EP0130179A1 - Method and device for monitoring processes - Google Patents

Abstract

In order to carry out process monitoring, respectively of machines controlled in real time, for example machine tools, the operating times, respectively the characteristic intervals of the process, respectively of the working phases of the machine, are stored. automatically in a processor and they are compared with threshold times, the latter being programmed in a process for learning the processor. In the event of inadmissible time differences, the processor can intervene directly in the process, for example to disconnect certain functions, or even the entire machine. An installation designed to carry out this process comprises a microprocessor (3) which is connected to the machine in question, respectively to its control device by data input channels (6, 7) for timing signals and which can act on the machine through an outlet channel (8). Other characteristic data of the process, such as pressure, temperature, etc., can be introduced into the processor and taken into account once transformed into numerical values. Thanks to an input-output unit (9), it is possible to record the inadmissible time differences between the timing signals and the threshold times, to display the instant states of the process progress and to enter the hand data.

Description

 Process monitoring method and apparatus

The invention relates to a method and a device for monitoring processes with simultaneous, successive and / or overlapping processes using a computer, in particular a digital microcomputer.

In known processes for process monitoring, characteristic process variables such as pressure, temperature or the like are measured and used to monitor the process.

The invention has for its object to provide a method and a device of the type mentioned, in which important times for the process are used for monitoring or diagnosis.

To achieve this object, a method according to the invention provides that the switching or interval times of a control device of the process are recorded, automatically stored and compared with target times, that the target times are based on specifications and / or taking into account the process recorded changes in time in the learning process (teach-in) are determined and stored in the computer, that after each process interruption, the target times are determined and stored again in the learning process, taking into account the new starting situation of the process, and that the process is initially, for some

Process cycles are operated exclusively during a learning phase based on a specified sample process and, at the same time as or before the end of the learning phase, a switch is made to monitoring mode.

The target times are preferably determined from a number of sample processes according to a statistical algorithm.

OMPI 2 00274

- 2 -

calculates and changes automatically during the process, e.g. B. by superimposing other characteristic quantities for the process, such as temperatures, pressures or the like.

A device for performing the method according to the invention is characterized in that a digital micro process computer is connected via digital data input channels for timing signals and a digital data output channel to a control device of at least one machine to be monitored and that at least one input / output unit is connected to the computer connected.

The input / output unit advantageously has a control unit and connected to it a printer for printing out deviations from the target times and a keyboard

Function keys for entering and actuating an alpha-numerical display of process data which cannot be automatically detected.

The control device can be controlled via a separate display and operating panel, which is expediently arranged directly on the machine or machines to be monitored, so that the operator in his vicinity finds all the information necessary for assessing the functional state of his machine.

The method and the device according to the invention allow for the first time a time diagnosis with statements about errors, disturbances and trends of the process to be monitored. /

The method and the device according to the invention are used with particular advantage for monitoring controlled machine tools or transfer lines. Here, z. B. Changes in feed and cutting times or in maintenance and repair

^OM '' Monitor times by continuously comparing the times mentioned with the specified setpoints during ongoing operation. Time deviations that exceed specified deviations are logged and displayed. In the event of a fault, reactions can be initiated automatically, e.g. B. certain machine functions can be switched off.

In the particularly preferred use case for machine tools mentioned, maintenance cycles and required tool changes can be automatically displayed to the operating personnel as soon as the permissible service lives have been reached. It is also possible to supply machine data in accordance with VDI regulation 342 or in general BDE data (BDE = operating data acquisition) for proof of utilization and failure statistics on demand or at predefined time intervals.

The invention is explained in more detail below with the aid of schematic drawings of an exemplary embodiment for a device according to the invention.

Show it :

Figure 1 is a block diagram of an apparatus according to the invention.

2 shows a block diagram of an input / output unit in a device according to FIG. 1;

3 shows an implemented front plate of the input / output unit according to FIG. 2;

Fig. 4 is a diagram for explaining the function of the device according to the invention;

FIG. 5 shows a descriptive illustration of the device according to FIG. 1 as a subsystem when used for monitoring several machines and

OMPI FIG. 6 shows a clear illustration of an overall system with several subsystems according to FIG. 5, an alternative being indicated by dashed lines.

