EP0331674A1

EP0331674A1 - Process for driving computer-controlled final control elements and computer coupled to a final control element.

Info

Publication number: EP0331674A1
Application number: EP87907300A
Authority: EP
Inventors: Jurgen Brauninger
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 1987-01-22
Filing date: 1987-11-11
Publication date: 1989-09-13
Anticipated expiration: 2007-11-11
Also published as: DE3764807D1; US5031177A; DE3701699A1; EP0331674B1; JP2609266B2; WO1988005496A1; JPH02501938A

Abstract

Dans les procédés de régulation d'un organe de commande couplé à un calculateur avec auto-vérification du calculateur, on a jusqu'ici procédé de manière que l'organe soit maintenu dans une position de référence déterminée et ne soit livré à la commande du calculateur qu'après la fin de l'auto-vérification de celui-ci. Dans les systèmes complexes, il s'ensuit cependant un retard important. Selon le procédé, l'organe terminal de la chaîne est maintenu uniquement pendant une première étape de l'auto-vérification et est régulé jusqu'à la fin d'une deuxième étape par un signal provisoire de régulation à moins qu'un signal d'erreur ne soit apparu entre-temps.In the methods of regulating a control member coupled to a computer with self-checking of the computer, steps have so far been carried out so that the member is maintained in a determined reference position and is not delivered to the control of the calculator only after the self-verification thereof has ended. In complex systems, however, there is a significant delay. According to the method, the end member of the chain is maintained only during a first stage of the self-verification and is regulated until the end of a second stage by a provisional regulation signal unless a signal d The error has not appeared in the meantime.

Description

- Λ -

Method for controlling a computer-controlled actuator and a computer coupled to an actuator

The invention relates _{# to} a method for controlling a computer-controlled actuator according to the preamble of claim 1.

Computer-controlled actuators are used in many areas of technology where the manipulated variable depends on numerous other boundary conditions, including setpoints and controlled variables.

The computers require a high computing speed and the ability to reliably process large data sets. This requires a large amount of memory on the computer side. Although it has been possible to reduce the failure rate of computer systems more and more in spite of increasing complexity, the possibility of a circuit-technical error has to be expected. While circuitry errors were usually immediately recognizable in the case of analog circuits which were customary in the past, this is often different in the case of errors in digital computer systems. There, an error, for example in a memory location for a rarely called program routine, can remain undetected for a long time and then occur completely surprised. This can result in control commands that cause considerable consequential damage.

In order to detect such errors at an early stage, it is customary to subject the computers to a self-test from time to time in the case of complex, safety-relevant computer-controlled actuators. Here, a test program run durch¬ be addressed in which all the modules in succession with vorge transferred data and it proceeds only in ^'real processing of the data after completion of self-tests in a main program. In the case of complex computer systems, such a self-test takes a certain time in which the computer is not available to carry out its actual task "in order to control the actuator in an uncontrolled manner. To avoid this time, it is held in a defined reference position and only released for control signals of the computer after the self-test has run for the first time or again.

One area of application for computer-controlled actuators is e.g. B. an electronic diesel injection for diesel engines. The computer-controlled actuator is arranged in a control unit which actuates the control rod for an injection device. The required position of the control rod can depend on the engine temperature, the fuel temperature, the air temperature, the torque and the speed.

The computer's self-test is expediently carried out at the start of each start-up process. However, it has been found that the time required for this can have adverse effects. The long start-up process can unsettle impatient drivers and lead to the assumption that there is a fault. Drivers who are already familiar with the starting behavior will be less willing to avoid the delay time during traffic jams, traffic light stops or stops before level crossings, for environmental reasons turn off the load. Frequent short-distance journeys also place a high load on the starter and the battery. If extreme cold or a weak starter battery is added, the long self-test time can lead to the energy provided by the battery for starting the engine being exhausted prematurely.

