OPTIMIZATION OF MAINTENANCE OF GENSETS USING EXPERT SYSTEMS FOR FAULT DIAGNOSIS

Size: px

Start display at page:

Download "OPTIMIZATION OF MAINTENANCE OF GENSETS USING EXPERT SYSTEMS FOR FAULT DIAGNOSIS"

Kathleen Green
5 years ago
Views:

1 JT Volume Issue 1, March 2017 Te of article OPTIMIZATION OF MAINTNANC OF GNSTS USING XPRT SYSTMS FOR FAULT DIAGNOSIS OPTIMIZACIJA VZDRŽVANJA GNRATORSKI NOT Z UPORABO KSPRTNGA SISTMA ZA DIAGNOSTICIRANJ NAPAK Željko ederić R, Dejan Barešić 1, Venco Ćorluka 2 Kewords: Baes theorem, fault diagnosis, genset, eert sstem Abstract The usage of eert sstems simlifies the determination of the technological condition of the genset, i.e. the fault diagnosis. This results in the otimization of maintenance in technical workshos, which articularl comes into la in militar sstems in which a great number of technical resources must be maintained. The eert sstem is based on the Knowledge Base, obtained b researching and collecting data from rofessionals who have ears of eerience in the maintenance of electric generators. The Knowledge Base is made to list the most common faults first; then the hotheses are associated with a certain robabilit. The rocessing of collected information is carried out according to Baes method, which is based on the classical robabilit theor. In order to decide uon a failure haening within the contet of certain facts, a calculation of true values is carried out, which will hel in develoing an ideal diagnostic rogram during further research. R Corresonding author: Dr. Željko ederić, Associate Professor, J.J. Strossmaer Universit of Osijek, Facult of lectrical ngineering, Comuter Science and Information Technolog, Kneza Trimira 2B, Osijek, Croatia, -mail address: zhederic@etfos.hr 1 Universit of Zagreb, Croatian Militar Academ Dr. Franjo Tuđman, Zagreb, Croatia, dejan.baresic@suradnik.unizg.hr 2 J.J. Strossmaer Universit of Osijek, Facult of lectrical ngineering, Comuter Science and Information Technolog, Osijek, Croatia, venco.corluka@etfos.hr JT 3

2 Željko ederić, Dejan Barešić, Venco Ćorluka JT Vol Issue 1 Povzetek Uoraba eksertnega sistema za diagnosticiranje naak v generatorskih enotah omogoča oenostavitev določitve stanja osamezne enote in morebitno diagnostiko naak. Rezultati otimizacije vzdrževanja v servisnih delavnicah so izredno omembni v vojaških sistemih, kjer je otrebno ohranjati nenehno večje število narav v stanju riravljenosti. ksertni sistem temelji na odatkovni bazi znanj, ki je ridobljena reko zbiranja odatkov iz večjega števila oravljenih reizkusov in izkušenj strokovnjakov z večletnimi izkušnjami s odročja diagnostike generatorskih enot. Predlagan eksertni sistem ri vzdrževanju narav izvede regled seznama najogostejših naak in samodejno izvede diagnostiko z določeno verjetnostjo. Obdelava zbranih odatkov oteka o Baes-ovi metodi, ki temelji na klasični teoriji verjetnosti. Za določitev možnosti naake na odlagi zbranih dejstev se izvede izračun dejanskih vrednosti, kar omogoča nadgradnjo eksertnega sistema z nadaljnjimi raziskavami. 1 INTRODUCTION This aer shows the otimization of the maintenance sstem of gensets in technical workshos. It is articularl suitable for use in the armed forces, where a great number of mobile gensets is resent. The are necessar in case of greater damage and failures in the electrical sstem, which ma occur in the event of natural disasters and war. Therefore, their reliabilit and availabilit are vital, in both civilian and the militar structures. Due to that, regular maintenance is being carried out, which requires certain financial resources. In addition to such resources, an imortant issue is also the involvement of secialized staff in the maintenance rocess. Since a large amount of resources is involved, ever simlification of the maintenance rocedure, as small as it ma be, is essential in saving financial resources and engagement of material, technical and human resources. In comle sstems, such as gensets maintenance rocesses, the identification of the technical conditions of equiment is the most time consuming, i.e. fault diagnosis. To formulate a diagnostic algorithm of the equiment states, a sstematic stud of the behaviour in good working condition and during a failure has been carried out. Collecting data on the failures and their indicators was derived from three ke sources: - technical documentation with the regulations on the qualit of the roduct, - documents about receiving resources for reair in the ast five ears, from which the reasons of arrival can be seen, as well as the methods carried out in order to correct the deficiencies, - b conducting research among the emloees with ears of eerience in the maintenance of gensets. B consolidating the collected data, a Knowledge base has been formed, shown artiall in Table 3. The data were rocessed according to Baes' theorem, and true values were obtained, which will hel in further research to develo an ideal rogram and secial instructions for searching and removing irregularities in the assembled form. This rovides a comrehensive understanding of the resources technical condition before dismantling and achieves the greatest effect in reducing the duration of a reair. In addition to that, the described rocedures enable the determination of the scoe and comleit of reair and the necessar tools, equiment and sare arts in advance. 36 JT

