Note: Descriptions are shown in the official language in which they were submitted.
CA 02738199 2014-09-15
WO 2010/033529
PCT/US2009/057066
SYSTEM AND METHOD FOR CONTROLLING A
CONVEYOR BELT CONDITION MONITORING SYSTEM
REFERENCE TO RELATED APPLICATIONS
FIELD OF THE INVENTION
[0002] The invention relates to systems for monitoring industrial conveyor
systems and, more particularly, to a system and method for analyzing signals
from
conveyor belt system sensors to provide automatic fault monitoring and signal
analysis.
BACKGROUND
100031 Conveyor belts and conveyor systems are well known systems used for
the
transport of a variety of materials and products. Conveyor belts are designed
and used in
heavy materials transport applications such as coal mining, ore mining, cement
manufacturing operations, and the like. In many such applications, conveyor
belts are
located in underground mines where access to long stretches of belt and
conveyor
components is severely limited. As can be appreciated, unexpected failures in
conveyor
belts in these limited access areas can be dangerous and can also cause
substantial
production delays.
100041 As a result, methods and systems have been developed to monitor the
condition of conveyor belts in operation to predict when failures may occur.
If
predictions are accurate, the conveyor system can be stopped and the belt
repaired at a
1
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
predetermined or more easily accessible location within the mine or above the
ground
surface in the case of underground conveyors. While current systems offer some
degree
of automated monitoring, there is still a need for a fully automated belt
monitoring
system that is capable of analyzing a variety of sensor data indicative of
belt condition,
and of triggering alarms to alert a user of the belt condition. There is also
a need for an
automated system that can sense a dangerous or imminent failure condition and
can
automatically stop the conveyor so that catastrophic system failure does not
occur. Such
a system should be sufficiently sensitive so that it successfully detects
actual problem
conditions, but should also be sufficiently discriminating that it avoids
alarming or
tripping where such responses are not warranted. Such a system should also
meet or
exceed applicable industry standards and regulations as well.
SUMMARY OF THE INVENTION
[0005] The disadvantages heretofore associated with existing systems are
overcome by the disclosed design for a conveyor belt monitoring and analysis
system.
An automated industrial monitoring system is therefore disclosed, which
includes a
plurality of sensors for measuring a magnetic field in a conveyor belt
component and a
controller for receiving signals from the plurality of sensors. The controller
is in
communication with a processor executing instructions for (i) populating a
table with
event data representative of the signals, (ii) comparing a first set of event
data from the
table with a second set of event data received from at least one of the
plurality of sensors,
and (iii) triggering a user alarm if the second set of event data deviates
from the first set
of event data by a predetermined amount.
[0006] In one aspect of the invention, the system may be effective for
automatically transforming the event data into a graphical representation for
display to a
user.
[0007] In another aspect, the system may be effective for automatically
assigning
the event data a rating corresponding to the magnitude of difference between a
second
2
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
waveform and a first waveform corresponding, respectively, to the second set
of event
data and the first set of event data,
[0008] In another aspect, the system may include a graphical user
interface
connected to the PLC so that a user may interact with and control said system.
[00091 In still another aspect, the plurality of sensors may be effective
for
analyzing the conveyor belt by defining separate charnels, and each of the
channels
corresponds to an endless longitudinal section of the conveyor belt. A
graphical
representation of each of the channels may be automatically generated by the
system for
simultaneous display to a user.
[0010] In another aspect, the system includes recording means for
automatically
recording and storing event data relating to the conveyor belt component,
including the
date and location on the belt corresponding to the event data, when the system
is
operating in a learning mode or a monitoring mode, such that new event data
relating to
the conveyor belt may be recorded and stored for comparison with subsequently
obtained
event data relating to the belt,
[00111 In another aspect of the invention, the conveyor belt may include
radio
frequency identification (RFID) chips. The chip(s) may provide information to
the
system relating to the chip, a rip panel associated with the chip, or a belt
splice associated
with the chip.