Long self-test times can also have an adverse effect in other areas, e.g. B. in computer-controlled actuators in production plants or in the drive technology of rail-bound public transport, watercraft or aircraft.

The invention has for its object to improve a method for controlling a computer-controlled actuator so that the disadvantageous effects of a time-consuming self-test for the control of an actuator are avoided without the safety gained by the self-test being restricted.

This object is achieved in a method according to the preamble of claim 1 by the features specified in the characterizing part. The invention makes use of the knowledge that the self-test of a computer-controlled actuator is passed as a rule, so that the time required for the self-test can subsequently be regarded as wasted. In order to shorten this period of time, the actuator is controlled with a preliminary control signal. This can e.g. B. can be chosen so that the actuator ensures a temporary function of the unit actuated by it. Because of the short time at which the provisional control signal is present, it can be accepted that this value deviates more or less from an optimal value. However, in order not to suffer any loss of security with these measures, the components of the computer which are responsible for carrying out the test themselves are subjected to the self-test in a first period of time controlled with the provisional control signal.

This measure also ensures that the subsequent second section of the self-test can be carried out in a controlled manner and that the provisional control signal can be switched off again in the event of an error message. With this measure, the delay time for a control can be shortened considerably. tion of the actuator, since the time required to test the memory space required for the self-test and the time to test the associated modules is considerably less than the time to test the entire remaining program memory. In practice, the method according to the invention can be used to achieve a reduction in the delay time until the final actuation of the actuator by 1/10 to 1/30 of the time required for the entire self-test.

Further developments and advantageous embodiments of the method according to the invention result from claims 2-11.

The invention further relates to a computer coupled to an actuator according to the preamble of claim 12.

In this regard, it is the object of the invention to improve a computer coupled to an actuator of the type mentioned above in such a way that the disadvantageous effects of a delayed release of the actuator for control signals of the computer are avoided without the security gained by the self-test being compromised - is restricted.

The invention solves this problem in a computer coupled to an actuator according to the preamble of claim 12 with the features given in the characterizing part.

With this configuration, the actuator can be controlled with a preliminary control signal before the self-test is completed. As a result, the time until the actuator is released for control signals from the computer can be bridged. In principle, the control signal can be freely selected. It is expediently chosen such that it temporarily ensures the functionality of the unit actuated by the actuator. A loss of safety is not associated with this measure, since after actuation of the actuator with the provisional control signal, the self-test can be continued and, in the event of a fault, the actuator can be switched over again so that it assumes the defined reference position.

The invention is explained below on the basis of exemplary embodiments which are shown in the drawing. - 0

Show it:

1 shows a flow chart for the course of a self-test in a computer-controlled actuator as part of a control unit for an electronic diesel injection in a diesel engine,

FIG. 2 shows a flow chart similar to FIG. 1, but additionally for the execution of a self-test for a monitoring circuit, and

3 shows a block diagram of a computer-controlled actuator according to the invention.

The flowchart in FIG. 1 begins at an initial state 10 of the computer, into which the computer z. B. was offset by a reset command after applying an operating voltage. In the next method step, which is designated by 12, the actuator is brought into a defined reference position and held there. This position corresponds to a zero quantity of diesel fuel. During this phase, the starter turns the engine without it being able to start. In a subsequent method step, which is designated by 14, a self-test routine of a first part of the program memory is run. The memory locations of the computer-side memory are checked for function in which the program part of the self-test program is stored.

Such a memory test can e.g. B. be carried out in such a way that the entire program memory is added up and compared with a comparison sum, or the comparison sum is recorded in a program memory cell and is selected such that zero results when the total program memory content is summed up.

A comparison for errors is then carried out in a method step denoted by 16. If an error is found, the computer is returned to its initial state, which is designated by 10. If no error is found, the method step designated 18 is carried out. It is a function test of other components of the computer, e.g. B. a read-write memory chers, a timer, or an analog-to-digital converter. After this test has been completed, an error check is carried out again at 16. In the event of an error, a reset command is issued which resets the computer to its initial state 10; if no error is found, the next method step follows.

