JPH01255905A - Trouble diagnostic device for machine - Google Patents

Abstract

PURPOSE:To correctly discover a trouble position by fetching action pattern information from a programmable controller at the time of a trouble, a comparing the action pattern information at the time of normality and displaying difference information. CONSTITUTION:A programmable controller generates the action control information of a machine based on input information from a machine which becomes a sequence controlled system and a sequence set beforehand. A storage means consists of input information and action control information. The action pattern information is stored in a time series at the time of the normal action of a machine. An information fetching means fetches the machine from the programmable controller and the action pattern information at the time of the trouble. A difference detecting means compares the action pattern information at the time of the trouble and the action pattern information at the time of normality and detects the difference between both pieces of information. A display means displays the difference information.

Description

[Detailed Description of the Invention] (Industrial Application Field) The present invention is capable of quickly and instantaneously detecting the cause of a failure when the machine stops, for example, when sequentially controlling a machine such as a manufacturing device using a programmable controller. The present invention relates to a machine failure diagnosis device that displays

(Prior art) Conventionally, for example, as in Japanese Patent Application Laid-Open No. 62-32510,
It is well known that the alignment circuit and storage device 1 permanently installed in the machine display the operating pattern of the machine at the time of a failure and before and after the failure, thereby making it easier for maintenance personnel to diagnose machine failures. There is.

(Problems to be Solved by the Invention) However, such diagnostic devices only display the operating pattern of the machine at the time of a failure and before and after the failure, and ultimately human judgment is required. Furthermore, these displays merely display the results of failures that actually occurred. Therefore, in order to diagnose a faulty part, this person must become familiar with the patterns of all processes and rely on experience and intuition. Basically, with this conventional technology, there is no standard for how a machine should behave under normal conditions, so it has no choice but to rely on human experience and intuition.

Therefore, the present invention was proposed in order to eliminate the drawbacks of the above-mentioned conventional examples, and its purpose is to provide a failure diagnosis device for a machine that is sequence-controlled by a programmable controller, which is equipped with a standard operation pattern during normal operation, By comparing this standard operation pattern with the operation pattern at the time of failure, we have created a machine failure diagnosis system that eliminates the need for failure diagnosis that relies on experience and intuition, and allows anyone to easily discover the failure part of the machine. The point is to make a suggestion.

(Means and operations for solving the problems) As shown in FIG. a programmable controller that generates operation control information for the machine based on the set sequence and controls the sequence operation of the machine; and operation pattern information consisting of the human power information and the operation control information.
storage means for storing time-series information during normal operation of the machine; information retrieval means for retrieving the operation pattern information at the time of machine failure from the programmable controller; information on the operation pattern at the time of failure, and operation during normal operation. The present invention is characterized by comprising a difference detecting means for comparing the pattern information and detecting a difference between the two pieces of information, and a display means for displaying the difference information.

Therefore, since the device itself estimates the faulty part, humans are freed from relying on experience and intuition for fault finding, and more accurate and quick fault finding becomes possible.

(Example) Examples according to the present invention will be described below with reference to the accompanying drawings.

Configuration of Diagnosis System> FIG. 2A is an overall diagram of the failure diagnosis system according to the embodiment. In the figure, 10 is a programmable controller that is connected to a machine (not shown) and performs sequence control of the machine. Reference numeral 11 denotes a diagnostic unit that is detachable from the programmable controller 10, and is a unit that monitors signals input and output to the programmable controller 10 for sequence control of the machine. A graphic console (hereinafter abbreviated as GC) 13 is connected to this failure diagnosis unit 11 via a communication cable 21 . This GC13 has a liquid crystal display 14. A keyboard 15 is mounted, and a memory cassette 16 consisting of a RAM backed up by a battery is provided in a removable manner. The GC 13 also includes a printer interface 18.
A printer 20 can be connected through the floppy interface 17, and a floppy disk drive 19 can be connected through the floppy interface 17.

In the embodiment of FIG. 2A, the diagnostic unit 11 and the GC1
3, memory cassette 16, etc. constitute a mechanical failure diagnosis device.

FIG. 2B is a front view of the diagnostic unit 11, and "
The in-operation JILED 11a lights up when the diagnostic unit 11 is in operation or ready for operation.

