DE10115897C2 - Method and device for providing information for the analysis of faults in a technical installation - Google Patents

Description

The invention is directed to a method for extraction of information for the analysis of errors in a technical system, after detection of a Malfunction caused by the control or monitoring device an error message is sent to an operating unit and displayed there, as well as a device for Carrying out this method with at least one device for Control and / or monitoring of a system and with a loading Service unit for receiving and displaying error messages gene.

Automatic control and monitoring of machine Today, plants are indispensable in many industries been to high productivity while maintaining high Ensure precision requirements. However, one automatic control, however sophisticated, the occurrence of Never rule out accidents as they become more complex Plants also increase the probability of failure. All the way it is more important that the Found the cause as quickly as possible and effectively eliminated can be so that the downtime is limited or a Accumulation of errors is avoided. For this purpose the device controlling / monitoring the state al potentially error-causing control and sensor signals saved at the time of the accident, so that by off read this information conclusions on the actual Cause are possible. On the other hand, the involvement causes technically trained maintenance personnel sometimes don't insignificant period of time, and not uncommon during that time searched for the fault by the operating personnel on site. If the supposed error was then corrected, the  Plant mostly started again in the hope that now can continue to be produced. However, if the real one The cause of the error has not been eliminated or only incompletely eliminated there is often another accident, and again the current sensor signals are saved, whereby the signal states at the first actual error be written and then arrive too late for that en maintenance personnel are irretrievably lost. This is then dependent on empirical troubleshooting, and Already from the supposed time reduction a considerable increase delayed restart.

DE 197 17 138 C1 describes a method and system for Archiving of process signals from a technical system known. For this purpose, the are recorded as a process signal stream Process signals in a number of buffers behind sets, each process signal from a selectable intermediate memory selected and based on the process signal The process signal type in one of these characteristics archiving memory is stored in chronological order.

DE 196 13 666 A1 describes a data processing system for Process data of a technical system known. Here is to Achieve transparent and flexible data access Archive system with an operator panel via a data server connected, based on one of parameters for data selection created request process data from the archiving system reads out and transfers them to the operator panel.

This results from the disadvantages of the prior art Problem initiating invention, a method and a pre to find direction, with which especially in short succession of the errors that occur are the triggering errors things and other information are preserved to be there draw conclusions about the behavior of the system can.  

This problem is solved with the invention Procedure in that, in addition to displaying the error mel an entry of the error with Zu on the control unit record information in a message archive and an assignment of the Error message about potential sources of error with the help of a stored network normal form takes place.

In contrast to the prior art, the invention is satisfied not with the temporary storage of the signal values at the time of the occurrence of the accident (so-called initial values) in the control unit but it sees the creation of a Archive or a list of all faults or errors that have occurred incidents that are relevant to the ac current values are added without the early values entered here would be overwritten. Such a thing The archive forms a source of information for the service personnel le, which is still an afterthought even after a long period of time analysis of the causes of errors or a statistical analysis cal evaluation makes it possible to identify particularly frequent sources of errors len to be able to recognize. According to the invention is advantageous the type of dependency of an error on potential mistake sources from a preferably within the scope of the operating unit stored normal network form determined. A Network normal form is a regulatory or other signal structure of the control and / or monitoring program and shows for one internally generated signal or one signal output, especially for each message, its logical and / or mathematical creation or education law the input signals. Such network normal forms are usually only in the actual control or over security devices stored, but can be used for loading acceleration of the flow of information and especially in the With regard to a total failure of a control or over monitoring device also in the control unit saved and therefore available at all times.  

It has proven to be convenient for the operator a filtering of the connected control and / or monitoring devices incoming messages For example, based on predefined message classes in order to enter messages from other distinguish between different messages. The one that maintains the message archive Control unit receives from the connected control tion and / or monitoring devices a variety of Messages that, for example, include status messages as well as error messages may include maintenance instructions, etc. To reduce Storage requirements for the message archive are planned that only "real" error messages are entered and other operating messages may be displayed as before or further processed in any other way. To one especially individual filtering of incoming messages allow different types of messages to classes can be summarized, for example according to the degree of the disturbance, where then the messages to be entered in the message archive are clear can be selected individually.