3 Otimization of Maintenance of Gensets Using ert Sstems for Fault Diagnosis 2 GNST MAINTNANC PROCSS The maintenance rocess is a comlete series of organizational and technical measures eloits to establish the functionalit of fault resources, i.e. to reair corruted technical features after the resources usabilit has assed or after the aearance of random defects. The maintenance rocess consists of a series of different stes connected with the reair, manufacturing of arts, their control, storage and transort. This rocess also includes the develoment of technical documentation, lanning, standardization, the sul of materials, sare tools and accessories as well as the refinement and develoment of technological rocesses and diagnostic methods, [1]. Due to the large number of technical resources in militar sstems, methods are continuousl elaborated, which contributes to the otimization of maintenance and the work roductivit of technical workshos. The imortant oint in otimization is the shortening of reair times, which can be significantl affected b the imrovement of technological rocesses and diagnostic methods. For now, the greatest effect is achieved b an elaborated rincile of reairing a set Figure 1, in which eert sstems for failure diagnosis can be alied, which significantl reduces the overall reair time. Receit of the resource for reair Prearation Prirema zafor oravak the reair Diagnostics of the genset it the assembled state Disassembling the genset into comonents n-1 n Reserve of the assembl in the reair rocess Fault Correct Diagnostics Prirema ofza theoravak comonents Prearation of roer comonents for assembl Assembling the genset Testing and adjusment Figure 1: Schematic of the technological rocess on the rincile of reairing a set On receit of the resource for reair, a rotocol for the descrition of the status is made. Net, the reair and diagnostics are reared in the assembled state, which reresents a comle roblem, eseciall in comle sstems such as gensets. Their comleit lies in that the are comosed of several technicall different subsstems: diesel engine, snchronous generator with ecitation, control rotection, and connecting and other circuits. This means that for their JT 37

4 Željko ederić, Dejan Barešić, Venco Ćorluka JT Vol Issue 1 diagnostics the criteria which mark the ermissible and imermissible states of each mentioned subsstems must be known. Therefore, in order to reduce the time to detect fault arts, it is necessar to develo an otimal rogram for fault diagnostic and on the basis of that rogram to reare a summar of secial instructions for the search for fault arts. The use of secial instructions for troubleshooting turns random search stes into re-lanned sequences of actions. It is usuall easier to locate defective arts when checkoints are determined, which are used to check for ermitted values of technical requirements. The main tasks of diagnostics in the assembled state are: - check of the working abilit, - identification of fault circuits, - determination of the circuits that need no reair. The diagnostics in the assembled state is carried out according to the aroriate instructions for each te of genset. The data on irregularities with which the resource is sent and the data on the time and mode of eloitation hel to reair diagnose the fault. The diagnosis is carried out in a secific order, which ensures the comleteness of the eected checks and ecludes the reetition of certain other checks. A roerl conducted diagnostics rovides a comrehensive insight into the state of the resource and enables: - determination of the scoe and comleit of the reair, - determination of necessar tools, equiment and sare arts, - determination of emloees with certain qualifications electricians, machinists, mechanics, - definition of the time required for the reair. After the diagnostics and detailed insight into the state of the resource, the disassembling continues with the net, so far mostl well elaborated ste according to the schematic of the technological rocess from Figure 1. 3 XPRT SYSTM FOR DIAGNOSIS OF T GNST STAT 3.1 Knowledge base The eert sstem for fault diagnosis is based on a reliable Knowledge base, in which the diagnostic features are shown. These data were obtained b studing the standard and the manufacturer's technical documentation, [2, 3, 4], worksho roduction orders from the revious five ears and the imlementation of surves of emloees who have ears of eerience in the maintenance of gensets. The eert sstem is designed to ensure the raid determination of genset condition correct, indication of failure-needed maintenance, faultneeded service. This is an additional tool that connects users and maintenance technicians. Ticall, these are two technical levels and are not at the same location neither under the same management. The user a soldier in the described case with a basic knowledge of oeration during the dail work notes "disturbances" in the work of the genset, which are 38 JT

5 Otimization of Maintenance of Gensets Using ert Sstems for Fault Diagnosis categorized in the grou of asil Detectable Failures DF. These are linguistic variables with distinctl subjective characters; the are shown in Table 1. Table 1: amle of linguistic variables that describe the grou of asil Detectable Failures with distinctl subjective character The engine can't start or has difficulties starting The engine stos a short time after the start Low engine erformance Low oil ressure The engine rotational seed dros fast The engine overheats The engine throws out thick black smoke Low generator voltage idling at nominal seed The generator roduces no voltage The main circuit breaker can't be turned on Checking the static frequenc characteristics - doesn't fulfil the conditions Based on the detection of disorders from the DF grou, the overseer ert Sstem focuses further rocedures toward maintenance technicians, who reresent a higher level of technical knowledge of the genset a technician in the Reair Institute in the described case. In this rocess, linguistic variables that describe disorders categorized in grou of ard Detectable Failures DF shown in Table 2 and associated robabilit of failure from the knowledge base of an eert sstem have been used. The imortance of resolving the roblems of large logistics sstems with the use of eert sstems that conventional linguistic variables shown in Tables 1 and Tables 2 convert into digital data should be emhasized. This enables the use of software tools for conducting the control, monitoring and lanning of logistics in sstems that satiall and hierarchicall belong to different sub-sstems of management. The general objective is to imrove the management of the overall sstem militar sstems in the described case. Table 2: amle of linguistic variables that describe a grou of ard Detectable Failures without subjective character Damaged fuel sul ieline Corroded or loosened connections on the batter Discharged batter Fault on the electric starter Dirt fuel refining sstem Dirt air refining sstem Fault injectors Lack of oil in the motor Dirt oil refining sstem Fault oil um Rotational seed regulator fault Low engine comression Loose contact on the voltage regulator The generator is demagnetized Fault mode selector JT 39

6 Željko ederić, Dejan Barešić, Venco Ćorluka JT Vol Issue 1 In the Knowledge Base artiall shown in Table 3, the most common faults are shown first, then the hotheses are associated with a certain robabilit. The table is just a segment of the Knowledge Base, which will be further eanded during future research. Table 3: Knowledge Base with entered fault robabilities 3. 2 Baes' theorem imlementation The rocessing of collected information is carried out according to Baes' method, which is based on the classical robabilit theor. With the hel of the conditional robabilit, the robabilit of an event is determined; if it is known that the event has occurred. vent is considered a fact that is easil obtained, whereas reresents the robabilit of the 40 JT

7 JT 41 Otimization of Maintenance of Gensets Using ert Sstems for Fault Diagnosis state of the genset, i.e. the robabilit of a certain failure which needs to be defined. This robabilit is called conditional robabilit :. 3.1 If two events are indeendent, then: According to the definition, the revolution, i.e. robabilit of the event under the condition that has occurred is: From 3. Because of the commutativit: 3.6 Including 1.6 in 1.1, we obtain the simlest form of Baes' theorem: 3.7 B eanding the members of the conditional robabilit, we get Baes' theorem []: ~ ~ 3.8 From Baes' theorem, we obtain the Baesian method in roduction sstems: If is true, the conclusion is with the robabilit. With the hel of Baes' theorem, we can conclude the robabilit of. B interreting this theorem with the formula 1.7: of the theorem marks an evidence or a fact of the theorem marks a hothesis 3.9 ~ ~ 3.10

8 Željko ederić, Dejan Barešić, Venco Ćorluka JT Vol Issue 1 In order to decide uon the failure haening with certain facts, a calculation of real values is carried out, as it is elaborated in [6], Chater 7. For the uroses of determination of correct failure robabilities series of data were collected amle case In order to clarif functioning of the Baes theorem, the imlementation into ert Sstems for fault diagnosis rocedures for one common event is resented. Using genset for the energ sul of sensitive communications equiment requires frequenc control of ower sul b using built-in indicators frequencies or ortable devices for measuring frequencies. The user soldier on the ground noted that the frequenc of ower sul is outside the ermitted areas: DF Checking the static frequenc characteristics does not fulfil the conditions. Based on that, the user logs in reort into ert Sstem, which in addition to the descrition of failure also records the time, date, location of failure, and characteristics of genset reviousl classified information in knowledge base. Case: Checking the static frequenc characteristics does not fulfil the conditions ert Sstem based on knowledge base information of robabilit Table 3 shaded row and column conduct following calculation: dirt fuel refining sstem thetest of thestaticfrequenc characteristics doesn't fulfil the conditions dirt fuel refining sstem 0.8 thetest of thestaticfrequenc characteristics doesn't fulfil theconditions not dirt fuel refining sstem 0.2 thetest of thestaticfrequenc characteristics doesn' t fulfil theconditions dirt fuel refining sstem if it is obviousthat thetest of thestatic frequenc characteristic doesn't fulfil theconditions 42 JT

9 Otimization of Maintenance of Gensets Using ert Sstems for Fault Diagnosis dirt fuel refining sstemwith thefact that it fulfils the staticfrequenc characteristics Comaring and ields the following conclusion: The fact that the test of the static frequenc characteristics does not fulfil the conditions increases the robabilit of a dirt fuel refining sstem b aroimatel 3.1 times. Table 4 shows the true values for the other cases when the test of the static frequenc characteristics does not fulfil the conditions. This means, b reading Table 4 and comaring the values to those in Table 3, we can, in addition to the abovementioned case, conclude the following: The fact that the test of the static frequenc characteristics does not fulfil the conditions increases the robabilit: - of a damaged fuel sul ieline b 2. times, - of fault injectors b 1. times, - of a rotational seed regulator fault b.2 times, - of a low engine comression b 3. times, - of a loose contact on the voltage regulator b 1. times. The same rocedure must be used to treat the easil observable facts mentioned in Table 1 and in the first row of Table 3. The given results are, after quick and eas rocessing b a comuter, the foundation for develoing an ideal diagnostic rograms. From Table 3 and the resented eamle case, it is evident that the static calculation checking the frequenc characteristics is not associated with all failures listed in Table 2 and in the left column of Table 3. This is consistent with the fact that, for eamle, failure of the starter has no effect on the stabilit of the frequenc of the generator ower sul. Some failures have higher and some lower correlation with asil Detectable Failures, which are listed in Table 1 first line in Table 3. For eamle: checking static voltage characteristics is not associated with failure of the starter, whether associated with dirt fuel because at nominal loads it can lead to instabilit of diesel engine oeration or the major collase of genset drive rotational seed. This directl, to a large etent, affects the characteristics of genset frequenc, and to a small etent, affects the stabilit of voltage, as shown in Table 3. The data were confirmed during measurements imlemented at genset insection, [4]. It was determined that a voltage regulator retains a constant voltage to a certain degree of rotational seed collase, but voltage deviation does noticeabl occur at higher and faster changes in rotational seed. JT 43

10 Željko ederić, Dejan Barešić, Venco Ćorluka JT Vol Issue 1 Table 4: Calculation of true values for all cases when the test of the static frequenc characteristics doesn't fulfil the conditions Checking the static frequenc characteristics - doesn't fulfil the conditions ~ ~ ~ / 1= = = = = = = = = = = = = = = CONCLUSION Due to the large number of technical resources in militar sstems, methods are continuousl elaborated, which contributes to the otimization of maintenance. The greatest effect is achieved b the rincile of reairing a set, for which the failure diagnosis in the assembled state has a significant role. Therefore, the stud of the behaviour of gensets in different states and the develoment of a Knowledge Base have begun, in which the hotheses are associated with a certain robabilit. With the hel of Baes' theorem true values were calculated, used together with certain facts, to correct the robabilities listed in the Knowledge Base. The Knowledge Base, containing the elaborated Baes' theorem and the gathered true values, is the basis for the continuation of the research and develoment of ideal diagnostic rograms and secial instructions for diagnosing and troubleshooting. 44 JT

11 Otimization of Maintenance of Gensets Using ert Sstems for Fault Diagnosis References [1] M. Jakočić: Održavanje naoružanja, Zaovjedništvo rvatske konene vojske, Zagreb, 2006 [2] ISO through International Organization for Standardization. Genėve [3] Uute za rukovanje i održavanje s izvješćem o isitivanju elektroagregata P-B40.R1 KONČAR-RS, Rijeka 199 [4] Proisi o kvaliteti roizvoda, lektroagregati izmjenične i jednosmjerne struje, snage do 1 MW, Rade Končar, Rijeka, 1984 [] htt:// [6] F. Jović: ert Sstems in Process Control, Chaman & all, London, 1992 JT 4

MaxxForce TM 11 High Pressure and Oil. Engine Systems

2004 2006 2007 MaxxForce TM 11 High Pressure and Oil System MaxxForce Diagnostics 13 Engine Systems A N AV I S TA R C O M PA N Y Study Study Guide Guide TMT-120717 Study Guide MaxxForce High TM Pressure