The error/detachable LED 11b is used to detect when a monitoring time over error occurs during diagnosis, or when the diagnostic unit 11 is connected to the programmable controller 10.
Lights up when the device can be removed from the device. The unit 11 can be removed by turning on the detachment PB11d. For example, when the 7-segment display LED 11c displays Px, it indicates that the unit 11 is in the teaching mode, and when it displays Dxx, it indicates that the unit 11 is in the diagnostic mode. Switch 11el? ,
When you defeat the lr diagnostic unit J, it becomes a diagnostic mode. Connector llf is an R5422C communication interface connector.

FIG. 3 shows a block diagram of the GC13. The GC 13 is the core of the fault diagnosis system, and its functions are as follows. That is, in the above teaching mode,
Input signals as input information from a normal machine and output signals as operation control information for the machine are collected in chronological order via the diagnostic unit 11 and stored in the memory cassette 16, so that the machine can When a failure is suspected, the above input and output signals at the time of failure are collected, this signal data is combined with the input/output signal data during normal operation, the failure location is estimated, and the liquid crystal display is displayed. display on the display 14. Now, as shown in FIG. 3, the configuration of the GC 13 includes a CPU 30 that manages overall control, a ROM 31 that stores programs to be described later, a RAM 32 that stores intermediate data, and data from the diagnostic unit 11. It consists of an R3422 interface 33 as a communication means for collecting information, a keyboard interface 34, a liquid crystal interface 35, a connector 36 for attaching and detaching a memory cassette, and the like.

FIG. 4 shows the connection relationship between the machine 1 as a sequence control target, the programmable controller 10, and the diagnostic unit 11. As is well known, the programmable controller 10 monitors an input signal from the machine 1 representing the state of the machine 1 and sequentially controls the operation of the machine according to a built-in program. Machine 1 operates based on output signals from programmable controller 10. Since these input signals and output signals define the operation of the machine 1 at a certain point in time, the diagnostic unit 11 monitors these input signals and output signals and sends them to the GC 13 as an operation pattern signal of the machine 1. The feature of this embodiment is that in a normal state, the status signal from the diagnostic unit 11 is stored as an operation pattern in the memory cassette 16 (or floppy disk 19), and when a failure occurs, the status signal is stored as an operation pattern at the time of failure. The idea is to compare the operating pattern during normal operation with the normal operation pattern and use it for fault diagnosis.

<Teaching> First, the memorization and registration (teaching) of normal operation patterns will be explained.

FIG. 5 is a menu screen displayed on the liquid crystal display 14 used by the operator to specify the signal to be registered as an operation pattern among the input/output signals shown in FIG. 4. . This type of signal specification allows you to specify the signals to be monitored for each machine, taking into account that a signal that is necessary for one machine may not be necessary for another machine, and allows you to specify the signals to be monitored for each machine.
This is to allow for versatility. In this embodiment, each of the four signals can be specified, and in FIG. Those labeled o and Il mean that they are used as output signals.

Signals that are not synchronized with machine operation may cause false detection, so they must be excluded from the registered patterns, and the operator must
For such asynchronous signals, select Ir on the menu screen.
Mark N and Q.

The reason for distinguishing between the input signal ■ and the output signal O is as follows. The failure diagnosis system of this embodiment is based on the premise that it is the machine 1 that fails and that the programmable controller 10 etc. do not fail, so the output signal as the operation control signal for the machine 1 from the programmable controller 10 is It is reliable. The state of the machine 1 can then be monitored by: - an input signal to the programmable controller indicating the state of response of the machine 1 to the operation control signal; That is, what operation the machine 1 should perform now is indicated by the output signal pattern as an operation control signal. As mentioned above, since it is assumed that the programmable controller 10 operates normally, this output signal pattern will remain the same regardless of whether the machine 1 is operating normally or malfunctioning, as long as the machine performs the same work. , the output signal patterns match perfectly. Therefore, this output signal pattern serves as a reference for determining the actual operating state of the machine, which changes from moment to moment, at the current time. Therefore, in normal times and in times of failure,
By comparing the input signal pattern as the response of the machine 1 to the above output signal, it can be seen why the machine 1 does not respond normally. This is why we distinguish between input signals and output signals.