According to the invention it is further provided that when entering of an error message to be entered in the message archive other signals at the time the error occurred read out and archived together with the error message become. These sig usually referred to as initial values nale represent an image of the state of the system when the Malfunction and thus remain for later error analysis available.

The entries in the message archive can be of great advantage can be displayed on the control unit at any time. In order to the on-site maintenance technician is able to moved to all of the In to be able to access formations, especially early Here error messages, for example similarities of the error picture to be able to determine and then quickly identify the cause to be able to. This function compares with previous ones  Error messages can also be automated, so that e.g. may be the same when pressing a key or button previous error messages are listed and can be viewed are cash. This type of error handling can go even further supported by that after fixing a bug in the measures taken for the message archive be entered properly, so that by reading this Zu sentence information further conclusions about possible causes of the current error can be made.

Further advantages result from the fact that an advertisement of the faults that have occurred are possible in the form of a list. A such list forms an overview, which also re conclusions about the system components that are particularly at risk of failure allowed.

The idea of the invention permits further training in that for every disruption that occurs, causal networking with all potential sources of error including their status in The time at which the error occurred can be displayed. This also poses a measure to facilitate the localization of errors cause. Because in some plants there are different signals triggering the accident are assigned the same error messages arranges, for example: "Packing machine defective", and therefore it is partial, from this summary report by tracing against the network structure the real thing, the accident to be able to locate the triggering signal. Achieve here the initial values, which are also saved, are not inconsistent substantial contribution to the rapid finding of the exact Fault location.

Through a statistical evaluation of the type and frequency of the fault skill can be determined whether, for example, particularly frequently failing device parts are undersized or on one other deficiency suffer.  

If the result of the statistical evaluation by the Be service unit can be displayed, can be used for complex applications system components that are particularly at risk of malfunction, so that the errors in a confusing constellation start at the most likely trigger location can.

Finally, it is in accordance with the teaching of the invention that the Content of the message archive for further evaluation, e.g. via an interface to another computer device is transferable. This is a constant link to one Control center or the like possible, so that, for example, a maintenance technicians from a central location already a preliminary can carry out an investigation of the accident in order to, for example On-site service personnel call the rich to be able to give timely treatment of the malfunction. other On the other hand, it is also possible in this way to set criteria for a further plant optimization or information for further To collect dispositions.

A device for performing the Ver driving, with at least one device for control and / or Monitoring of a system and with an operating unit for Receipt and display of error messages stands out there by from that in the context of the control unit a memory or storage area for storing archived files Error messages including additional information is provided. By a corresponding here in the context of the control unit Storage space can be long-term Storage of the message archive ensured for years become.

Other characteristics, details, advantages and effects on the basis of the invention result from the following Description of a preferred embodiment of the Er and the drawing. Here shows:

Fig. 1 is a schematic signal flow diagram of the inventive method, and

Fig. 2 is an exemplary screen display when an error message.

In Fig. 1 the control and monitoring technical part 1 is shown an industrial plant. Various parts of the system are controlled by dedicated control devices 2 , 3 according to the production requirements. These are preferably programmable logic controls installed on-site in control cabinets, which ensure largely automatic operation of the relevant system parts. Possibly pure monitoring devices 4 can also be provided in order to monitor various production parameters by means of sensors and, if necessary, to report overruns or the like. These devices 2-4 usually work independently of one another and are exclusively coupled to a central control unit 5 , where, for example, the current status of the various parts of the system can be viewed from a central location in the production hall or a control room or the like. Transferred to this end, the control and / or monitoring devices 2-4 in regular Zeitabstän the corresponding request on the part of the operating unit 5 at the occurrence of special events and / or data 6-8 to the central control unit. 5 There, in turn, you can react to changed situations and individual parameters of the control devices 2-3 can be changed or other functions can be activated or certain system parts can be stopped. Here, the current data 9 are made visible to the monitoring or control room personnel on a screen 10 .

When a fault occurs, the control unit 2 , 3 in question automatically processes an error handling program, preferably shuts down the system part in question or operates it with reduced power, switches over to an emergency unit or the like, and at the same time an error message 6 , 7 is sent to the Operating unit 5 forwarded and displayed on the screen 10 . Similarly Form 4 registered divisional also overruns made visible, in special cases, also to produce an addi tional error message on a screen 10 of a monitor.

As soon as an error message is visible on the screen 10 in the control room, a maintenance technician can go to the relevant part of the system in order to search and rectify the cause of the error. However, since a certain period of time necessarily passes, it cannot be ruled out that, despite an automatic countermeasure by the control device 2 , 3, further errors occur in the part of the system that is not functioning properly. It can therefore happen that at the point in time when the maintenance technician arrives on site, a completely different, secondary-triggered error message is already displayed or can be read out on the control device 2 , 3 in question, the original, all subsequent error messages following the cause of the fault was overwritten by the most recent error message. In this case, it is hardly possible for the technician to find and rectify the actual cause of the fault in a short period of time; rather, he is dependent on a visual assessment of the system and has to examine it because the control device 2 , 3 in question does not further assistance can do more.

In order to counter this disadvantage, in the context of the system according to the invention, a memory 12 coupled to the operating unit 5 is provided, which has enough memory space 13 for a message archive. In this message archive 13 , all of the error messages arriving from the control and / or monitoring devices 2-4 and possibly selected by the operating unit 5 are stored.

Furthermore, the network normal forms of the control and / or regulating algorithms or other links implemented on all control devices 2 , 3 can also be stored in the memory 12 . The control units 2 , 3 are in a position to determine the part of the signal flow diagram relating to this error message when an error message to be stored occurs and to determine all the signal values relevant for this purpose and to store them as initial values. Once this has been done, the operating unit 5 requests the at least temporarily stored initial values recorded and at least temporarily stored by the relevant control device 2 , 3 at the time the error occurs and likewise enters them into the message archive 13 .

From the initial value states of the potential, error-triggering signals, the operating unit 5 can then recognize which part of the system caused the error message, and it can make this information visible on the screen 10 . Since all of this can be done in a fraction of a second, the maintenance technician in the control center can immediately narrow down the cause of the error before he goes to the relevant part of the system. On the other hand, he has sufficient time to examine the first error that has occurred since subsequent errors that occur at short intervals do not overwrite the relevant entry in the message archive. If the technician is on site, he can, for example, use a portable computer 14, for example with a radio interface, to connect 15 to the operating unit 5 and get the interesting data from the message archive 13 on the spot on the screen. If a screen is present, the relevant information 6 , 7 can also flow via the control device 2 , 3 there and be made visible.

Fig. 2 is a diagram 16 shows the screen surface of the screen 10, after a plurality of error messages WUR generated at short time intervals by the same controller 2, 3 to 17. As already stated above, the old or first error message 17 remains stored in the message archive 13 and can therefore be displayed simultaneously with all subsequent error messages 17 . Such a list can be seen on the screen display 16 at the top. In addition to the code 17 of the relevant error message, a date column 18 and a time column 19 are provided in order to be able to read off the exact time an error occurs. This information can be read from the archive 13 at any time later. From code 17 a certain message text 20 results, which is predetermined and gives a very rough overview of the situation. A status column 21 shows whether the relevant error is still pending, and an acknowledgment column 22 shows whether the cause of the error has been remedied by the operating or maintenance personnel and then acknowledged.

As can be seen from column 20 , the message texts 20 output in the case of an error code 17 are extremely poor and often give no indication of the actual cause of the error. For this reason, an additional function is provided in that further information on this error can be displayed by selecting a specific error line, for example in the example shown in the upper line relating to the first error at 13: 13: 39.17. These additional information representations can take place in different modes, for example according to windows 23-25 .

In the window display 23 , in addition to the already known information of the error code 17 and the malfunction point 18 , 19, a verbal comment 26 is output which, for example, contains the results of an internal or automatic error analysis performed by the operating unit 5 .

In the second representation according to the window 24 , a contact plan for the current error message 17 can be seen, which means that this error message is output when the flag 27 (M777.3) is high. This in turn results from the OR operation 28 of two flags 29 , 30 , of which the former, namely M555.1, has a fault 31 , ie a high level. In addition, as in the window 23 , a concretized message text 32 assigned to the flag M555.1 identified as the cause can also be output.

As an alternative to this ladder diagram representation of the signal flow graph triggering the error message, the same information can be seen in the window 25 as an instruction list. It is also known here from the operator column 33 that the flags M555.1 and M666.2 are linked to flag M777.3 by an OR link. The current states of the signals 34 are shown in column 35 , and from this as well as from a fault, in particular lightning symbol 36 in the line of flag M555.1, it can be seen that this was responsible for the output of the error message. It can be seen from comment 32 that production did not stop at this error message because there was no raw material (M666.2), but rather because the protective grille was not closed.

To generate this information, the operating unit 5 uses the network standard forms stored in the memory 12 and the message codes 17 and signal states 35 entered in the message archive 13 , which immediately after the error message arrives from the control device 2 , 3 concerned be requested before they can be overwritten by a subsequent error message.

As you can see, the arrival of the second error message after just under a minute is completely irrelevant, the operating unit 5 has now had enough time to determine the signals 28 , 30 in question and their signal values from the relevant control device 2 , 3 based on the network normal form to retrieve and save by storing in the message archive 13 for the following investigations.

Claims (12)

1. A method of providing information for the analysis of errors in a technical installation, wherein

• a) after detection of a malfunction, an error message ( 17 ) is sent ( 6-8 ) to an operating unit ( 5 ) and displayed ( 10 ),
• b) an error message ( 17 ) with additional information is entered in a message archive ( 13 ), and
• c) the type of dependency of an error message ( 17 ) on potential error sources ( 29 , 30 ) from a stored ( 12 ) normal network form of the technical system is determined.

2. The method according to claim 1, characterized in that an entry in the message archive contains additional information on at least the type, time ( 18 , 19 ) and an error message ( 17 ). 3. The method according to claim 1 or 2, characterized in that the operating unit ( 5 ) filters incoming ( 6-8 ) error messages ( 17 ) on the basis of predefined message classes in order to enter messages ( 13 ) into the message archive ( 13 ). 17 ) to be distinguished from their reports. 4. The method according to claim 1, 2 or 3, characterized in that upon receipt of an error message ( 17 ) to be entered in the message archive ( 13 ), the states ( 35 ) of potentially error-causing signals ( 34 ) at the time ( 18 , 19 ) of the Error occurrence read out ( 6-8 ) and archived ( 13 ) together with the error message ( 17 ). 5. The method according to any one of the preceding claims, characterized in that the entries in the message archive ( 13 ) can be displayed ( 10 , 16 ) at any time on the operating unit ( 5 ). 6. The method according to any one of the preceding claims, characterized in that a display ( 10 , 16 ) of the faults that have occurred in the form of a list ( 17-22 ) is possible. 7. The method according to any one of the preceding claims, characterized in that the causal networking ( 27 , 28 ) with all potential sources of error ( 29 , 30 ) together with their status ( 35 ) at the time ( 18 , 19 ) for each fault ( 17 ) ) the occurrence of the error can be indicated ( 10 , 16 , 23-25 ). 8. The method according to any one of the preceding claims, characterized in that a sta Statistical evaluation of the type and frequency of faults is taken. 9. The method according to any one of the preceding claims, characterized in that the result of the statistical evaluation can be displayed ( 10 ) by the operating unit ( 5 ). 10. The method as claimed in one of the preceding claims, characterized in that the content of the message archive ( 13 ) can be transmitted ( 15 ) via an interface to a further computer device ( 14 ) for the purpose of further evaluation. 11. Device for providing information for the analysis of errors in a technical system, with

• a) a control and monitoring device ( 2-4 ) for detecting a malfunction,
• b) a memory ( 12 ) or memory area ( 13 ) for storing
  • 1. a network normal form of the technical system and
  • 2. a message archive ( 13 ) for error messages ( 17 ) including additional information ( 18-22 , 26-36 ), and
• c) an operating unit ( 5 ) for receiving error messages ( 17 ) from the control and monitoring device ( 2-4 ) and for displaying the error messages ( 17 ), the type of dependency of an error message ( 17 ) on potential error sources ( 29 , 30 ) is determined from the stored ( 12 ) network normal form of the technical system.

12. The apparatus according to claim 11, characterized in that the entry of the error ( 17 ) with additional information in the message archive in the context of the operating unit ( 5 ).