According to FIG. 1, a microcomputer 3 is connected to a ring bus 1 via a bus coupler 2, the microprocessor known, not shown, namely a CPU, a ROM, a RAM, a C-MOS RAM and a clock with calendar contains.

Several such computers 3 can be connected to the ring bus 1, so that numerous machines (e.g. eight machines per computer, see FIG. 5) can be monitored in a large production hall, and the data provided by the computers can be compared or otherwise evaluated . Instead of one ring bus, several

Computer 3 can also be connected in a star configuration to a central data processing system.

The computer is connected to one (or more) machines to be monitored via a modular system of interface modules 4 to 8. 4 and 5 indicate two interfaces each for 16 analog input channels, via which measured values for non-time process variables such as pressures, temperatures o the like.. The ^"machine to be monitored in the computer are fed one.

The reference numerals 6 and 7 denote interfaces for 32 digital input channels each, via which time measurement signals such as switching or cycle times of the machine to be monitored or a control device therefor are fed into the computer 3 _^ . The processed signals are positive or negative switching edges of digital pulses.

It goes without saying that the analog incoming measurement signals are converted into digital signals in the interfaces 4, 5

OMPI must, so that they can be processed in the digital microcomputer 3, while such a conversion is not necessary for the digitally arriving timing signals to the interfaces 6 and 7.

The reference numeral 8 denotes an interface for 8 digital output channels, via which the computer can automatically act on the machine to be monitored or its control device, for example can switch off certain functions or the entire machine.

An input / output unit 9 is connected to the computer 3 via serial data channels 10, 11, via which data can be successively transferred from and to the computer.

2, the input / output unit 9 comprises an input / output control device 12 to which the following components are connected:

a status indicator 13, which indicates faults or irregularities on the machine to be monitored, for example by means of lamps;

a printer 14 which records time deviations from the setpoints preprogrammed in the computer;

an alpha-numerical display 15 for displaying entered and recorded values; ^.

a recorder 16 for the documentation of analog measured values or the like;

a function keypad 17 for manual input and retrieval of machine data not automatically transmitted to the computer;

WIPO a numeric keyboard 18 for keying in digits, for example for time setting, channel polling or the like.

The corresponding reference numerals are entered in the representation of the front plate of the input / output unit (FIG. 3) and are not explained again.

The described device works as follows:

All changes in time on the machine to be monitored are automatically carried out via the interfaces 6 and 7 or by manual input via the function keys 17 with high resolution, e.g. B. 10 ms, recorded as time measurement values. To this end, other typical process variables can be introduced over the interfaces 4, 5 and influence the target time specification in the computer in any desired manner. The duration and frequency of the recorded processes are determined and stored from the measured times.

The diagram according to FIG. 4 illustrates the chronological sequence of machine cycles when the setpoint is specified using the described device. In Fig. 4 and the following text, t means: time

A: Number of machine cycles ok: absolute upper limit = upper spontaneous value limit

(relative) oD: upper drift value limit (relative)

X _n : mean (calculated in computer 3)

Xn: measured value

X _s : setpoint (calculated in computer 3) uD: lower drift value limit (relative) uS: lower Sporitan value limit (relative) oA: upper absolute limit uA: lower absolute limit

OMPI In the computer 3, target times are first formed as follows in the learning process ("teach-in method"): at the start, the times for work steps to be monitored are measured over several machine cycles, and target values and permissible deviations, which are initially, from the mean values and scatter are specified, calculated. After a certain number of machine cycles, e.g. B. three, is automatically switched to monitoring mode ("diagnosis"), the work in the learning process over a number of machine cycles, eg. B. six, preferably is continued. The target times determined in this way or also specified during installation of the system are continuously compared with the measured times. Time deviations that go beyond predefined deviations are output via the printer 14 and are therefore available as a log.