While viewing the setting screen as shown in FIG. 5, for each signal, designate whether or not to exclude it as a registered pattern, and designate input I and output 0. Also, from among the signals specified above, a start signal and an end signal that rise for each series of operations of the machine are specified, and their addresses are specified.

At the rising edge of this activation signal, the GC 13 starts monitoring input/output signals every 30m5, takes in data and writes it into the memory every time these input/output cover patterns change. This monitoring every 30 m5 is repeated until the end signal rises or pattern detection is performed a maximum of 3000 times. FIG. 6 shows the input/output cover turn table thus obtained. In the example in Figure 6, with the specifications in Figure 5, 0
Address 000000 Address 777777 778 input/output signals are specified, and the data collected by the above-mentioned monitoring at intervals of 30 m5 are grouped together with the same output signal pattern and assigned one cycle number. In the example of FIG. 6, the output signal pattern specified as "0" is the same for the first sampling data and the second sampling data, so cycle 1
is assigned. And the duration of the cycle is 60 m5. Cycle number 2 is assigned to the third to sixth sampling data.

FIG. 7 shows an output cover turntable obtained by extracting only the output signal (marked as "0") from the input/output cover turntable obtained as shown in FIG. The reason why we focus only on the output pattern is that, as mentioned above, the output signal defines the operation of the machine. That is, in the example of FIG. 7, the signal designated as the output signal (address OOO2) is determined by the designation shown in FIG.

0003003 address...102 address...77
Among the pattern group (assigned to address 7), those with the same pattern are grouped together and assigned one cycle number, and then arranged in chronological order. The cycle numbers of the input/output cover turntable and the output cover turntable thus obtained are the same.

The "cycle name" in FIG. 7 is the name of the cycle that is displayed on the display 14 and registered by the operator by inputting it from the keyboard 15 when the output cover turntable is created. .

In the above manner, the input/output cover turntable (Fig. 6)
An output turntable (Figure 7) was created. By the way, the duration CT of one cycle on these tables obtained by teaching in this way may be slightly different when the actual machine operates. Therefore, it is necessary to take into consideration the variations in these durations and set a reference duration CT2 within which it can be determined that the machine is operating normally. This reference duration CT2 can be set for each cycle, but in this embodiment, the range of ±50% of the duration CTI obtained with the output cover turntable is set as the reference duration for all cycles at once. Time was set as CT2.

For example, in the example of FIG. 7, at cycle number r, CT2
= range of 30ms to 90m5.

In this way, the teaching of the reference pattern table as shown in FIG. 8 is completed. That is, in the reference pattern table of FIG. 8, the input and output cover turntables as shown in FIG. 6 and the output cover turntables as shown in FIG.
It is completed as one table and stored in the memory cassette 16 so that it can be indexed by N.

As mentioned above, the reference pattern table in FIG. 8 can be indexed by cycle number CN, but since the output signal pattern in the input/output cover turntable is the same as the output signal pattern in the output cover turntable, the following For convenience of explanation, a reference pattern indexed by a certain cycle number CN is expressed as DP[CH3, its input signal portion is expressed as DPI[CH3, and its output signal portion is expressed as DPO[CN].

<Control Procedure> The teaching control procedure will be explained according to teaching FIGS. 9A and 9B. Furthermore, the designation of input signal equipment and output signal O, and the designation of signals to be excluded from registration N, which were explained in relation to FIG.
It is assumed that this has already been completed.

First, in step SL and step S2, the diagnostic unit 1
1, the operation pattern (input signal and output signal) of PClo is read, and it is checked whether the specified activation signal mentioned above is "1" during the operation pattern. Unless this activation signal is “1”, step Sl, step S2
Repeat.

When the start signal is detected, in step s4, wait for the sampling interval of 30 m5 to elapse, and when this time elapses,
In step S6, the PCIO operation pattern is fetched, and in step S8, it is stored in the RAM 32 in chronological order. These 30
The acquisition of the motion pattern every m5 is continued until an end signal is detected in step SIO or until the data collection of 3000 samples is completed in step S12. This completes the collection of raw data for registration. Next is the 8th
This is editing of the reference pattern table shown in the figure.

In step S14, the first sample data is stored in RAM3.
2, the duration time Cr t is initialized in step S16, and the cycle number counter CN is initialized in step S18.