[0012] In yet another aspect of the invention, a method for controlling a
conveyor
belt condition monitoring system is provided. The method is useful for
controlling belt
monitoring systems of the type that produce signals from conveyor belt system
sensors to
provide automatic fault monitoring and signal analysis. The inventive method
comprises
providing a programmable logic controller (PLC) for receiving the signals. The
controller is in communication with a processor executing instructions for:
(i) populating
a table with event data representative of the signals; (ii) comparing a first
set of event
data from the table with a second set of event data received from at least one
of the
3
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
sensors; and (iii) triggering an alarm if the second set of event data
deviates from the first
set of event data by a predetermined amount.
[0013] In another aspect, the method comprises calculating a slope
measurement
at a plurality of points along a first waveform and at a plurality of points
along a second
waveform to determine deviations in event data.
[0014] In another aspect, the method includes recording first positive
and first
negative maximum amplitudes and second positive and second negative maximum
amplitudes of each of said signals in order to determine deviations in event
data.
[0015] In yet another aspect, the method may include automatically
recording and
storing event data relating to the conveyor belt component, including the date
and
location on the belt corresponding to the event data when the system is
operating in a
learning mode or a monitoring mode so that new event data relating to the belt
is
recorded and stored for comparison with subsequently obtained event data
relating to the
conveyor belt. The event data relating to the conveyor belt component may
comprise an
event waveform, and the method may include automatically recording said event
data in a
belt data table/belt map in the processor.
[0016] In still another aspect, the method may include operably
connecting the
system to a facility-wide monitoring system, an Intranet, a virtual private
network and/or
the Internet.
[0017] One object of the invention is to provide an improved system and
method
for controlling a conveyor belt condition monitoring system, which is capable
of
providing automatic fault monitoring and signal analysis. Related objects and
advantages
of the invention will be apparent from the following description.
4
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
BRIEF DESCRIPTION OF THE DRAWINGS
[0018] The details of the invention, both as to its structure and
operation, may be
obtained by a review of the accompanying drawings, in which like reference
numerals
refer to like parts, and in which:
[0019] FIG 1 is a schematic of a conveyor belt monitoring system;
[0020] FIG. 2 is an exemplary computer display generated from data
collected by
the system of FIG. 1;
[0021] FIG. 3 is an exemplary depiction of a event observed by the system
of
FIG. 1;
[0022] FIG. 4 is a flowchart describing exemplary processing steps for
analyzing
data collected by the system of FIG. 1;
[0023] FIG. 5 is a flowchart describing exemplary processing steps for
analyzing
alarm conditions for a conveyor belt monitored by the system of FIG. 1;
[0024] FIG. 6 is a flowchart describing exemplary processing steps for
analyzing
data representative of events occurring in a conveyor belt monitored by the
system of
FIG. 1;
[0025] FIG. 7 is a diagram of an exemplary data structure and indexing
technique
for use in handling data collected by the system of FIG. 1;
[0026] FIG. 8 is an exemplary depiction of an alternative pattern
recognition
technique for analyzing the event of FIG. 3;
[0027] FIG. 9A is an exemplary computer display showing the system
operating
in learn mode, in which four events have been learned and are populating the
belt data
table;
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
[00281 FIG. 9B shows the customization screen for setting alarm and trip
tolerances associated with a particular conveyor belt, in which parameters
governing
three belt channels are being configured;
[00291 FIG. 9C illustrates a screen for a general configuration for alarm
and trip
tolerances for the non-historical monitoring configuration;
10030] FIG. 9D shows a fiill screen display of an exemplary splice joint
event,
wherein the individual channels are shown in different colors; and
[0031] FIG. 9E shows a further refinement of the display of FIG. 9D in
which
each channel is represented separately.
DETAILED DESCRIPTION
[0032] An improved system and method are disclosed for collecting,
analyzing
and managing data relating to the physical condition of conveyor belting
systems.
