Umted States Patent [191
`Schricker et al.
`
`IIlIII 1111 11 um IIIIIIIIIILICIIIISIIIIIL |I|l11||||||||||||||n|| 11 Ill
`[11] Patent Number:
`5,561,610
`[45] Date of Patent:
`Oct. 1, 1996
`
`[54] METHOD AND APPARATUS FOR
`
`4,583,176
`
`4/1986 Yamato et al. .................. .. 364/431.11
`
`INDICATING A FAULT CONDITION
`
`[75] Inventors: David R. Schricker, Dunlap; Satish M. .
`
`Sheny’ East Peona’ both ofm'
`m,
`
`[a
`[21] Appl. No.: 268,693
`_
`[22] F1led:
`
`Jun. 30, 1994
`
`6
`[51] Int. Cl. ................................................... .. G08B 23/00
`[52] US. Cl. ..................................... .. 364l551.01; 340/679
`[58] Field of Search ................................... .. 340/679, 680;
`364/550, 551.01, 55192
`
`[56]
`
`References Cited
`
`U'S- PATENT DOCUMENTS
`Re_ 31,750 11/1984 Morrow
`3,257,652
`6/1966 Foster .......... ..
`3,298,010
`1/1967 Dubosq et al.
`3,882,305
`5/1975 Johnstone ........ ..
`3,946,364
`3/
`COdOHlO C11 31. ..
`
`364/550 X
`340/68OX
`340/679 X
`235/151.11
`340/1725
`
`4,093,939
`4,184,205
`4,542,461
`
`. . . .. 340/52
`6/1978 Mitchell . . . . .
`364/550 X
`1/1980 Morrow .......... ..
`9/1985 Eldridge et al. ...................... .. 364/424
`
`4,749,987
`
`6/1988 IShii . . . . . . . . . . . . .
`
`. . . . . .. 340/587
`
`364/551.01
`
`4,787,053 11/1988 Moore
`,
`,
`
`glatsushim erruyer .... ..
`4,967,381 10/1990 Lane 61111.
`364l551.01
`11212811222122.1119 .......... "222/22;
`5,303,163
`4/1994 Ebaugh et a1.
`364/550
`
`Primary Examiner—Edward R. Cosimano
`Aztomey, Agent, or Firm—-Steven R. Janda; Mario J. Donato
`
`[57]
`
`ABSTRACT
`
`An apparatus is provided for indicating a fault condition in
`a machine having a plurality of parameters each having
`levels being dependent upon machine performance. The
`apparatus includes a sensor adapted to produce an electrical
`signal in response to the level of one of the of parameters;
`and a processor for identifying a trend in the parameter level
`in response to the electrical signal, calculating the duration
`and 81999 of the trend; and determining Whether =1 ?rst
`warning threshold is exceeded in response to the duration
`and slope of the trend, the ?rst warning threshold being a
`function of duration and slopa
`
`16 Claims, 10 Drawing Sheets
`
`104.11.1
`
`I CALCULATE DISTANCE OF LAST POINT 1X10 4 11 2
`
`104.1113]
`
`OBTAIN WEIGHTING
`FACTOR BASED ON
`DISTANCE
`
`DETERMIN
`DURATIO
`
`SLOPE AND
`OF TREND
`
`OBTAIN SIGNIFICANCE
`FACTOR BASED ON
`SLOPE AND DURATION
`
`104.11.14
`
`j“ CALCULATE WARNING PARAMETER 1
`104.115
`
`WARNING
`LEVEL GREATER
`THAN WARNING
`
`STORE FAULT CONDITION
`
`104.11.11
`
`RETURN
`
`Page 1 of 16
`
`Hyundai Exhibit 1005
`
`

`
`US. Patent
`
`Oct. 1, 1996
`
`Sheet 1 0f 10
`
`5,561,610
`
`Slit.
`
`0 O0
`
`285m
`
`%
`
`:28 Li! @2058:
`
`
`520:
`
`a ”
`
`l | L
`
`_
`
`Page 2 of 16
`
`

`
`U.S. Patent
`
`Oct. 1, 1996
`
`Sheet 2 of 10
`
`5,561,610
`
`row
`
`lmlmh
`
`ZSSmImE
`
`Q 0333 420:5?