In step S20, the next sample data is stored in the RAM 32.
Load from. In step S22, it is checked whether there is a change in the pattern of the output signal portion between the sample data read in step S14 and the sample data read in step S20. Detection of this change is performed by sequentially comparing corresponding portions in accordance with the output signal designation order shown in FIG.

If there is no change in the output pattern, the duration CT is updated by 30 m5 in step S24.

In step S38, it is confirmed that the input/output cover pattern change check for all sample data has not been completed, and the process returns to step S20, where the next sample data is read from the RAM 32.

As long as there is no change in the output signal pattern, step S24
Then, the duration time CT is updated.

A case where it is determined in step S22 that there is a change in the output signal pattern will be described. At this time, in step 326, the duration CT is updated, and then in step 328, this CT is updated.

The reference duration CT is calculated based on the following. In this embodiment, as described above, CT z is within a time width of ±50% of CT. In step S30, the input/output cover turn (
In step S32, only the output signal portion (denoted as DPO) is extracted from the output signal portion (denoted as DP), and the duration c”r*,
Input/output cover turns DP and output cover turns DPO are written and registered in the memory cassette 16 according to the cycle number CN as shown in FIG. In step S34,
Increment the cycle number CN and proceed to step S3.
At 6, CT I is cleared for counting the new cycle duration CT1.

Then, through step S38, the process returns to step S20, and depending on whether or not there is a change in the output cover turn, the duration time CT is updated or the input/output cover turntable D is updated.
P. Register the output cover turntable DPO.

Diagnosis-'p In this embodiment, two failure diagnosis modes are prepared. One is to switch the diagnostic unit 11 to PCIO when a failure occurs.
One is a spot diagnosis mode in which the diagnostic unit 11 is attached to the PC 10 and performs failure diagnosis, and the other is a continuous diagnosis mode in which the diagnosis unit 11 is always attached to the PC 10 and only the attached PC is constantly monitored. . In the former spot diagnosis mode, constant monitoring cannot be performed, but it is possible to diagnose multiple PCs by connecting one set of diagnosis units and C13 to a malfunctioning PC and changing the database in the memory cassette 16. In the latter constant monitoring mode,
Although only one PC is monitored, it is effective against intermittent failures.

First, the spot diagnosis mode will be explained according to the flowchart of FIG. This spot diagnosis is performed by finding the output signal pattern RPO when the machine is stopped in the reference pattern (Figure 8) (identifying the cycle number CN) and comparing it with the actual input signal pattern RPI when the machine is stopped. The faulty part is estimated through comparison with the input signal pattern DPI during the cover turn.

Now, when a failure occurs, the machine is stopped. The operator attaches the diagnostic unit 11 to the PC of the stopped machine, sets the switch ``lie'' to the r diagnostic unit J side, and presses the ``spot diagnosis J key (not shown) on the keyboard 15 of the GC 13 to start the GC 13. Start.

When the GC 13 is activated, in step S50, the operator is prompted to input from the keyboard which station's machine is stopped. This is because the memory cassette 16 stores a database for each station. In step S52, the installed diagnostic unit 11
The current input/output cover turn RP of the actual machine 1 held by the PCIO is read through the PCIO. For convenience, the output signal portion of the input/output cover turn is referred to as RPO, and the input signal portion is referred to as RPI. In step S54, the cycle counter CN is initialized to "1".

In step S56, RPO and DPO [CN
] Compare with. That is, the cycle number is increased by 1 until the output signal portion DPO[CN] in the reference pattern indexed by the cycle number CN matches the pattern RPO.
The count is counted up one by one (step 558), and an output pattern having the same pattern as the output signal state held by the current PCIO is searched from the database, and its cycle number CN is detected.

When such a cycle number CN is detected, the process proceeds from step S56' to step S62, where CN is incremented by "1". Then, in step S64, the actual input signal pattern RPI read in step S52 is
and the input signal pattern D of the database for the next cycle.
PI [Compare with CN+1 and find out the discrepancy.

FIG. 11 shows a diagram illustrating the above control procedure part in more detail. When the number of cycles is CN-n, the reference pattern database should be created as follows.