The implementation of method steps 12, 14, 16, 18 and again 16 in the form of predetermined self-test routines corresponds to a first section of the self-test. In the subsequent method step 20, the actuator is now controlled with a provisional control signal. In the application example, this corresponds to a starting quantity of diesel which is sufficient to start the engine safely. A second section of the self-test is carried out while the engine is already running. This is considerably more time-consuming than the previous self-test carried out in the first section. In the second section of the self-test, a program routine is run through, in which the remaining program memories are checked for function. This is represented by method step 22. After this method step has been completed, an error check is carried out as symbolized by 16. In the error In this case, the computer is reset to its initial state; in the other case, the actuator for control signals of the computer is released by the computer going into the main program. This is symbolized by 24. The second section of the self-test ends when the test to be carried out at 16 is successful.

In a modification of the flow diagram shown, the second section of the self-test can also take place after the end of the test 16 following the method step 14 and the method step 18 can be assigned to the second section of the self-test. In this case, method step 20, by which the actuator is actuated with a preliminary control signal, would have to be shifted upwards.

In a further modification, the computer can be set to a state other than the initial state in the event of error messages which only appear after several self-test routines which have been carried out without errors. A state is expediently selected which corresponds to the last error-free self-test section.

Since during the first section of the self-test the If the engine is cranked by the starter, error messages can occur as a result of voltage dips, without causing damage to the computer. In this case, the method according to the modified execution prevents test routines which have already been successfully run through from being repeated and thereby wasting unnecessary time. However, this configuration also requires special control measures, which in turn require program storage spaces. Based on the consideration that such errors rarely occur anyway, the general resetting of the computer to the initial state is not to be regarded as disadvantageous.

In a further development of the invention, this fact is saved as a switching state when the computer is reset due to newly applied operating voltage. This storage takes place, for example, in method step 12. Before the provisional activation of the actuator in method step 20, a comparison 26 is carried out in this case, in which the switching state is evaluated as a switching criterion. The memory is then deleted in a method step 28. The memory is also erased if an error has already occurred before comparison 26 is reached. If it is now established in comparison 26 that the self-test routines have been successfully carried out immediately after the operating voltage has been applied for the first time, method step 20 is then carried out. In the other case, method step 20 is not carried out. Resetting the computer to the initial state on the basis of an error message therefore means that the actuator is held in the defined reference position during the subsequent self-test routines. In relation to the application example, this means that the engine receives zero quantity until the entire self-test has ended.

This measure prevents repeated error messages in the second section of the self-test from running through a loop in which the actuator is controlled again and again with the provisional control signal. In the application example, this would mean that the engine is cyclically injected with the starting quantity of diesel. Since the starting quantity is dimensioned much higher than the idling quantity, this would result in the engine starting up and over revving, which can lead to considerable damage. In this case, the safety aspect has priority over ^• a reduction in the time for actuating the actuating link with control signals.

In order to be able to monitor the security of a computer-controlled actuator even during operation, a monitoring circuit (watchdog) is often provided. This checks z. B. arithmetic routines or sync pulses and triggers a reset command in the absence of such signals. This prevents that the computer due to an external disturbance, z. B. is brought into a state by a voltage drop, voltage peaks or by electromagnetic interference, which he can no longer leave by himself.

If the monitoring circuit is to be included in the self-test, z. B. can be proceeded according to a flow chart according to FIG. 2. The computer is programmed in such a way that it omits an arithmetic routine or a synchronous signal, so that the monitoring circuit initiates a reset command.