The log essentially contains the following parameters:

- Number of signal changes

- signal duration

- Totalization of the signal duration

- target-actual comparisons

- Averaging the signal duration - Trend value formation

- Warning upper and lower limits

- Upper and lower switch-off limits

- "Emergency stop" limits

- degree of utilization

* r

The following is an example of a protocol for monitoring a machine tool for one working day:

Thursday, 10/22/83 8:27:14 AM Power On Turn on the

Systems

31 sec: 058-063 10ms: 125-127 - machine status

3 layers: 08:30, 12:30, 13:10, 17:30 Anze gen ⁱ

- ÜREA

OMPI 274

- 8th -

062 hours actual 088.5 rest 031.5 min 126 hours actual 230.4 min

127 Average 000410 msec: Diagnostics setpoints and tolerances

127 Trend target 000490 msec limits

127 Drift: -000190, +000240 msec

127 Spont: -000280, +000350 msec

127 Trend setpoint: 000490 msec: Past values of the diagnosis 1. 000440 msec channel

2,000,430 msec

3,000 450 msec

4. 000420 msec '

5,000410 msec 6,000,420 msec

7. 000400 msec

8. 000410 msec

Thursday, 22.10.83 12:16 14 3 Teach "teaching" by Di

Start gnos channels

Thursday, 10/22/83 12:16:46 PM 3 diagnostic start

Thursday, 10/22/83 12:18:22 PM 3 Teach-

The End -

Thursday, 22.10.83 13:14:27; drift tolerance - 127 drift: -000210 msec limit undershot

Thursday, 10/22/83 1:27:18 pm: Spont-Tolerance- 127 Spont: +000360 msec Grerfze overlapped

Thursday, 10/22/83 2:47:11 p.m.

063 St. Target time reached 120.0 min

Thursday, 10/22/83 2:47:13 PM; number of items reached

062 pcs. Target number 030 pcs reached

^< δU £ E

OMP A 062 measured value 00047 sec A 127 measured value 000 430 msec A 126 measured value 001940 msec A 063 measured value 00103 sec

Group statistics: 3 Thursday, 10/22/83 12:30 p.m.

03:41 machine switched on 03:10 automatic mode

00:05 03 The "Test" program is running

00:28 04 Tool change

00:10 05 Tool shortage

00:26 06 Material shortage 00:16 07 Electrical fault

00:00 08 Mechanics fault

00:24 09 Hydraulic fault

00:17 10 maintenance

.04: 00 Planning time 00:31 Total malfunction

00:19 machine rest time

79% utilization rate

The maximum of 256 input signal lines that can be connected in the largest standard device configuration can be used multiple times according to the definition when entering and dividing the channels. A signal input can thus be used simultaneously for the counting, measuring, testing and monitoring process. It is "possible, for example, to r the first flank of the machine tool

Use the switch-on signal and the rear flank of the switch-off or changeover signal of the work spindle control for the machining process for the following measurements:

BURE OP - Number of processing cycles

- Duration of the processing cycle

- Total processing times

- Determination of the average cycle time

- Trend statement about the cycle time.

In addition, the following data important for production can be determined or defined, e.g. B .:

- Number of workpieces manufactured - Number of workpieces still to be manufactured at a previously entered upper limit

- Monitoring the number of pieces of a lot size to be manufactured

- Detection of the tool engagement time, neglecting the start-up and run-down phases

- Determine the tool life

- Monitoring tool change intervals

- Monitoring of maintenance intervals.

If there are processes that are constantly recurring and that move within certain limits in terms of their timing, these can be used for the so-called cycle time analysis. Freely controlled motion sequences can be sensibly monitored, i. H. Forced operation or speed-controlled processes are excluded in most cases.

Due to the temporally high resolution of 10 msec and the algorithm implemented in computer 3, further statements about the function can be made by recording several successive processes.

In the example of the processing unit of a machine tool, these are e.g. B. the uniformity of the movement timing, the influence of the duty cycle on

OMPΓ__ the cycle time, the statement about the tool state in comparison to the "optimal" or set-up state, the monitoring of the correct functioning of guides, switches, switching valves, sensors or the like, which are involved in the movement time sequence, the determination of inadmissible machining allowances for the workpieces, the monitoring of the material differences in different workpiece batches.