Input signal pattern DPI.

is input to the PCIO from the machine I, the PCIO outputs the output turn DPO, according to its program.

is output to machine 1. Here, the cycle number is n
After counting up to +1, the machine 1 side responds by sending an input cover turn DP n+ to the PCIO. The PCIO then outputs D P On,I to machine l according to its program settings. If the machine is actually stopped at CN=n, step S5
The cycle number n should have been found in the loop from step S6 to step S60. Therefore, in that case, DPOI,=RPO.

It is. When machine 1 fails, machine 1 receives an input signal pattern RPI, . Since it cannot respond with 1, it means that PCIO is stopped at RPOn. Therefore, by comparing RPOI and D P Oo-+ , it becomes clear what kind of input signal is not in the desired state. This comparison shows that RPO, and DPO
This is done by taking an exclusive OR with n, , , and an input signal whose OR is "1" indicates the cause of the failure.

The input signal considered to be the cause of the failure found in this manner is displayed on the display 14, for example, in a negative video display. In addition, cycle numbers, station names, and other input/output signals may also be displayed.

In step S66, input of a "continue" key (not shown) on the keyboard 15 is awaited. When this key is input, the control proceeds to step 560=>step S56, and then searches for the cycle number of the DPO with the matching RPO. This is done for the following reason. In the reference database created by teaching, there is not necessarily one output pattern DPO having a certain pattern.

Conversely, it is more common for the same output pattern to appear over and over again. On the other hand, in the spot-type diagnosis described now, in such a case, it is not possible to determine how many times the same pattern RPO has occurred. Therefore, by returning from step S68 to step S60, the same output turn RPO
For all cycles with , the RPI in that cycle is
and the DPI of the reference pattern (step 564)
/, the discrepant portion is displayed, and the final decision is left to the operator.

1 Grave Superintendent Yu Meng 2 Next, we will explain the continuous monitoring mode. In this mode, if the input signal from machine 1 changes within the above-mentioned standard duration time 072, the PCIO will also change its output signal pattern accordingly, so the GC 13 will change the output signal to the above CT 2 time. This system constantly monitors whether or not there has been a change in the fault condition, and when there is no change, notifies the occurrence of a fault condition and displays the fault location. For example, if a certain relay in machine 1 deteriorates and the relay sometimes takes about 2 seconds to operate when it should normally operate within 1 second, the reference duration CT2 may be changed to, for example, 1 second. If you set it to .5 seconds, you can find the failure location. In such a case, the spot diagnosis described above cannot detect a failure because the relay is operating in an irregular manner over time.

To carry out this constant monitoring, the program shown in FIG. 12 is started by pressing a key (not shown) on the GC 13.

The number of the station where the diagnostic unit is installed is designated in step S80. In step S82, the cycle number CN is initialized to "1". In step S84, the actual input/output cover turn RPV (=RPn) of PClo is read as first data via the diagnostic unit 11. In step S86, at least the reference duration 07
2, a monitoring timer WDT is activated to monitor whether the input/output cover turn RP from the PC side has changed.

This timer WDT (not shown) is a timer that continues to count up until it is reset once it is activated.
The count value can be read by the CPU.

Step S88 takes a reasonable amount of time (for example, 30 m5
). If this time has elapsed, in step S90, the PClo
The input/output cover turns RPn, , (=RPt) are read. Then, in step S92, this RP, l+1 (
The output signal portion RPO of =RP2) is compared with RPO8 read in step S84 to determine whether there has been a change. In this step S92, generally the RPO
,,1=RPO.

Find out whether or not.

There is no change (RP On++ = RP O,)
If so, in step S93, the reference duration CT 2 corresponding to the current cycle number CN is retrieved from the memory cassette 16.
Read [CN] and proceed to step S94.
The WDT is read to check whether the timer value WDT exceeds the reference duration CT2 [CH3. As mentioned above, this reference time is set within the range of ±50% of the duration CT obtained by teaching, so wD'r>cTz [CH3 means that PClo The sequence cannot proceed further on the side.

When WDT≦Crt[CH3, the process advances to step S99 to replace RPn, I and RPn in order to detect a change in output pattern next time, and returns to step S88 to repeat the above-described control.

If WDT>C70[CH3 is detected in step S96, the process advances to step S98 and displays that a time-over error has occurred. The details are shown in FIG.