Specifically, the system collects data from sensors mounted adjacent to a
conveyor belt
and transmit the data to a computer system capable of analyzing the signals to
determine
whether a damage condition (e.g., rip, reinforcing cord break, splice failure)
exists or will
soon exist in the belt. The analyzed data is presented to a user in an
interactive display
format to enable the user to make affirmative decisions regarding further
operation of the
belt. The system also may be capable of predicting an imminent failure of the
belt and
may automatically stop the belt to avoid or minimize catastrophic damage to
the
conveyor system and/or loss of the materials being carried on the belt.
[0033] Referring to FIG. 1, an exemplary conveyor belt monitoring system
1 is
shown mounted adjacent to a conveyor belt 2 that moves in the direction shown
by the
arrows as a result of the rotation of one or more pulleys 4. The system 1 may
comprise a
permanent magnet 6 that spans the width of the belt 2 and is positioned above
or beneath
the belt in sufficiently close proximity that it magnetizes a portion of the
belt 2.
Specifically, the permanent magnet 6 will cause the magnetization of metal
elements
6
CA 02738199 2014-09-15
WO 2010/033529
PCT/US2009/057066
within the belt such as belt reinforcement cords, rip panels and splice
joints.
[0034] A series of sensors 8 are positioned across the width of the belt 2
at a
location downstream of the permanent magnet 6. These sensors 8 detect
disruptions in
the magnetic fields induced by the permanent magnet 6. Such disruptions are
known to
be indicative of certain conditions in the belt. For example, the disruptions
can represent
a break or damage in one or more of the belt reinforcing cords. They can also
represent a
belt rip panel, or a belt splice joint. As will be described in greater detail
later, disruption
magnitudes and trends can be used to predict the short term and long term
health of the
conveyor belt 2.
100351 The sensors 8 may be divided into groups, or belt channels, across
the
width of the belt so that two or more sensors 8 represent a single data
channel for
purposes of the analysis. In one exemplary embodiment, three channels divide
the belt
into three sections ¨ left, center, right ¨ for analysis. As will be
appreciated, the sensors 8
may comprise electric coils, or they may be Hall Effect sensors.
[0036] The sensors 8 may be connected to a processing system 10, which in
the
illustrated embodiment is a programmable logic controller (PLC) rack system.
The
sensors 8 may be connected to the processing system 10 via a high speed analog
input 12
(where the sensors comprise coils), or to an Ethernet input 14 via an Ethernet
switch 16
(where the sensors comprise Hall Effect sensors). The signals received from
the sensors
are then directed to the CPU 18 for processing.
100371 An encoder 20 may be used to obtain information on the velocity of
the
belt, and its signal may be input to a counter module 22 in the processing
system 10.
Encoder 20 may be integrated with the belt drive, as shown in FIG. 1.
7
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
[0038] The system 1 may further comprise a camera 24 for transmitting
visual
information regarding the top surface 25 of the belt 2 to enable the user to
correlate
information provided by the system with a visual indication of the belt at a
desired
location. The signal from the camera 24 can be routed to the processing system
through
the Ethernet switch 16. In addition, the camera 24 can be controlled (e.g., to
take a
snapshot at a targeted location) via a relay output module 26 of the
processing system 10
such that it can be operated to automatically take a picture based on a
triggering event
tied to the processing system. The relay output module 26 may also provide a
connection
to a belt control system 28 to stop the belt 2 where the processing system 10
predicts an
imminent failure condition in the belt. The belt control system 28 may be part
of a larger
mine monitoring system,
[0039] As will be appreciated, more than one camera 24 may be provided.
Any
or all of these cameras may be high definition cameras that are position
adjustable and
incorporate a zoom function. Such an arrangement may allow the system (or a
user) to
adjust the camera(s) to focus on particular areas of interest. The information
from the
camera may be incorporated into an e-mail and sent to a local or remote user.
[0040] A graphical user interface or user input and monitoring console 30
may be
provided to enable a user to interact with and control the processing system
10. In
addition, the system 1 may include a local display 32 that can be positioned
near the belt
2. The console 30 and local display 32 may be connected to the processing
system via
the Ethernet switch.