All of these processes are monitored automatically in the automatic operation of the machine system.

The time measurement values are recorded during the "teach-in" phase, the device automatically determining the corresponding upper and lower limits from these time measurement values

(Curves aO, oD, uD, uS in Fig. 4) determined in which warned, the next cycle prevented or even the machine "emergency stop" is switched.

The monitoring mode is usually started when the system is released by the installer as working properly and begins to produce in automatic mode.

From the first working cycles, the device puts a "tie" on each state defined for the diagnosis (cf. FIG. 4), the limit values formed from the first 16 machine cycles, based on the mean value, the last 8 Measurements are carried out. The range of the permissible percentage error is selectable. However, this is only defined once in the input definition and is retained each time the diagnosis is started again, unless it is changed again by manual input. In addition to these limit values carried above the "current" mean value, the device also offers Possibility to specify limit values manually. These then have priority over the limit values determined automatically by the device.

5 _A ork defined for the diagnosis channels inner¬ half of the calculated Frontiers (there are six tolerance thresholds can be selected), so Will not intervene in the procedure (4 curves aO, undated, uD, uS, OA, uA in Fig.). Slow changes such as B. can be caused by temperature influences, 10 then do not lead to a "reaction" of the device as soon as they do not exceed or fall below the manually entered fixed values (Fig. 4).

The automatic monitoring of temporal limit values of a movement or function offers the possibility of preventive diagnosis. Irregularities can thus be detected, observed and examined early on by the device during the operation of a system. This makes it possible in many cases to intervene before the machine fails completely. This means that longer periods of downtime or even consequential damage can be avoided, which means a corresponding loss of production.

It can e.g. B. may occur that at Werkzeugs.chaden on a 25 unit tools of the other units are moved to "breakage" due to the lack of pre-processing before the machine is stopped. This can be avoided by monitoring with the device described if a corresponding time overshoot or overshoot r 3.0 is connected to the relevant unit or function.

If the total cycle time were monitored, this irregularity would not be able to be recognized. The result would then be either the production of rejects from incorrectly machined parts or, if the parts were not fully machined, a reproduction of the corresponding one

- ^ URE OMPI Processing step.

The advantage of diagnosis using cycle time analysis is above all the possibility of early detection of an occurring error and the possibility of intervention before a real failure. The detected misconduct with the corresponding output on the printer does not immediately determine the faulty part, but "circles" the fault. The monitored time channel contains malfunctions of various components, which can play a role with regard to the recorded time interval, e.g. B. Malfunction of a control valve, falling oil pressure, faulty accumulator, falling supply pressure, feeding of the guide due to insufficient lubrication, irregularities of proximity switches due to moisture, tool wear, etc.

In the event of a fault, the computer 3 can automatically initiate reactions and output them to the machine via the interface 8. For example, the entire machine can be switched off.

The device supplies machine data according to VDI regulation 3423 or generally BDE data for utilization evidence and failure statistics on demand by pressing the corresponding function key from the keypad 17 or at predetermined time intervals.

The operation of the device does not require any special knowledge because each function key of the keypad 17 is assigned a desired function specified above the key.

After the start of the monitoring operation ("diagnosis"), the device runs automatically in the respective state and

Q PI reports itself automatically in the event of deviations or at times when the operator has to carry out work such as tool change or maintenance.

The output of diagnostic or machine data is triggered by pressing the corresponding function key from the keypad 17. Entries are shown in the display 15 and can be corrected before they are entered into the computer.

The device described can be used advantageously for all processes or workflows in machines in which digital pulses are used for control, such as in machine tools, handling and assembly machines, transfer lines, test and test stands and the like. like.

It is also possible to monitor several machines independently of one another from one computer, e.g. B. with regard to quantity, service life, maintenance cycle, tool change and diagnosis. An example of the monitoring or diagnosis of eight machines M1 to M8 by means of a device according to FIG. 1 is shown in FIG. 5. The same reference numerals are used in FIG. 5 for the same components as in FIG. 1. These components and their interconnection are not described again.