If it is detected in step S92 that RPOll□≠RPO0, there has been a change in the output pattern at least within the reference time, so step S92 is detected.
00, clear the timer WDT, and step 5
At 104, the cycle number is incremented. In step 5106, RP, .

and RPrl are exchanged, and in step 3108,
Verify that the machine has completed one full cycle by checking whether CN exceeds the maximum value. When all cycles are completed, CN is set to "1" in step 5109 for monitoring the next cycle.

In this way, monitoring continues until a timeout occurs in step S96.

Modification> In the above embodiment, as shown in FIG. 6, a new cycle number is assigned to the input/output cover turntable each time the output cover turn changes. Even if the cycle numbers are the same, the input signal patterns may differ from each other.

In the above-mentioned fault diagnosis, if there are two or more different input signal patterns for the same cycle number in the reference pattern table, the last input cover turn DPI is compared with the input signal pattern RPI from the actual machine, I tried to find the faulty part. In this modification, if there are two or more different input signal patterns for the same cycle number, index the different DPIs of the input signal patterns one by one from the top, and compare this with the RPI. The system displays mismatched input signals as faulty parts. The instruction to index from the top is given from a predetermined key on the keyboard 15.

Further, as another modification, the reference pattern database stored in the memory cassette may be stored in a floppy disk. Further, the display device is not limited to a liquid crystal display, and may be a CRT device.

(Effects of the Invention) As explained above, the machine failure diagnosis device of the present invention diagnoses the machine based on a machine as a sequence control target, input information from the machine, and a preset sequence. a programmable controller that generates operation control information and controls sequence operations of the machine; and operation pattern information consisting of the input information and operation control information;
storage means for storing time-series information during normal operation of the machine; information retrieval means for retrieving the operation pattern information at the time of machine failure from the programmable controller; information on the operation pattern at the time of failure, and operation during normal operation. The present invention is characterized by comprising a difference detecting means for comparing the pattern information and detecting a difference between the two pieces of information, and a display means for displaying the difference information.

Therefore, since the device itself estimates the faulty part, humans are freed from relying on experience and intuition for fault finding, and more accurate and quick fault finding becomes possible.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the configuration of the present invention, FIG. 2A is an overall configuration diagram of the failure diagnosis system of the embodiment, FIG. 2B is a front view of the diagnostic unit 11, and FIG. 3 is a configuration diagram of the GC 13 of the embodiment. Fig. 4 is a diagram showing the connection state of the machine, PC, GC, etc. in the embodiment, Fig. 5 is a diagram showing the input/output signal specification menu, Fig. 6,
7 and 8 are diagrams explaining the configurations of the input/output cover turntable, the output cover turntable, and the reference pattern table, respectively; FIGS. 9A and 9B are flowcharts related to the control procedure for teaching; 12 and 13 are flowcharts of control related to failure diagnosis, respectively, and FIG. 11 is a diagram explaining the principle of failure diagnosis. In the figure, 1... Machine, 10... Programmable controller (PC), l 1... Diagnostic unit, lla, ll
b...LED, llc...Status display LED, lld
...Removal button, lle...switch, llf.
... Cable connector, 13... Graphic console (GC), 14... Liquid crystal display, 15... Keyboard, 16... Memory cassette, 17... Floppy interface, 18... Printer interface, 19...floppy disk drive, 20
...Printer, 21...Communication line, 30...CPU
, 31... ROM132... RAM, 33... Communication interface, 34... Keyboard interface, 35... Liquid crystal interface, 36...
It is a connector. Figure 2B 0000 ~ 0003 ・ ・ -r I 00000
4～0007-、、、・NI 000010 N
00 + 3-r I I I○774 ＾ノ 0
777-・-NNNOFigure 5Figure 9AFigure 13

Claims (1)

[Claims] (1) Generate operation control information for the machine based on a machine as a sequence control target, input information from the machine, and a previously set sequence, and control the sequence operation of the machine. a programmable controller; storage means for storing operation pattern information consisting of the input information and operation control information in chronological order during normal operation of the machine; a means for extracting information, a means for detecting a difference between the operation pattern information at the time of failure and an operation pattern information at the time of normal operation, and a display means for displaying the difference information. A machine failure diagnosis device characterized by: