時(shí)代;2015年06期

4 鄭多;王偉;;M1080B無(wú)心外圓磨床改進(jìn)[J];哈爾濱軸承;2015年02期

5 馬強(qiáng);趙熹;;基于機(jī)器視覺(jué)的牛頓環(huán)自動(dòng)測(cè)量系統(tǒng)研究[J];電子技術(shù)與軟件工程;2015年05期

6 尤睿驊;王樹(shù)林;呂冬;周宏達(dá);;電控高壓共軌柴油機(jī)噴油嘴針閥表面質(zhì)量的研究[J];機(jī)械設(shè)計(jì)與制造;2015年02期

7 陳杰;;淺談圖像分割原理和方法[J];信息與電腦(理論版);2010年03期

8 Valtteri Heiskanen;Kalle Marjanen;Pasi Kallio;;Machine Vision Based Measurement of Dynamic Contact Angles in Microchannel Flows[J];Journal of Bionic Engineering;2008年04期

9 蘇健凌;梁建;;機(jī)器視覺(jué)技術(shù)在煙草行業(yè)的應(yīng)用[J];廣西質(zhì)量監(jiān)督導(dǎo)報(bào);2008年06期

10 陳建偉;劉瓔瑛;;基于機(jī)器視覺(jué)技術(shù)的大米品質(zhì)檢測(cè)綜述[J];糧食與食品工業(yè);2008年03期

相關(guān)博士學(xué)位論文 前3條

1 周文舉;基于機(jī)器視覺(jué)的在線高速檢測(cè)與精確控制研究及應(yīng)用[D];上海大學(xué);2014年

2 于忠黨;基于支持向量機(jī)的機(jī)械零件計(jì)算機(jī)視覺(jué)檢測(cè)若干關(guān)鍵技術(shù)的研究[D];吉林大學(xué);2009年

3 張娜;超高速數(shù)字CMOS圖像傳感器關(guān)鍵技術(shù)研究[D];天津大學(xué);2008年

相關(guān)碩士學(xué)位論文 前10條

1 區(qū)炳煜;基于機(jī)器視覺(jué)的鋸條缺陷檢測(cè)系統(tǒng)開(kāi)發(fā)[D];廣東工業(yè)大學(xué);2016年

2 焦亮;基于機(jī)器視覺(jué)的精密零件高精度測(cè)量與自動(dòng)化裝配系統(tǒng)的軟件及實(shí)驗(yàn)研究[D];華南理工大學(xué);2016年

3 龔香;運(yùn)動(dòng)車(chē)輛目標(biāo)提取與跟蹤技術(shù)研究[D];電子科技大學(xué);2015年

4 劉國(guó)陽(yáng);基于機(jī)器視覺(jué)的微小零件尺寸測(cè)量技術(shù)研究[D];哈爾濱工業(yè)大學(xué);2014年

5 龔聰;基于機(jī)器視覺(jué)的高精度尺寸檢測(cè)方法與實(shí)現(xiàn)[D];廣東工業(yè)大學(xué);2014年

6 董富強(qiáng);基于機(jī)器視覺(jué)的零件輪廓尺寸精密測(cè)量系統(tǒng)研究[D];天津科技大學(xué);2014年

7 李慧德;基于機(jī)器視覺(jué)的在線尺寸測(cè)量方法研究[D];天津理工大學(xué);2014年

8 楊戈;基于機(jī)器視覺(jué)的機(jī)械零部件精密測(cè)量系統(tǒng)設(shè)計(jì)與開(kāi)發(fā)[D];南京信息工程大學(xué);2013年

9 張美靜;亞像素邊緣檢測(cè)技術(shù)研究[D];沈陽(yáng)理工大學(xué);2013年

10 廖維;基于機(jī)器視覺(jué)的工件加工尺寸在線測(cè)量方法與系統(tǒng)研究[D];武漢理工大學(xué);2012年

，

本文編號(hào)：1371713