In addition to Figure 1, note the following:

The input / output unit 9 is arranged separately from the computer 3 in the example in FIG. an additional desk housing in the master cabin. The maximum distance between the computer 3 arranged in the vicinity of the connected machines M1 to M8 and the input / output unit 9 should not be more than 100 m. The distance between the computer 3 , .. ... ,

- 1 5 -

and the respective machine should generally not exceed 25 to 30 m in order to avoid the risk of interference on the signal lines from the machine to the device.

An additional printer 23 is provided in order to convey the malfunctions that have occurred to the operating personnel in the area of the machine.

An operating panel 21 with input buttons for identifying a fault and a fault display is provided on each machine. The control panel 21 is connected to the computer.

6 shows an overall system in which a plurality of computers 3 1 to 3 are connected via a bus line 1 (dashed line) or in a star shape to an intermediate computer 25 for data concentration and storage. This intermediate computer has conventional input and output devices such as tape 26, floppy disc 27 and printer 28.

The intermediate computer is connected to a computer system 29 via V-24 interfaces.

Furthermore, several screen terminals 30- ~ 30 can be connected to the intermediate computer, so that access to machine and production data at different locations y. becomes possible, e.g. B. in the master's office, in the maintenance office, in the production planning office, in the personal service accounting office etc.

Additional possibilities of the overall system shown in FIG. 6 consist in the return of stored measured values as setpoints for the machine function. The one EP82 / 00274

- 1 6 -

richter can thus significantly shorten the setting phases of workpieces that have already been manufactured when setting up a new one. This possibility arises in the subsystem according to FIG. 5 only when the recorded and printed values of the "old" setting are presented on a corresponding data strip.

y-

Claims

PT / EP82 / 0027-17 claims

1. A method for monitoring processes with simultaneous, successive and / or overlapping processes with a computer, in particular a digital microcomputer, characterized in that the switching or interval times of a control device of the Processes are recorded, automatically saved and compared with target times, so that the target times are determined and stored in the computer on the basis of specifications and / or taking into account changes in the learning process (teach-in) recorded in the process, that after each Process interruption, the target times are again determined and stored in the learning process, taking into account the new start situation of the process, and that the process is initially operated for a few process cycles only during a learning phase, starting from a predefined sample process and simultaneously with or before the end of the learning phase on monitoring operation is switched.

2. The method according to claim 1, characterized in that the target times are calculated from several sample processes according to a statistical algorithm and are changed automatically. - i '

3. The method according to any one of claims 1 to 2, characterized in that several processes are monitored simultaneously.

4. The method according to any one of claims 1 to 3, characterized in that in addition to the times, other characteristic process variables are measured and processed in the computer to influence the target times.

5. Device for performing the method according to one of claims 1 to 4, characterized in that a digital micro process computer (3) via digital data input channels (6,7) for time measurement signals and at least one digital data output channel (8) with a control device (20) at least one machine (M1 to M8) to be monitored is connected and that at least one input / output unit (9) is connected to the computer (3).

6. The device according to claim 5, characterized in that analog data input channels (4, 5) are also provided for transmitting characteristic, non-temporal measurement data of the process to the computer (3).

7. The device according to claim 5 or 6, characterized gekenn¬ characterized in that the input / output unit (9) has a control unit (12) and connected to it a printer (14) for printing out deviations from the target times and a keyboard (17,18 ) with function keys (17) for

Entering and operating an alpha-numerical display (15) of process rates which cannot be automatically detected. ^r i

8. Device according to one of claims 5 to 7, characterized in that the or each machine to be monitored (M1 - M8) is assigned a separate display and operating panel (21) with a fault display and identification buttons for recording the fault.

OMPI

9. Device according to one of claims 5 to 8, characterized in that a plurality of micro-process computers (3), which accordingly downstream control devices (20), input / output units u. have, via an intermediate computer (25) for formatting,

Intermediate storage of data or the like is connected to a central computer (22).

10. Application of the method according to one of claims 1 to 4 and / or the device according to one of claims 5 to 9 for monitoring controlled machines, in particular feed and assembly devices, machine tools and transfer lines.

y t '