[0041] Providing a local display may be advantageous for applications in
which
the console 30 is located in a remote control room or other remote location
away from the
conveyor belt. In one example, the console 30 may be located in a different
geographic
location from the conveyor belt, and may be connected to the processing system
10 via
virtual private network, Intranet, the Internet, or other data communications
system. In
such instances, the local display 32 would provide local operator access to
the
information generated by the processing system 10.
8
CA 02738199 2014-09-15
WO 2010/033529
PCT/US2009/057066
[0042] Referring now to FIG. 2, an exemplary user/operator display screen
is
shown. This is an example of the display that would be shown on the user input
and
monitoring console 30. Portions of this display would also be available for
viewing via
the local display 32,
[0043] In this exemplary display, the screen is cut into four windows. The
first
window 34 shows a picture or live video feed from the camera 24. The second
window
36 contains information about the belt, including the total number of splices
in the belt,
the total belt length, distance to the next belt event and distance to the
next splice. For
the purposes of this disclosure, an "event" may be defined as anything that is
observed on
the conveyor belt, good or bad. Examples of observed "events" include (1)
splice joints,
(2) rip panels, and (3) generalized belt damage. The system 1 monitors these
events to
ensure that the user is aware of any changes in the events (e.g., rips in rip
panels,
unacceptable increase in damage, splice joint degradation) so that appropriate
action may
be taken. In some cases, the system 1 can operate to shut the belt down
automatically if
the magnitude of a change in a particular event exceeds a predetermined
amount.
[0044] The second window 36 may also contain a count of the total number of
events recorded for the belt (number of rip panels, number of damage spots,
etc.). It may
also contain information regarding the number of alarms triggered for the
belt. Where
the belt contains radio frequency identification (RED) chips associated with
belt splices
or rip panels, information regarding the RFIDs and their associated splice/rip
panel may
be provided.
Soft keys 37 at the bottom of the second window of FIG. 2 may be provided to
allow
toggling between different display information to enable the user to customize
the display
to show a desired set of data for the monitored belt.
[0045] A third window 38 may provide a graphical representation of the
signals
received from the sensors 8. This representation can be shown as a real time
feed, or it
can be stopped to enable the user to view a particular event in detail. In the
illustrated
9
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
embodiment, a graphical representation of each of three channels is provided,
although
greater or fewer than three channels may also be shown, depending on the
system. Three
"events" are shown in the illustrated screen. The left two events 40, 42
represent rip
panels, while the rightmost event 44 represents a splice joint.
[0046] The fourth window 46 may provide a listing of events recorded for
the
particular belt 2 being monitored. As can be seen, a listing of the events
(splice, rip
panel, damage) is shown, as is the distance of each event from a known
location on the
belt (e.g., a splice joint). This list may scroll so that the topmost event
listed corresponds
to that event's graphical depiction in the third window 38. For events that
represent belt
damage, the listing may also provide more detailed information about where
across the
belt the damage exists (e.g., right, left, center). Additionally, for events
that represent
splice joints, information may be provided to identify the type of splice
joint present (e.g.,
B92F).
[00471 The user may view information about a particular "event" in the
listing
simply by clicking on that "event," whereupon the graphical signal
representation will be
shown above it in the third window 38. Likewise, standard control buttons 48
are
provided in the third window 38 to enable the user to control the graphical
signal
representation. As will described in greater detail later, a variety of other
display options
may also be provided.
[00481 Referring now to FIG. 3, a generalized description of signal data
collection for an exemplary splice joint "event" is shown. The graph shows the
signal
history, generated on a single channel, resulting from a magnetized splice
joint passing
the sensors 8. Where the sensors 8 provide more than one channel of sensing,
the FIG. 3
data recording function is performed for each channel individually, enabling
on-going
monitoring of the conveyor belt events on a per-channel basis.
[00491 For a length of conveyor belt 2 that is not damaged or does not
have a rip
panel or a splice, the steady state reading from sensors 8 is nominally zero.
Thus, to
sense the beginning of an event the system 1 sets high and low reading lirnits
50, 52 and
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
begins recording data only when that high or low limit is crossed for a
predetermined
minimum period of time (see, Event Begin Time Limit, in FIG. 3) to avoid false
events.
In the illustrated embodiment, splice event begins at 54 where the signal
crosses the high
limit 50 on any channel being recorded. At this point 54, the system sets a
distance
marker, which is a reading from the encoder 20, and begins recording signal
data (i.e.,
summing signal amplitude data to determine an average and also to look for
maximum
positive amplitude) until the signal crosses back down through the low limit
52. Average
and first maximum amplitude (positive) 56 of the signal are then stored.
[0050] When the signal passes through the low limit 52, the system also
sets a
first zero crossing point 57 and again records signal data (once again,
summing signal
amplitude data to determine an average and also to look for a maximum negative
amplitude) until the signal passes back across the low limit 52. The system
sets a second
zero crossing point 59, and stores average and first maximum amplitude
(negative) 58 of
the signal. The signal again passes through the high limit 50 and signal data
is once again
recorded (summing and searching for the second maximum positive amplitude)
until the
signal crosses back down through the high limit 50, whereupon the second
average and
second maximum positive amplitude 60 of the signal are recorded. When the
signal
remains between the high and low limits 50, 52 for a predetermined period of
time (see,
Event End Time Limit in FIG. 3) on all recorded channels, the event ends 62,
the system
sets a second distance marker to identify the boundaries of the event, and
average and
maximum amplitudes are stored.
[0051] The system 1 stores all of this infomiation as an "event," noting
the time
and location on the belt. Thus, baseline data regarding this particular splice
joint is used
to compare against data collected for that splice joint the next time the
joint passes the
sensors 8. Any changes from previous readings may be noted, and the user can
be alerted
of significant changes from previous readings. For example, each time an event
is
triggered, the system I may send an alarm such as sending a warning e-mail,
taking a
photographic snapshot of the belt using the camera 24, or some other action.
The specific
11
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
action required may be pre-set by the user (e.g. fifteen percent increase or
decrease in
maximum amplitude from previous reading).
[0052] Thus, a user need only look at the event and sign off on it if it
goes above
a certain value. A historical catalog of such acceptances of a particular
event is stored
and is available for viewing via the display screen (FIG. 2). The algorithm
will then only
signal an alarm based on the last previous event (i.e., fifteen percent
greater or less than
the last accepted event). Of course, this can be modified so that an
additional five percent
increase/decrease will alarm, etc. For significant changes (i.e., associated
with an
imminent catastrophic damage condition), the system 1 may cause an automatic
trip of
the conveyor belt 2 by sending an appropriate signal to the belt control
system 28.
[0053] Hence, the system 1 may, for example, be set up to require a trip
(i.e., belt
stop) if the maximum signal amplitude is twenty percent greater or lower than
the last
previous run, and may require merely an alarm if the maximum signal amplitude
is ten
percent greater or lower than the last previous run.
[0054] It will be appreciated that although the foregoing description
relates to the
data recording process for a splice joint "event," similar principles are
applicable when
recording damage-type events and rip-panel events. Referring now to FIG. 4,
the data
processing flow of the system 1 will be described in greater detail.
[00551 At step 100, input mapping is performed, in which sensor data paths
are
mapped to the appropriate locations in processor 18 of the PLC. Speed scaling
of the
signals is implemented in this block to normalize the signals from the sensors
based on
the speed at which the belt is traveling. This may be important because signal
voltage
increases as belt speed increases. Thus, while a sensor may output a maximum
100
millivolt (mV) signal for a particular "event" at a belt speed of one
meter/second (m/s),
the same event may generate a 300 mV signal in the sensor at a belt speed of
three m/s.
The inventors have discovered that an appropriate normalization is achieved by
dividing
the sensed signal by velocity.
12
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
[0056] At step 200, the event detection block determines whether an event
has
occurred. As previously noted, an event is detected when the signal from a
sensor goes
above the high limit 50 or below the low limit 52 set by the system for a
predetermined
period of time. When an event is detected on at least one channel, the system
considers
an event to have occurred on all channels and thus all channels will be
recorded. It is
expected that the high and low limits 50, 52 may not be universally equal for
all conveyor
belts, and may depend on the nature (type, age, etc.) of the belt being
monitored. Belt-
specific high and low limits will, therefore, likely be set during an initial
conditioning
step. Customizing the high and low limits 50, 52 to a particular belt will
ensure that only
valid events are detected.
[0057] At step 300, the event processing block collects information
regarding the
event waveform. This event waveform information is placed in a temporary table
for all
enabled channels until the event is complete, whereupon the data is written
into the belt
data table/belt map in the processor memory 18.
[0058] At step 400, the event end block detects the end of an event. This
occurs
when the signal is "flat-lined" for all channels (i.e., the signal is between
the high and
low limits 50, 52). At this point the system 1 determines what kind of event
took place
(e.g., rip panel, splice, general darnage) based on the general form of the
waveform. For
example, a splice joint event may have three maximum amplitudes and two zero
crossings, while a rip panel event may have two maximum amplitudes, only one
zero
crossing, and a zero signal in the center channel. Generalized damage may be
anything
that does not conform to the splice or rip-panel pattern. The system is
capable of
recognizing these characteristics and classifying the event appropriately. At
this point the
data from the temporary table is placed in the belt data table/belt map.
[0059] At step 500, the system determines whether, based on the data
collected
during the event, an alarm or trip should be ordered. Referring to FIG. 5, the
alarm
decision is made depending on whether the system is operating in the
"learning,"
"searching," or the "monitoring mode." The learning mode is implemented during
the
initial conditioning of the belt, where the system is initially learning the
events that exist
13
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
on the belt. Learning is automatic and is typically performed on initial
system startup
with a new belt (or where the new system is implemented to begin monitoring a
belt that
has already been in service). Thus, upon the initial revolution of the belt,
the system 1
will sense the signals from the sensors 8 and will populate the belt data
table/belt map
with an initial set of events.
[00601 Searching mode is implemented, not on a new belt, but rather when
the
system 1 encounters an unexpected event (e.g., a splice or rip panel in an
unexpected
location). Such an occurrence of an unexpected event is an indication that the
system 1
has lost its registration of the belt, and thus a searching of the events on
the belt is
performed to reset the system.
[00611 When operating in learning and searching modes, alarms are set
only for
rip panel events. (Block 510) Thus, if the rip panel event signals exceed the
preset
limits, an alarm or trip is ordered at step 520. Although the tolerance is
rather wide, this
presents a degree of protection for the rip panels until the monitoring mode
is established
or reestablished.
100621 When operating in inonitoring mode, (block 530), sensor readings
for
current "events" are compared to the readings from the stored data table for
the same
event. If the sensor reading exceeds the alarm limit (e.g., ten percent higher
than the last
revolution), then an alarm is ordered. Likewise, if the sensor reading exceeds
the trip
limit (e.g., twenty percent higher than the last revolution), then a trip is
ordered. Data
relating to the alarm or trip is written into an Alarm/Trip Table 540 indexed
to the
particular event. This data is then written into the belt data table 550.
[00631 In one embodiment, the tolerances for each waveform are customized
per
event, and are not generic for all of the events in the belt. Thus, the high
and low
tolerances for alarm and trips may be customized for each particular event
(rip panel,
splice, generalized damage) on the belt. They may even be different within
groups,
providing different tolerances for different types of splices or rip panels.
This allows a
tighter control over the rria?dmin allowable sensor reading for each rip panel
or splice.
