工程机械液压系统的远程视情维护方案

本资料包含pdf文件1个，下载需要1积分

Sep．2013机 床 与 液 压Hydromechatronics Engineering Vo1．41 No．18DOI：10．3969／j．issn．1001-3881．2013．18．005Remote Condition-based M aintenance Approach to HydraulicSystem of Construction M achineryZHAO Xiuxu ，HU Zhemin ，JIANG Jihai ，ZHANG Qingrun。

． School of Mechtronics Engineering．Wuhan University2．Harbin Institute of Technology，School ofMechanical1．Introductionof Technology，Wuhan 430070，China；and Electrical Engineering，Harbin 150001，ChinaAbstract：Hydraulic system has been widely applied in high-power and automated constructionmachinery．But there are many dificulties in its fault diagnosis because of the complicated struc-ture of hydraulic system．Especialy in the working site，it may cause the shutdown of work whenthe fault can’t be removed timely．In this study，an extended failure mode and efects analysis(FMEA)method is presented to integrate expe~knowledge and working data．A remote condi-tion—based maintenance approach is presented，and an application system with FMEA knowledgebase has been developed for construction machinery hydraulic system．The system structure andsome key problems such as data acquisition，fault features extraction，as wel as FMEA knowl-edge base is discussed．An application example shows the system is content.

Key words：construction machinery，hydraulic system ，remote condition—based maintenance，ex-tended FMEA，FMEA knowledge baseHydraulic system has advantages of high power，smal volume，high accuracy，so it is widely appliedin hgh—power and automated construction machinery『1]．While its structure becomes more and morecomplicated，there are more and more difficulties tocontrol the working process because of its closedstructure．It is also the reason why fault diagnosis ofhydraulic system has been widely concerned by re—searchers in recent years.

Generally，there are model-based method，sig—na1．based method and knowledge．based method usedin fault diagnosis of hydraulic system[2]．Practical—ly，there are some disadvantages in each method，socombination of them is always adopted．Because theReceived：2013—05—24Sponsored by Education Department of P．R．China&Guangdong Prov．Corporation Research Project(2010B090400548)ZHAO Xiuxu．E—mail：zhaoxiuxu### whut．edu．an.

complexity and nonlinearity of hydraulic system，knowledge·-based method has been used widly togeth-er with the other methods．For example，Azadeh es—tablished a fuzzy inference system for pump failurediagnosis supporting by knowledge[3]．Dong presen—ted a statistical hidden Markov model to perform bothdiagnosis and prognosis for hydraulic pump[4].

Jiang proposed a combination method of wavelet anal-ysis and exponentially weighted dynamic kernel prin—cipal component analysis to identify fault of hydraulicpump[5]．Angeli summarized study on online expertsystem as parameter estimation， statistical method，knowledge—based method，neural network and quali-tative simulation and conclused that it tend to COU—pling use of these diagnostic techniques[6].

Generaly，knowledge-based method fully utili—zes both expels’knowledge and diagnostic strategy，provides intelligent data analysis ability for diagnosisand maintenance，avoids the error in application withmodeling and issue on how to select the fight instru—ment and method of signal analysis，and developsrapidly[6]．It can locate and clear faults fast，espe—26 Hydromechatronics Engineeringcially in a complex and nonlinear system．Typically，the knowledge integration is a key issue．A complexbase of experience knowledge and analytical knowl—edge wil improve the accuracy and reliability of diag—nosis.

As an actuator，the status of hydraulic cylinderaffects the system performances，the common faults ofhydraulic cylinder are leakage and creeping whichlead the lost of thrust or velocity and unsteady work-ing[7]．It wil have serious influence on the per.

form ance of system，causing it is important to diagno—sis it timely and accurately． Liu applied Bayesiannetworks to build a cylinder creeping model to detectcauses[8]．Goharizi established a wavelet trans—form—based approach and empirical mode decomposi-tion and Hilbert spectrum to detect internal leakagein hydraulic actuators[9—10].