`
`E5: 5355 £3“
`
`
`35% 5! 552m |\+
`
`
`
`i 023? #2952
`
`= a a a
`
`a a = =
`
`Ella .L @ kW
`WV
`
`Page 3 of 16
`
`

`
`US. Patent
`
`Oct. 1, 1996
`
`Sheet 3 of 10
`
`5,561,610
`
`212
`
`READ. MACHINE
`214"\ PARAMETERS
`
`DEPENDENCIES
`SATII'SPFIED
`
`STORE
`PARAMETER \"ZZO
`
`COUNTS BEEN
`‘REAgHED
`
`222
`
`DETERMINE AND
`-<—- STORE TREND \224
`DATA
`
`Page 4 of 16
`
`

`
`U.S. Patent
`
`Oct. 1, 1996
`
`Sheet 4 of 10
`
`5,561,610
`
`
`
`
`
`BEEExpo:.._o5252mozmsomm
`
`ooooo%ooooooO00
`
`oMo%oowe
`
`0oo
`
`3fT'lV/\ CINEIELL
`
`Page 5 of 16
`
`Page 5 of 16
`
`

`
`U.S. Patent
`
`Oct. 1,1996
`
`Sheet 5 of 10
`
`5,561,610
`
`lmlmlm
`
`ZOEEDQ
`
`m” /// //;
`llm-oll
`lmzo
`$0.0
`
`m. TI
`
`BdO'IS
`
`Page 6 of 16
`
`

`
`US. Patent
`
`Oct. 1, 1996
`
`Sheet 6 of 10
`
`5,561,610
`
`Page 7 of 16
`
`

`
`US. Patent
`
`061. 1, 1996
`
`Sheet 7 0f 10
`
`5,561,610
`
`E15_Y_
`
`100
`
`INITIALIZE DEFAULT FIT
`PARAMETERS AND ZERO
`ACCUMULATORS
`
`OBTAIN OPTIONAL
`OVERRIDES FROM
`DEFAULTS
`
`102
`
`LOAD X,Y DATA
`PAIRS
`
`I
`I
`I
`I
`
`SEGMENT X,Y DATA
`/ WRITE OUT RESULTS A104
`
`CLOSE OUTPUT FILES A1 05
`
`106
`
`Page 8 of 16
`
`

`
`U.S. Patent
`
`0a. 1, 1996
`
`Sheet 8 of 10
`
`5,561,610
`
`ZERO X—-MINUS
`LINE
`J10“
`STATISTICS
`
`‘21-’- E
`__.I.I: _ _.
`
`104.2
`SET X-MINUS
`sTART TO f
`BEGINNING OF
`X,Y DATA
`
`104.3
`
`STORE LAST __1049
`X-MINUS LINE
`'
`I
`WAIT UNTIL A
`NEW X,Y PAIR
`IS RECIEVED
`AND LoAD THE
`X,Y PAIR
`—104.10
`
`THERE MORE
`UNPROCESSED X,Y
`PAIRS THAN THE
`
`X—PLUSSZgVlNDOW
`I?