14
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
[0064] When the system is operating in monitoring mode (530), comparison
of
the current event waveform to the historical waveform stored for the same
event is
performed as soon as the event is identified rather than waiting for the event
to end to
perform the comparison. This can result in a savings in computing time.
[0065] At step 600, the event data handling block manages the data by
determining whether the event is a new event or an old event (i.e., one that
has previously
been detected). During the initial startup of the belt, the system is in
learning mode 610
in which it stores (at step 620) all the data about the splices, rip panels
and existing
damage in the belt data table/belt map. While this learn mode occurs during
startup, it
can also occur after splice or rip panel configuration changes.
[0066] After the initial learning mode is completed, the event data
handling block
runs in "monitor" mode 630. In lnonitoring mode, the system monitors the
events as they
occur, determining whether they are new events or old events. At step 640, the
system
determines whether the event is new, old, or indicative of a loss of
registration. Thus, if
the detected event is expected (i.e., the system detects a splice, and the
next expected
event at the sensed location based on historical performance is a splice),
then the system
continues on to the next event. If the detected event is a new event (e.g.,
generalized
damage is detected but there is no expected event at the sensed location),
then the event is
added to the belt data table/belt map 550. This is typically the case when new
damage to
the belt is observed for the first time. An event is considered to be "old" if
an event of its
type has been identified in the same location during previous revolutions of
the belt. If
the system detects a rip panel or splice at a location on the belt where,
based on historical
data, no such rip panel or splice is expected, then the system is diverted
into "searching"
mode. Such a condition may indicate that the system has lost its registration.
100671 When a loss of registration occurs, the system will first attempt
to
compare the incoming events with previously stored unique identifiers. These
unique
identifiers (e.g., RFID chips, mechanical uniqueness such as two rip panels in
a row
followed by a particular type of splice, splice signature, rip panel
signature), are
originally stored when the system 1 is operating in learning mode, and the
identifiers are
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
associated with events specific to a particular belt or component on the belt.
Thus, to re-
register the belt in searching mode, incoming events are compared against
these unique
identifiers and once a match is determined, the system can resume operation in
monitoring mode. The unique identifiers can be identified in a separate table,
or they
may be identified in a particular location in the belt data table/belt map. If
the belt cycles
through a complete revolution and no event match is obtained, then the system
may
return to learning mode to re-learn the events on the particular belt.
[0068] At step 700, the output mapping block outputs the result of the
event
analysis to the user. This output can be an alarm, an e-mail alert or message,
a trip order
to stop the belt, or the like.
[00691 Referring to FIG. 7, the belt data table/belt map 550 will be
described.
The data table 550 will contain information regarding each of the events
detected in the
belt. The data table 550 is originally populated with event data when the
system is
operated in learning mode. New data regarding an event is written into the
data table 550
when an alarm condition occurs for the event and the user acknowledges the
event as
acceptable so that belt operation may continue.
[0070] The belt data table 550 contains information regarding all of the
events
observed on the belt (e.g., waveform, etc.). This table is originally
populated with event
data when the system is operated in learning mode. As new events such as belt
damage,
for example are observed, data representative of these new events are written
into the belt
data table 550 in descending order.
[0071] A scrolling index 560 may then be used to provide a cross-
reference
between the events, in the order in which they occur on the belt, and the
location of the
respective event data in the belt data table 550. Thus, in the initial belt
learning mode,
the system registers the events in order (events "1" ¨ "5" in FIG. 7). Then,
in monitoring
mode, when a new event is observed (e.g., event "6" in FIG. 7), the new event
identifier
is placed into the scrolling index in a manner that represents its physical
location on the
16
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
belt (Le., new event "6" is observed between known events "2" and "3"). The
scrolling
index is referenced by the system 1 when the system searches for the next
expected event.
100721 Thus, as new events are observed, reorganization of the scrolling
index is
all that is required. No reorganization of information in the belt data table
550 is
necessary, which saves substantial processing time.