~From the cost perspective，condition—based ma—intenance(CBM)is a maintenance program that rec-ommends maintenance actions based on the infor ma—tion collected through condition monitoring． It at—tempts to avoid unnecessary maintenance tasks bytaking maintenance actions only when there is evi-dence of abnormal behaviours of a physical asset．ACBM program，if properly established and effectivelyimplemented，can significantly reduce maintenancecost by reducing the number of unnecessary sched—uled preventive maintenance operations[1 1]．In thisstudy， a remote condition—based maintenance ap—proach for hydraulic system in construction machineryis presented，an application system has been devel—oped．In the following sections，the system structure，and the key problems in this approach are discussed，a case study is presented in section 4．Finally，weconclude the paper by reiterating our main contribu—tions.

production life cycle．The main objective o±FMEA isto discover and prioritize the potential failure modes(by computing respective RPN)，which pose a detri—mental effect on the system and its performance．Areliability centered maintenance methodology basedon FMEA is developed for improvement[13]．An in—tegration model of FMEA and QFD(Quality FunctionDeployment)is presented in literature[14]．Thesemethods are connected by feedback relation in eachcycle to improve product quality.

FMEA provides the ability to identify potentialfaults，evaluate severity，occurrence and detection ofthe fault，determ ine the effects，and propose solvingmeasures，a coresponding knowledge base can storeuseful information and support the decision makingprocess[1 5]．But traditional FMEA is lack of logfault symptom and monitoring information， cannotmeet the need in this study．To solve the problem ，we propose extended FMEA to integrate expert knowl-edge and working data for condition-based mainte-nance．As shown in Tab．1，some typical failuressuch as sealing and leakage etc．are listed．And al—so，Sensor Installation describes the type and site ofsensors，and Detection Parameters is a record of sec—tions of features to achieve threshold judgment．Therestill need some field to explain the relation betweensensor and feature analyzed in Section 4．3 and 4．4while are not given in the table.

Expert knowledge is the core of database，butit’s ambiguous．During the analysis，we calculatedthe values of severity， occurrence， detection andRPN by fuzzy comprehensive evaluation method，en-suring suitable safety assessment[1 6].

3．System design and Key problems discus-sion2．Integration of fault knowledge based on 3·1·System structureextended FM EAPracticaly，different equipments need to makedifferent maintenance decision under different situa—tions． For industrial application．cost considerationalways plays the major role，while it is reliability formilitary equipment[1 2]．That is，the troubleshootingdecision should be a deliberated tradeof with systemstructure，eficiency，and cost.

FMEA is usually used as design FMEA，proce-dure FMEA and equipment FMEA，performing in theAs shown in Fig．1，the architecture of systemconsists of 3 layers，namely data layer，technologylayer and application layer.

Data layer is mainly responsible for data acquisi—tion．In most practical cases．inaccurate fauh analy.

sis may be caused by the inform ation come from onlyone type of sensor；sometimes it is unilateral andfuzzy．So it is necessary to collect the related datawith muhi-sensor．In this system，it contains pres—sure sensors，displacement sensors，wireless data ac—quisition card and the front end software．Data acqui—ZHAO Xiuxu，et al：Remote Condition—based Maintenance Approach to HydraulicSystem of Construction Machinery 27sition card with corresponding program collects thedisplacement and pressure sensors readings of hy—draulic cylinder through Wi-Fi， making diagnosingdecision、under multi—task mode
． With LabVIEWplatform，seamless connection between card and soft—ware can be realized． Wavelet packet method is a—dopted to process the non—stationary signals for thefault features extracting of hydraulic system.

Tab．1 Examples of e~ended FMEA knowledgelu弛 M。de C
ause S O D RPN M
ode ID Efl_ectsSensorInstalationDetection RecommendedParameters ActionsCylinderCylinderCylinderCylinderAbrasionof rodsealingDestructionof rodsealingMediuminternalleakageCO1 ／CO2 CO1C03 C01Serious C04 C02intemalleakageC02； 2 7 7 98C03；，7 2 3 42 ／(Some 3 5 2 30systemfailure)(Some 8 2 2 32system .

failure)DisplacementSensor onrod；PressureSensor onoutput；DisplacementSenS0ron rod；Pressuresensor on．
output；，，[130，124]；[80，4O]；[2．40，2．50][124，0]；[40，0]；[2．50，。]Replacesealing；Replacesealing；Stop to checkrod，sealingand block；Stop to checkrod，sealingand block；Fig．1 Framework of SystemTechnology layer is the core of system ，inclu-ding the engine of fault analysis and inference ma—chine．A threshold judgment is run in front end soft-ware firstly to detect fault and get its fault features.

Then the fault features will be transferred to back endservice to identify the fault level and stored in data-base．Treating，maintaining and preventing measuresare generated by FMEA knowledge base and sent tothe front end software.

Application layer provide the interactions be—tween front end or back end users．For front end us—ers．the result of diagnosis is send back to software.

For back end users，the statistical data or cha~swhich ale important for product designing and upgra—ding can be obtained and summarized.

3．2．Information integrationAlthough all modules are achieved in the sameWeb server of technology layer in this system．thedata exchange is relatively simple，by file streams orSQL server．Considering of the trend of CBM f condi—tion based maintenance) system and application ofdistilbute system．we still treat inform ation integra—tion as a key issue，attempting to satisfy the structureand standardization of MIMOSA OSA—EAI and ISO.

Fig．2 shows the inform ation flows among themodules． The critical functions are processed inWebservice layer． Because of data volume． thesefunctions may be processed in independent server，while the functions are repalceable and composable，28 Hydromechatronics Engineeringdemanding unifv and logical interfaces should bemade amaong them．In our study，the father functionrequest child function by S0AP／HTTP request．andthe child return XML file as a format result to father.

The data will be stored in database in the same time.

AI1 the function are caled by public interfaces．Web.

service API and Winforift API programed by ASP.

NET，in Fig．2．When user run a public interface，such as nlainteneance repol·t Interface．it wil actuallytrigger the nlainteneance report function．Firstly，thefunction searches database．if the data and informa—tion is absent，the child function，desicion supportfunction wj11 be run.

Fig．2 Information Integration Model3．3．Remote data acquisitionConsidering the practical working condition ofconstruction machiltory，the reliability and accuracyof hydraulic system may be influenced by contamina—tire oil． In this study，wireless data acquisition isused to collect data of system status，and remote ma—intenance support is developed to shorten the diagno—sis time and improve the precision of fault identifica—tion for user，and provide samples for system designerto optimize their product，saving wiring cost，achie—ving distributed measuring，and providing a fuly sys—tenm description.

In the working site，the wireless data acquisitioncards are set up for each machine，to transfer real—time data to front end software by Wi—Fi and achievenmhi—machine—monitoring． A multiple sensors ap—proach has been adopted to colect data．Thereinto，displacement sensors collect the position signal of thehydraulic cylinder piston rod，pressure sensors collectthe pressure signal of the outlet of the hydraulic cyl-inder．The sensors are connected to wireless DAQboard． With the software developed based on Lab—VIEW ， seamless connection with hardware can beeasily achieved and some signal process methods areinvoked directly.

Remote diagnosis technology based on Internet isa combination of Internet communication and fault di—agnosis technology． The opological structure of sys—tem is shown in Fig．3，the system contains web usersand winform users which we treat as back end opera—tors belonging to suppliers，and at the same time，machine users is front end operators．These two typeoperators have diferent issues on data，that is，frontend operator focuses on data acquisition and featureextraction，and back end operator focuses on fauh di-agnosis and statistics.

M achlneLfse_rask andDataApplicalion DataServer ServcT固一；l=：=LabVJEW VI蜀 1专=： 一皇塞
=；；；} “ ；．-L ? jASPNET Winform W ebWebSen,Ice FunctionFig．3 Remote Fault DiagnosisAccording to above description，the back endoperator is designed as clients and browsers in C／B／Sstruc,ture．If they are authorized．web server can re—ceive the fault features of acquisition task sent fromfront end， after diagnosing and reasoning process，and then push the result back to front end and storein database． Users．containing Web ones antj Win.

form ones，can summarize the statistical data or chartand manage the system and database respectively.

3．4．Fault Features extractionUsualy，the signals of hydraulic system in con—struction nmchinery are non—stationary and noisy.

Wavelet analysis is a time—frequency analysis methodwhich can be applied to handle these signals effec—lively．It Cal f0CUS on details of the signal with auto—matically adaptation to the requirements of time．fre.

quency signa1．In the previous study，we are mainlyfoCUS on hydraulic cylinder and found leakage affectthe stability in all frequency bands．and some faultfeatures can be use~I to indicate the leakage level[1 7]．A hydraulic cylinder leakage test—bed havebeen designed，and the faults are identified by theconnections between leakage and features[1 8].

覆ZHAO Xiuxu，et al：Remote Condition—based Maintenance Approach to HydraulicSystem of Construction Machinery 29As shown in Fig．4，fault features can be ana—lyzed with the data collected on it．The active and fit-ring CUI'VeS indicates the connections between leakagelevel and fault features，which are useful for fault di·agnosis and prognosticaton.

Leakagerate／(L·min )3．5．FM EA knowledge base12O100萼s。

磐 60： 4o2o差 oUsing WA WP Decomposition VI in LabVIEW ，signals can be processed，and the fault features areextracted．Then，Displacement Analysis VI and Pres：sure Analysis VI can be established according to a—bove procedures.

Leakagerate／(L·min )Fig．4 Fault Diagnosis by Wavelet PacketIn this study．a FMEA knowledge base is estab—lished in SQL Server database to support the fault di—agnose and system maintenance.

Generally，faults are detected as system fauh orpart fault in hierarchy structure mechanica1．electricalequipment[20]．Thus，they can be divided to de。

scribe structure and failure separately and then con-nected by a relation table．The system is analyzed byhardware method to establish production structuremodel and failure mode1．In the structure mode1．itnot only expresses the mix and function．but also in—dicates the fault transmission implicitly by iterativerelations．The iterative relation should be containedtogether with information of single fault．Appling twomode1．description and relation of faults are represen—ted completely and accurately.

As pre．mentioned．there are interactions in dif-ferent failure modes，for example，part failure modecan lead system failure mode or other part failuremode．This coupling relation should be analyzed andstored in FMEA base．the relation is shown in Fig．5.

The relation contains indirect transmission which ac—cords to the production structure，and direct trans—mission which express by FMEA records.

4．Case studyAccording to aforementioned structure and meth—od，a remote diagnose and maintenance system forhydraulic cylinder is developed． The back end sys-tern achieved the functions of diagnose and mainte-nance based on ASP．NET，connecting to the front2 702 652 602 552 502 452 402 352 302 250 5Leakagerate／(L·rainend software easily and safely．As shown in Fig．10，the front end software is developed on LabVIEW ，re-alizing real—time data acquiring， primary diagnosewith threshold and interface to the back.

FunctionRelationEf_ectRelation l Maintenancomponen Relation ailure M oRelationStructure ModelRelation l l RelationDetection]l !Failure ModelFig．5 Structure Model and Failure ModeIThe system has been run on our test-bed to col-lect analytical data and test reliability． In applica—tion．the DAQ board is NI 9201．which is 8．channelanalog input at a maximum aggregate rate of 500 kS／s，the data is sent via Wi-Fi；A QH一400 gratingsensor is setted to measure the position of plunger，and a AK4a pressure sensor is setted to measure thepressure of rod end：Maximum travel of cylinder is300 mm，input pressure of system is up to 7 MPa.

We foCUS on interna1 leakage fault for its dificulty onobserving，especially minor and mediem leakage.

If user want to connect with server． a taskshould be created with an auto．generated number。

then choose or add the details testing parameters，such as machine type，cylinder type and oil type，and then receive the corresponding threshold if theconditions are matching.

After monitoring data and calculating features，d^垦 8 壶 8 S 口苗 未 、^一．呻g一．∞．I 0 g置4J0兽3oI30 Hydromechatronics Engineeringthe software would send judgment request with taskinformation including ID，parameters and features toWeb service which is running in the industry．to calldiagnosis and maintenance service．The service rec.

ognizes the fault by BPN and corresponding sampleselected by testing parameters and analyzes the deci—sion and maintenance measurements for the fault.

while the task inform ation will be stored in database.

and some knowledge will be updated such as occur.

rence in FMEA base．After the called service runningout，the result will be pushed to front end softwareand show tn us rs.