`
`YES
`ADD NEw X,Y
`PAIR To X-MINUS
`AND X-PLUS
`WINDOWS
`
`104'“
`
`DETERMINE
`IQTAHIQG
`
`104.4
`
`I
`
`INCREMENTALLY ADD
`NEW POINT TO
`STATISTICS AND
`CALCULATE NEW
`BEST FIT LINE
`
`I
`
`CALCULATE FIT OF
`X-PLUS ON X-MINUS_/‘104.6
`PROJECTION
`
`104.7
`
`104.8 »\ DEFINE X-MINUS
`AS NEW LINE
`
`Page 9 of 16
`
`

`
`U.S. Patent
`
`0a. 1, 1996
`
`Sheet 9 of 10
`
`5,561,610
`
`( START >—104.7.1
`
`FITSTATDENOM =
`XMINUSMSE + f104.7.2
`SLACKCONST *
`XMINUSNARTANCE
`+ BIAS SLACK
`
`104.7.3
`
`IS
`XPLUS.MSE
`4 ITSTATDT-éNOM > 0 m FITSTAT =
`K404.7.4
`
`104.7.7
`
`X—PLUS.QMSE > O
`
`RETURN .FIT
`FAILS (1)
`
`FITSTAT >
`FIT TOEERANCE
`
`104.7.5
`
`RETURN FIT
`PASSES (0)
`
`104.7,8
`
`Page 10 of 16
`
`

`
`U.S. Patent
`
`Oct. 1, 1996
`
`Sheet 10 of 10
`
`5,561,610
`
`______
`
`104.11.1
`
`(?g/LITE]
`.._-_-I:_I:_l _
`CALCULATE DISTANCE OF LAST POINT
`I
`DEIRIIIISEI ERIE/11K
`I
`ORAIIRS'SIEIRIR
`SLOPE AND DURATION X1044“
`I
`
`104.11.13
`T
`
`OBTAIN WEIGHTING
`FACTOR BASED ON
`
`104‘11‘2
`
`.
`
`1044-11-3
`
`COMBINE WEIGHTING FACTOR
`1Q4_11_14f AND SIGNIFICANCE FACTOR
`
`104115
`
`CALCULATE WARNING PARAMETER I
`
`LEVEL GREATER
`THAN CAUTION
`THRE$PHOLD
`
`RESET
`WARNING
`
`\104.11.7
`
`WARNING
`LEVEL GREATER
`THAN WARNING
`THRES'PHOLD
`
`INDICATE WARNING
`
`104.11 a
`
`104.11 a
`
`104.11.10
`
`EsToRE FAULT CONDITION |
`k-IOILILII
`@
`
`Page 11 of 16
`
`

`
`5,561,610
`
`1
`NIETHOD AND APPARATUS FOR
`INDICATING A FAULT CONDITION
`
`TECHNICAL FIELD
`
`The invention relates generally to devices for diagnosing
`and indicating fault conditions, and more particularly, to a
`method and apparatus for diagnosing and indicating a fault
`condition in response to the trend in the level of a machine
`parameter.
`
`BACKGROUND ART
`
`For service and diagnostic purposes, machines are some
`times equipped with sensors for measuring operating con
`ditions such as engine RPM, oil pressure, water temperature,
`boost pressure, oil contamination, electric motor current,
`hydraulic pressure, system voltage, and the like. In some
`cases, storage devices are provided to compile a data base
`for later evaluation of machine performance and to aid in
`diagnosis. Service personnel examine the accrued data to get
`a better picture of the causes of the failure or to aid in
`diagnosis. Similarly, service personnel evaluate the stored
`data to predict future failures and to correct any problems
`before total component failure.
`In addition, these stored parameters may be examined by
`service or supervisory personnel to evaluate machine and/or
`operator performance to ensure maximum productivity of
`the machine. These issues are particularly pertinent to over
`the-highway trucks and large work machines such as off
`highway mining trucks, hydraulic excavators, track-type
`tractors, wheel loaders, and the like. These machines rep
`resent large capital investments and are capable of substan
`tial productivity when operating. It is therefore important to
`predict failures so servicing can be scheduled during periods
`in which productivity will be less affected and so minor
`problems can be repaired before they lead to catastrophic
`failures.
`Similarly, it is sometimes advantageous to accumulate
`parameters only when the machine is in a particular oper
`ating .condition. This type of information is predominantly
`used during performance evaluation but may also be used in
`failure diagnosis and prognosis. For example, the length of
`time spent in a particular gear while the machine is loaded
`may be needed to evaluate machine performance.
`The present invention is directed to overcoming one or
`more of the problems set forth above.
`
`DISCLOSURE OF THE INVENTION
`
`In one aspect of the invention, an apparatus is provided
`for indicating a fault condition in a machine having a
`plurality of parameters each having levels being dependent
`upon machine performance. The apparatus includes a sensor
`adapted to produce an electrical signal in response to the
`level of one of the parameters; and a processor for identi
`fying a trend in the parameter level in response to the
`electrical signal, calculating the duration and slope of the
`trend; and determining whether a ?rst warning threshold is
`exceeded in response to the duration and slope of the trend,
`the ?rst warning threshold being a function of duration and
`slope.
`In a second aspect of the invention, a method of indicating
`a fault condition is provided. The method includes the steps
`of sensing a parameter having a level being dependent upon
`machine performance and responsively producing an elec
`trical signal, identifying a trend in the parameter level,
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`45
`
`50
`
`55
`
`65
`
`2
`calculating the duration and slope of the trend, and deter
`mining whether a ?rst warning threshold is exceeded in
`response to the duration and slope of the trend, the ?rst
`warning threshold being a function of duration and slope.
`The invention also includes other features and advantages
`which will become apparent from a more detailed study of
`the drawings and speci?cation.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`For a better understanding of the present invention, ref
`erence may be made to the accompanying drawings, in
`which:
`FIG. 1 is a high level diagrammatic illustration of an
`embodiment of the invention;
`FIG. 2 illustrates a plurality of connections to an elec
`tronic module in one embodiment of the invention;
`FIG. 3 illustrates an algorithm performed by the elec
`tronic module in one embodiment of the invention;
`FIG. 4 illustrates a plurality of sensor values and best-?t
`line segments;
`FIG. 5 illustrates warning levels as a function of slope and
`duration of trended data;
`FIG. 6 illustrates one instance of a group of data points
`being used to obtain a line segment;
`FIG. 7 illustrates a high level algorithm performed in one
`embodiment of the invention;
`FIG. 8 illustrates a portion of the algorithm of FIG. 7 in
`more detail;
`FIG. 9 illustrates a portion of the algorithm of FIG. 8 in
`more detail; and
`FIG. 10 illustrates a portion of the algorithm of FIG. 8 in
`more detail.
`
`BEST MODE FOR CARRYING OUT THE
`INVENTION
`
`Referring to FIG. 1, a machine prognostic system is
`shown generally by the number 10 and is a data acquisition,
`analysis, storage, and display system for a work machine 12.
`Employing a complement of on-board and off-board hard
`ware and software, the machine prognostic system 10 moni
`tors and derives machine component information and ana
`lyzes the resulting data to indicate impending component or
`system failures.
`FIG. 1 illustrates a variety of potential communication
`systems 14 that may be used to transfer data from the work
`machine 12 to a central computer system 16 for analysis. In
`the preferred embodiment, the data is transferred by the
`Qualcomm satellite system back to the central computer
`system 16. Alternatively, the data is transferred by a cellular
`telephone system or by storing data on a computer disk
`which is then mailed to the central computer site for analy
`srs.
`It should be understood that all aspects of the present
`invention could be located on-board the work machine 12
`thereby eliminating the need for a communication system
`14; however, the central computer system 16 allows an
`entire ?eet to be monitored at a central location.
`Subsets of the data are also transmitted to a display
`module (not shown) in the operator compartment of the
`work machine 12 for presentation to the operator in the form
`of gauges and warning messages. During normal operation,
`gauge values are displayed in the operator compartment.
`
`Page 12 of 16
`
`

`
`5,561,610
`
`3
`During out-of~spec conditions, alarms and warning/instruc
`tional messages are also displayed.
`In the preferred embodiment, sensed data is directly
`sampled by sensors 18 of a type well-known in the art for
`producing electrical signals in response to the level of
`operational parameters and includes pulse-width modulated
`sensor data, frequency-based data, ?ve volt analog sensor
`data, and switch data that has been effectively debounced.
`The sensors are connected to an electronic module 20 for
`delivery of the sensor signals.
`In the embodiment of FIGS. 1 and 2, the sensor signals are
`delivered to the electronic module 20 by either direct
`connection of analog sensors, connection by way of an
`RS485 link, or over a datalink governed by SAE speci?ca
`tions J1587 and 11708. A pushbutton is also included to
`trigger the acquisition of a snapshot of data. Connection is
`also provided from the machine battery and key switch to the
`electronic module 20.
`In the preferred embodiment, the electronic module 20
`includes a Motorola 68000 microprocessor, a lower level
`communications board (not shown) of a type well-known in
`the art, and a memory section 24 including high level ?ash
`memory and battery backed RAM. The electronic module
`also includes a pair of RS232 connections, one being avail
`able for connection to the satellite communications system
`21 and the other being available for connection to an
`off-board computer 22 used in download of data and ini
`tialization of the system. In the preferred embodiment, the
`off-board computer 22 is a laptop personal computer.
`To document the performance of the machine and/or its
`major components, performance baselines are stored in an
`array within the memory device located in the electronic
`module 20. These baselines are used during key, repeatable
`performance checks of the machine to help verify machine/
`component health and, as discussed below, are used as
`reference points to determine whether the machine is in an
`operating condition in which machine parameters are to be
`processed and stored.
`A subset of parameters for which trend data is to be
`produced is either prede?ned or de?ned via the off-board
`computer 22 or the central computer 16. Each parameter
`includes a dependency de?nition that identi?es the condi
`tions under which data will be stored for trending purposes.
`Typically, the dependency de?nition is selected to indicate
`the normal operating conditions of the machine; for
`example, when engine RPM are above a certain level and
`boost pressure is above a predetermined level. The trending
`de?nition for each parameter may vary and may be a
`function of several other machine parameters that shall be
`referred to as dependencies. Trend data is gathered and
`stored in memory as the speci?ed dependency de?nition is
`met over a speci?ed trend period, which is measured either
`in time, such as over a period of ten hours, or in counts, such
`as over a period of ten transmission shifts. Trend data is only
`obtained while the engine is running. Based on the specified
`trend type, the maximum, minimum, or cumulative value of
`data gathered during this period is then stored as a single
`trend point with counts to determine the average value
`and/or the points available. The determination of whether to
`use the average, maximum, or minimum value to obtain the
`trend point is based on the system designer’s decision
`regarding which type of calculation would provide the best
`indication of changes in engine performance or impending
`failures. It should also be understood that multiple values
`could be calculated for the same sensed parameter, i.e. trend
`points could be calculated to indicate both an average value
`and a minimum value for a designated machine parameter.
`
`10
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`4
`The overall trend is formed by storing a speci?ed number
`of points in the memory device depending on the size of the
`available memory area and the length of the desired histori
`cal data base.
`Trend data may be reset and the de?nitions may be
`rede?ned by the off-board system 22 via one of the com
`munication ports. For example, if a particular application of
`the machine requires a different dependency de?nition for
`one or more of the sensed parameters, the off-board system
`22 is used to modify the dependency de?nition by providing
`commands to erase a given array including a given depen~
`dency de?nition and replace that de?nition with a new
`dependency de?nition. Similarly, this function may be per
`formed by the central computer system 16 via the commu
`nication system 14.
`In addition to the trend data produced for sensed param
`eters, it should be understood that calculated values, such as
`net horsepower or driveline torque, may also be trended in
`a similar manner. Typically, these calculated values are
`determined by the electronic module 20 according to pre
`determined de?nitions in response to a plurality of sensed
`parameter signals.
`Referring now to FIG. 3, an algorithm incorporated in an
`embodiment of the invention and executed by the processor
`within the electronic module 20 to perform the above
`functions is now described. In block 212. The electronic
`module 20 determines whether the engine is running.
`Advantageously, the engine is determined to be running if
`engine speed exceeds cranking engine speed. If the engine
`is not running, then the algorithm will not proceed. If the
`engine is running, control proceeds to block 214 where the
`electronic module 20 reads the sensed machine parameters
`from the datalink or other inputs.
`For each of the sensed parameters, the electronic module
`20 determines at block 216 whether that parameter is to be
`processed to provide trend data. If trend data is to be
`provided, the trending de?nition is retrieved and the depen
`dency parameters are checked at block 218 to determine
`whether the dependency de?nition is satis?ed. If the depen
`dency de?nition is not satis?ed, control returns to block 212.
`The dependency de?nition for each operating parameter of
`interest is de?ned in terms of other sensed machine param
`eters. For example, the dependency de?nition for boost
`pressure may be satis?ed only when engine rpm is greater
`than a low operating speed and less than a high operating
`speed, when the engine rack setting is greater than a prede
`termined level, and when the jacket water temperature is
`greater than a prede?ned operating temperature. That is,
`values for boost pressure are only saved and processed for
`producing trend information when the above conditions are
`satis?ed. In this way, all boost pressure values used to
`produce the trend data will have been acquired when the
`engine is in the same general operating condition. It should
`be understood that the actual ranges, minimums, and maxi
`mums used in the dependency de?nitions are determined
`empirically to de?ne the operating conditions of interest and
`will vary from machine to machine and application to
`application.
`If the dependency de?nition is satis?ed, control proceeds
`to block 220 where the value of the sensed parameter is
`stored. This process is continued until, at block 222, either
`the time period over which each trend point is to be
`determined or the number of events for which each trend
`point is to be determined is reached at which point control
`proceeds to block 224 where the electronic module 20
`calculates and stores the trend point. The time period or
`
`Page 13 of 16
`
`

`
`5,561,610
`
`LII
`
`20
`
`25
`
`5
`number of events is selected in response to the designer’s
`desire for precision, the availability of memory space in the
`memory device, and the length of time or number of counts
`required to obtain meaningful trend points. The calculation
`of the trend point may include accumulating the stored
`values, selecting the maximum stored value, or selecting the
`minimum stored value. The calculated trend point is saved
`and the data array for that parameter is then cleared to allow
`for the storage of data for calculation of the next trend point
`for that parameter.
`Trend data obtained by way of the algorithm of FIG. 3 is
`illustrated in FIG. 4. While the illustrated data has a sub
`stantial variance, straight lines can be ?t to the data to
`illustrate the general trend of the data by known curve ?tting
`techniques, such as the least-squares method. The method of
`determining when a particular straight line trend should be
`terminated and a new line segment de?ned is described
`below in connection with the algorithm of FIG. 7.
`Based on the slope and duration of the trends illustrated
`in FIG. 4, certain judgements can be made regarding the
`likelihood of impending component or system failure. As
`shown in FIG. 5, functions may be de?ned in terms of slope
`and duration of particular trends whereby warning condi
`tions are indicated in response to either the parameter of
`interest changing at a very high rate for even a small period,
`or at a relatively low rate for an extended period. The area
`indicated by the letter “A” indicates the various combina~
`tions of slope and duration of trends associated with normal
`operating conditions. The areas indicated by the letters “B”
`and “C” indicate the caution region and warning region,
`respectively. That is, normal conditions are indicated if the
`magnitude of either the slope or duration of a particular trend
`is relatively small, but as the magnitude of the slope and
`duration increase, the likelihood of a fault condition
`increases.
`Turning now to FIG. 6, the method of de?ning a particular
`line segment associated with a trend in the sensed data is
`described. A ?rst region of data points in a series of data are
`selected and indicated by X—. A second region of data points
`immediately following X- are selected and indicated by X+.
`The sizes of X- and X+ are arbitrary but, in the illustrated
`embodiment, X- consists of 6 data points and X+ consists
`of the following 3 data points. In some embodiments, X+
`may include only a single data point; however, multiple
`points are desirable for X+ to reduce the likelihood that a
`single aberrant point will cause a line segment to be tenni
`nated.
`The data points in X- are used to calculate a best ?t line
`using a known technique such as least squared errors. Using
`the derived equation of the line, the data in X+ is tested to
`determine whether the line segment should continue such
`that the data points of X+ are added to the line segment data
`or whether a new line segment should be started and the
`previous segment terminated. This test is performed by
`using the following inequality:
`
`35
`
`45
`
`55
`
`Fit-Statistic =
`
`MSE(X+, x+ Based OITX-REETCSSI'OH)
`MSE(X—, x—) +
`deviation slack + bias slack)
`
`< Fit tolerance
`
`where: MSE is the mean squared error;
`deviation slack is a bias to the denominator of the ?t
`statistic to discount MSE changes when there is a high
`slope to the trend and in the preferred embodiment is
`equal to a slack constant multiplied by the variance of
`X—;
`
`65
`
`6
`?t tolerance is a constant controlling the tightness of ?t of
`the data allowed for a given line segment, in the
`preferred embodiment is equal to 4.8; and
`bias slack is a constant used to prevent a new line segment
`from being formed in the event that the MSE of X— is
`zero. In the preferred embodiment, bias slack is set
`equal to zero and decision blocks are included in the
`algorithm shown in FIG. 9 to handle situations where
`the ?t statistic denominator is equal to zero.
`If the inequality is true, the left most data point in X+ is
`moved to X- and a new data point is moved to the right most
`side of X+. If the inequality is false, the line segment is
`terminated and a new X— is de?ned beginning with the next
`data point following the de?ned line segment.
`In one embodiment, if the inequality is false, a ?t statistic
`base is set equal to the present ?t statistic. Data points are
`then continuously added to X- and a new ?t statistic is
`calculated when each new data point is added. Data points
`are no longer added to X- when the ?t statistic reaches a
`preselected ?t poomess percentage indicative of the mini
`mum ratio of the ?t statistic base to the ?t tolerance. This
`allows a subset of the data points in X+ to be added to the
`line segment if the trend actually changes in the middle of
`X+ rather than at the beginning of the X+ group of data
`points.
`The algorithms performed by the present invention are
`now described in connection with FIGS. 7 through 10. At
`block 100, the algorithm is started. At block 101, the default
`?t parameters are initialized and any accumulated values are
`set equal to zero. At block 102, the system determines
`whether there have been overrides of the defaults. At block
`103, data points are loaded into the computer system by way
`of one of the communication systems. For each of the data
`points, the time at which it was stored and its magnitude are
`arbitrarily chosen as X and Y. This data is next processed at
`block 104 which is described in more detail in connection
`with FIG. 8. Following processing of the data, the output
`?les are closed at block 105 and the algorithm is terminated
`at block 106.
`The ?mctions described in connection with blocks 104
`and 105 describe a system in which the analysis of data
`occurs olT-board the machine at a central computer 16. In an
`alternative embodiment, the analysis occurs on-board the
`machine in the electronic module 20 on a real-time basis. In
`this case, the results of the processing in block 104 are stored
`but not written out and the function of block 105 is omitted.
`Turning now to FIG. 8, the X- minus statistics are zeroed
`and the X- window is ?lled with the beginning data points
`at blocks 104.1 and 104.2, respectively. The computer then
`determines whether there are more unprocessed data points
`than the number of points to be included in the X+ window
`at block 104.3. If yes, the left most point in X+ is moved to
`the right most position in X- and a new data point is added
`to the X+window at block 104.4. New statistics are then
`calculated to arrive at a new best ?t line for X- using the
`least squared errors technique at block 104.5, although other
`curve ?tting techniques are equally suitable. The mean
`squared error of the data points in the X+ window is then
`calculated at block 104.6 with respect to the best ?t line
`derived in block 104.5 to determine the relative ?t of the data
`points in X+ with respect to the line determined using the X—
`data points as compared to the ?t of the data points in X—.
`At block 104.7, and as set forth in more detail in FIG. 9, a
`determination is made as- to whether the data points in X+
`are a part of the trend described by the X- line segment or
`whether a new line segment should be de?ned to describe a
`new trend.
`
`Page 14 of 16
`
`

`
`5,561,610
`
`20
`
`25
`
`7
`Block 104.7.1 is the start of this determination. In block
`104.7.2 the ?t statistic denominator of the equation dis
`cussed above is calculated. If the ?t statistic denominator is
`greater than zero at block 104.7.3, the ?t statistic is calcu
`lated at block 104.7.4 and compared to the ?t tolerance at
`block 104.7.5. If the ?t statistic is less than the ?t tolerance
`then the ?t passes at block 104.7 .8, otherwise the ?t fails at
`block 104.7.6. If the ?t statistic denominator is not greater
`than zero, then the ?t fails if the means squared error of X+
`computed at block 104.7.7 is greater than zero, otherwise the
`?t passes.
`If the ?t passes, then the warning status is determined at
`block 104.11. If the ?t fails, then the data points in X—— are
`de?ned as a new line segment at block 104.8 and the
`warning status is determined at block 104.11. If there are no
`unprocessed X, Y pairs in the X+ window, then the statistics
`corresponding to the last X— line is stored at block 104.9. At
`block 104.10, the system waits until a new X, Y pair is
`available and loaded and the process continues from 104.4.
`The detailed operation of the warning status determina
`tion block at 104.11 is set forth in FIG. 10 at block 104.11.1.
`At block 10411.2 the distance of the last point in X+ from
`the mean of the data points in X— and X+ is calculated in
`standard deviations. Alternatively, the position of the data
`point is determined with respect to prede?ned warning
`levels such that an indication is provided of the relative
`distance of the data point from a warning level.
`The slope and duration of the line segment is determined
`at block 104.11.3 from the equation of the X- line segment
`obtained using the least squared technique, although other
`curve ?tting techniques may be used. At block 10411.4
`based on this slope and duration, the computer obtains a
`signi?cance factor from a look-up table of a type well
`known in the art. The value of the signi?cance factor
`increases as slope and duration of the trend increase. That is,
`for any given slope, the signi?cance factor increases as
`duration increases; and for any given duration, the signi?
`cance factor increases as slope increases. The precise values
`used in the look-up table are selected as a matter of design
`choice such that faults are indicated only when there is a
`substantial likelihood that a failure is imminent.
`In one embodiment, the signi?cance factor itself is used
`to determine whether a fault condition is indicated. Alter
`natively, a warning parameter is calculated at block 104.115
`by multiplying the signi?cance factor by the distance from
`the mean in standard deviations obtained in block 10411.2.
`In another embodiment, a number weighting factor at block
`104.11.13 indicative of the relative distance of the data point
`from a warning level as computed in block 104.11.12 is
`combined at block 10411.4 with the signi?cance factor to
`produce a warning parameter at block 10411.5. The mean
`used in block 10411.2 may be one of either an exponential
`weighted average of data points, a mathematical average of
`a predetermined number of the most recent data points, a
`mathematical average of all data points stored since the last
`servicing of the machine, or any other similar procedure for
`determining a historical mean.
`The determination of the warning parameter adds a third
`dimension to the illustration of FIG. 5 such that the warning
`levels are a function of not only slope and duration, but also
`distance from the mean. In this way, a lesser signi?cance
`factor is required to indicate a fault condition when the
`parameter level is a substantial distance from the mean or
`near a warning level.
`The warning parameter is then compared to a caution
`threshold. In the preferred embodiment, if in block 10411.2
`the data point is below the mean, the resulting distance is
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`8
`expressed as a negative number and similarly, if the slope is
`negative, the resulting signi?cance factor is negative. This
`approach eliminates the false indications of fault conditions
`that may occur when the parameter level is low but is
`returning to the mean along a trend having a signi?cantly
`positive slope for a su?icient duration.
`If the warning parameter is less than the caution threshold
`at block 10411.6, then the line segment is in a range similar
`to the two-dimensional area “A” in FIG. 5 which indicates
`a normal operating condition. Any warning that had been
`active is then reset at block 10411.7.
`If the warning parameter is greater than the caution
`threshold at block 10411.6, then the line segment is outside
`the normal operating region. If the warning parameter is
`greater than the caution threshold but less than the warning
`threshold, then the line segment is in a region similar to the
`two-dimensional area “B” in FIG. 5 which indicates a
`caution region as shown at block 10411.8. If the warning
`parameter is greater than the warning threshold as indicated
`at block 10411.9, then the line segment is in a region similar
`to the two-dimensional area “C” in FIG. 5 which indicates
`a warning region. This warning is indicated at block
`104.11.10 and stored at block 104.11.11. As stated above, in
`an alternative embodiment, the signi?cance factor itself is
`compared to caution and warning thresholds and thereby fall
`into the regions identi?ed in FIG. 5. If the warning param
`eter is not greater than the warning threshold, control
`proceeds to block 104.11.11 where the fault condition is
`stored.
`The existence of fault conditions are indicated by any of
`a plurality of available wanting means, such as lights or
`horns, at either or both of the work machine 12 and central
`computer system 16. The existence of the fault condition is
`also stored for use by diagnostic personnel. Control returns
`to blo