[0073] FIG. 8 shows an alternative event analysis technique to that
described in
relation to FIG. 3 in which a pattern recognition approach is used to provide
finer-
sensitivity in event monitoring. In the FIG. 8 approach, the splice joint
event is sensed
and the system begins recording data when one of the high and low reading
limits 800,
802 is crossed. In the illustrated embodiment, the splice event begins at 804
where the
signal crosses the high limit 800. At this point 804, the system sets a
distance marker
(i.e., a reading from the encoder 20), and begins recording signal amplitude
data into an
array for pattern recognition. High and low pattern recognition limits are set
based on
expected slope measurements at a plurality of locations along the curve.
[0074] Any of a variety of well known techniques may be used to calculate
the
slope of the curve at discrete locations. If the measured slope for an event
deviates from
the expected slope measurement at any of these points, or deviates with
respect to an
expected change in slope over time, an alarm or trip is ordered. This approach
provides
enhanced sensitivity to the event monitoring process because it is capable of
identifying
small deviations in the event curve that would be masked using the averaging
procedure
in the technique described with respect to FIG. 3. This slope calculation
technique,
however is substantially more processor intensive than the technique of FIG.
3.
[0075] As with the technique of FIG. 3, this pattern recognition
technique may
also record first and second zero crossings, and may also store in the belt
data table 550
data representative of the measured positive and negative waveforms to enable
later event
comparison at the next revolution of the belt.
[0076] FIGS. 9A-9E show various alternative information configurations
for
display on the user input and monitoring console 30. Thus, FIG. 9A shows the
system
17
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
operating in learning mode, in which four "events" have been learned and are
populating
the list in the lower right quadrant of the screen. A corresponding graphical
representation of the input signals is shown above the list. FIG. 9B shows a
general
configuration of alarm and trip tolerances for the historical monitoring
configuration,
showing how the system can be customized to provide a plurality of individual
alarm and
trip tolerances on a per-channel basis. FIG. 9C shows a general configuration
of alarm
and trip tolerances for the non-historical monitoring configuration.
[0077] FIG. 9D shows a full-screen representation of a splice joint
event, with
individual channels being displayed in different colors. As set forth above,
the graphical
representation of each of the channels displayed correspond to an endless
longitudinal
section of the conveyor belt. This full-screen display may be activated by
simply double-
clicking on the graphical display quadrant of the normal display. FIG. 9E is a
further
refinement of the display of FIG. 9D in which each channel is represented
separately. As
can be seen, soft keys at the bottom of the screen allow one or more channels
to be
switched off so that the user can highlight and focus on a particular channel
of interest.
[0078] In addition to the benefits of the system described above, it is
also highly
compatible with existing mine monitoring systems because the inventive system
is PLC-
based, Applicants' new system, therefore, preserves the client's old system
while also
implementing the new features, (e.g., rip panels, monitoring features, control
devices
operating at the direction of the inventive system 1). For example, the PLC
components
used to construct the present system can be readily obtained from Rockwell
International
Corporation, and can thus be easily integrated into current mine systems using
software
sold under the trademarks RSVIEW or RSWORKS. The disclosed system display may
be a page "dropped into" an existing conveyor belt and/or facility-wide
industrial
monitoring system.
[0079] In addition, if applicable, the disclosed system can immediately
interact
with the facility controller to instruct the mine system exactly where to stop
the belt when
repairs are required. Thus, belts and their metallic and non-metallic
reinforcing
components, rip panels, and splices may be analyzed and the corresponding belt
systems
18
CA 02738199 2011-03-21
WO 2010/033529
PCT/US2009/057066
controlled and managed irrespective of the particular manufacturer of these
components
and systems.
[0080] It will be understood that the description and drawings presented
herein
represent an embodiment of the invention, and are therefore merely
representative of the
subject matter that is broadly contemplated by the invention. It will be
further
understood that the scope of the present invention encompasses other
embodiments that
may become obvious to those skilled in the art, and that the scope of the
invention is
accordingly limited by nothing other than the appended claims.
19