5．ConclusionsDue to the complicated structure of hydraulicsystem，there are many difficulties on its fault diag—nosis in the working site． In this study．a remotecondition—based maintenance approach is presented，an application system with FMEA knowledge base hasbeen developed for construction machinery hydraulicsystem．The front end of this system is based on Lab.

VIEW and aims to colect the operation data of cylin.

der with multi．sensor and extract fault features in thework site：the backend is based on ASP．NET andcan provide service to identify the fauh diagnosis andmake maintenance decision with the inform ation fromfront end via Interuet：a FMEA knowledge base is es.

tablished to support above procedures．With the con-nection among diagnosing center， suppliers， andconstruction machine users，economic benefits can beobtained by reducing the troubleshooting time andcost of maintenanReferences：[1] Wang C J．China Hydraulic Pneumatic Industry the Cur—rent Situation and Countermeasures 201 1 Hydraulic Pneu—matic Sealing Industry Working Conference Report．Hy—draulics Pneumatics＆ Seals，2012．1：4—8.

[2] Venkatasubramanian V，Rengaswamy R，Kavuri S N，eta1．A review of process fault detection and diagnosis PartI：Quantitative model—based methods． Computers &Chemical Engineering，2003，27：293—311.

[3] Azadeh A，Ebrahimipour V，Bavar P．A fuzzy inferencesystem for pump failure diagnosis to improve maintenanceprocess：The case of a petrochemical industry，ExpertSystems with Applications，2010，37：627—639.

[4] Dong M，He David．A segmental hidden semi—Markovmodel (HSMM)一based diagnostics and prognosticsframework and methodology． Mechanical Systems andSignal Processing，2007，21：2248—2266.

[5] Jiang W L．，Wu S Q，Liu S Y．Exponentially WeightedDynamic Kernel Principal Component Analysis Algorithmand Its Application in Fault Diagnosis．Journal of Me—chanical Engineering，2011，47(3)：63—68.

[6] Angeli C．Online expert systems for fault diagnosis intechnical processes． Expert Systems，2008，25：1 15— 132.

[7] Zhan C c．Reliability and Diagnostics on Hydraulic Sys—tem．Beijing：Metallurgical Industry Press，2009.

[8] Liu X L，Chen G M，Li F X，Zhang Q，Miao C C.

Fault Bayesian Networks and Its Application to HydraulicCylinder Creeping． Chinese Hydraulics＆ Pneumatics.

2012(8)：108—111.

[9] Goharrizi A Y，Sepehri N．A Wavelet—Based Approachto Internal Seal Damage Diagnosis in Hydraulic Actua-tors．IEEE Transactions on Industrial Electronics，2010，57：1755—1763.

[10]Goharrizi A Y，Sepehri N．Internal leakage detection inhydraulic actuators using empircal mode decompositionand Hilbert spectrum．IEEE Transactions on Instrumen—tation and Measurement，2012，61(2)：368—378.

[11]Andrew K S Jardine，Darning Lin，Dragan Banjevic．Areview on machinery diagnostics and prognostics imple—menting condition．based maintenance． Mechanical Sys—tems and Signal Processing，2006，20：1483—1510.

I 12 l Xia L H，Jia X S，Xu Y．Reasonable Choice and Deci—sion—making Flow of Equipment Maintenance Policy，JOURNAL OF GUN LAUNCH & CONTROL．2006.

(4)：63—68.

[13]Azadeh A，Ebrahimipour V，Bavar P．A Pump FMEAApproach to Improve Reliability Centered MaintenanceProcedure：Th e Case of Centrifugal Pumps in OnshoreIndustry．Proceedings of the 6th WSEAS InternationalConference on FLUID MECHANICS， FLUIDS ’09，2009.

[14]He Z，Zhao Y，Zou F．Study and Application of the In—tegration of QFD and FMEA in Product Design Phase.

Modular Machine Tool& Automatic Manufacturing Tech—nique，2007(4)：97—100.

[15]Zhao X X，Bai X L．The application of FMEA method inthe risk management of medical device during the lifecy—cle．2nd International Conference on e-Business and In—formation System Security，2010.

[16]Cui W B，Wu G T，Sun P T，Zhang Y Q．Ship safetyassessment based on FMEA and fuzzy comprehensive e—valuation methods．Journal of Harbin Engineering Uni—versity，2007，28(3)：263—267，276.

[17]Li R，Research on Method of Engineering Machinery Hy—draulic Cylinder’s Fauh Feature Extraction Based onWavelet Analysis，Wuhan：Wuhan University of Tech—nology，2012.

[18]Li Y，Wang L，Jiang J H．Experimental Research on In—ternal and External Leakage of Construction MachineryHydraulic Cylinder．Machine Tool& Hydraulics．2012，4O(12)：5—9． (Continued from 35 page)ZHANG Jinxi，et al：The Conception and Kinematic Simulation of a CamMechanism Continuously Variable Transmission 35performance of new—type CVT could be further betteron this basis，it might provide a new breakthrough forthe study of mechanical stepless transmissions.

References：[1] Tian J Y．The structure principle and design method ofvehicle automatic transmission[M]．Beijing：Beijing U—niversity Press，2009.

[2] Yang H R，The application and development trend ofmechanical stepless transmission[M]．[S．1．]：Academ-一 种新型无级变速器的构想与运动仿真张晋西 ，谭 芬，蔡 维重庆理工大学，重庆 40054[3][4][5][6]ic Communication．20U9.

Sun H，Chen Z M，Ge W J．Mechanical orinciple(rI1}leseventh edition)[M]．Beijing：Higher Education Press，20o6.

Shen Y S．The course of mechanical principle『M].

Beijing：Tsinghua University Press，2004.

Du L．Lin L．The structure parameter design of pulsingmechanical stepless transmissions[J]．Journal of YuzhouUniversity，2002(1)：21—24.

Zhang W H，Xie L Y，Cheng N S．卟 e experimental studyof metal belt CVT’S friction coeficient and transmissioneficiency[J]．Mechanical Design，2(I)6：41—43.

摘要：提出了一种新型机械式无级变速器的构想。该变速器通过采用空间锥形凸轮机构和摆杆与锥形凸轮接触部位的连续轴向变化来实现无级变速。改变锥形凸轮的长度或锥度能得到不同的变速比。用 ADAMS对该模型进行运动仿真和分析比较，获得 了较理想的匀速运动效果。这一构想可为无级变速器的研究提供一条新的思路。

关键词：无级变速器；锥形凸轮；变速比；运动仿真中图分类号：TH12；TP391．9 ‘(Continued on 30 page)[19]Wang L H，Guo J J，Wu B，Yang S Z．Correlative Di—mension Analysis on Dynamic Properties of Moving Hy—draulic Cylinder． Machine Tool& Hydraulics，2006，(9)：99—102，2006.

工程机械液压系统的远程视情维护方案赵秀栩 ，胡结曼 ，姜继海 ，张清润1．武汉理工大学 机电工程学院，武汉 430070；2．哈尔滨工业大学 机电工程学院，哈尔滨 150001[20]Zheng X J，Yang S Z，Shi H M，Zhou A F．The Hierar·chy Diagnostic Model for Diagnostic Expert Systems.

Journal of Huazhong University of Science and Technolo—gY，1987，15(2)：9—14.

摘要：液压系统在大功率和 自动化工程机械 中得到 了广泛应用。但在工程实际中，由于结构复杂使得液压系统的故障诊断非常困难。特别是在现场施工过程中，如果液压系统的故障得不到及时处理，会导致设备长时间的停工而严重影响到施工效率。为此，提 出了一个远程视情维护解决方案，通过对失效模式与影响分析(FMEA)方法的扩展和工程机械液压系统 FMEA知识库的开发 ，实现专家知识与设备工作过程 中实时采集数据的集成。介绍了系统的整体结构设计、数据采集、故障特征提取以及 FMEA知识库的建立等关键技术，并给出了一个工程实际案例。

关键词：工程机械 ；液压 系统；远程视情维护；扩展 FMEA分析；FMEA知识库中图分类号：TH137