In this case, a comparison 34 is carried out between method steps V2 and 14, by checking whether a reset command is issued by the monitored circuit has taken place. If this is the case, the computer goes into the main program 24, ie the actuator is released for control commands from the computer. However, this comparison is only carried out when all other self-test sections have been carried out successfully. The method steps are carried out in the same order as has been described in FIG. 1. If the self-test of the first and second sections was error-free, the actuator is first temporarily brought back into the defined reference position and held there in a method step 30. This corresponds to the delivery of zero quantity of diesel. In a subsequent method step 32, the wait for the arrival of a reset command, which must be triggered by the monitoring circuit. After completion of process step 12 in response to such a reset command, a check is then carried out at 34 as to whether the reset command has taken place through the monitoring circuit and has been transferred to main program 24. An exemplary embodiment of a computer-controlled actuator that is shown as a block diagram in FIG. 3 is finally explained.

It includes an actuator 34 for actuating a control rod of an electronic diesel injection and a computer 36. In a control line 38 of the actuator 34, a switch 40 is inserted, which is switched by the computer 36 via a control line 50 and can assume three states.

As long as there is no switching signal at switch 40, it is in the position shown. In this position it is at reference potential 42, which corresponds to a definition of a defined reference position. If the test routines of a first section are passed and if the initial state from which the computer started the self-test was due to the first application of the operating voltage, the switch 40 is switched to the second position. In this position, the actuator 34 is connected via a line 44 to a generator 46 which generates a preliminary control signal. In the application example, this corresponds to a starting quantity of diesel. Only when the computer has successfully completed all sections of the self-test does the switch 40 reach the third position, in which the actuator 34 is released for control signals from the computer 36, which it receives via a control line 48. The method according to the invention and the computer coupled to an actuator thus enable the actuation of the actuator 34 triggered by the self-test to be shortened considerably. The remaining time in which no injection of diesel takes place in the application example corresponds approximately to that Time that the starter needs to accelerate the engine to the starting speed. After the engine has started, the further self-test can then be carried out without potential faults due to voltage dips caused by the starter. In the course of this further self-test, any errors that occur lead to an immediate shutdown of the provisional control signal, that is to say the starting quantity of diesel. As a result, security is not compromised.

The measures achieved with the invention in the application example of the diesel engine thus contribute to the fact that the immediate starting behavior of the engine encourages the driver to switch off the engine during short stays and thus to reduce the environmental impact. ^{'Furthermore,} SERS increases the life of Anlas¬, relieves the battery, particularly at kenfahrten Kurzstrek- and the starting of the engine also enabled when the battery is low.

Claims

Patent claims

1. A method for controlling a computer-controlled actuator during a self-test of the computer, the actuator being initially held in a defined reference position during the self-test and released for control signals of the computer after passing the self-test, in particular for control an electronic diesel injection for diesel engines, characterized in that the actuator is held in place only during a first section of the self-test and is then actuated with a provisional control signal until the end of a second section of the self-test or until a previous error message is received.

2. The method according to claim 1, characterized shows that the first section of the self-test is formed at least from test routines for checking the program memory part for the self-test program.

3. The method according to claim 2, characterized gekenn¬ characterized in that the first section of the self-test from test routines for checking the program memory part for the self-test program and other assemblies of the computer other than the other program memory parts is formed.

_ •

4. The method according to claim 3, characterized in that the second section of the self-test is formed from test routines for checking the rest of the program memory part.

5. The method according to any one of claims 1-4, characterized in that the self-test in the event of an error message is repeated from an initial state of the computer from which the self-test was started and is then continued if there are no errors, and that Actuator is held in the defined reference position during the entire repeated and continued self-test.

6. The method according to any one of claims 1-4, characterized in that the self-test in the event of an error message is repeated from the last error-free test routine and is continued with no errors, and in that the actuator is repeated during the entire repeated and continued self-test in the defined reference position.

7. The method according to claim 5, wherein the computer is put into the initial state from which the self-test is started by each new application of the operating voltage, characterized in that the new application of the operating voltage is stored as a switching state and after error-free Sequence of the first section of the self-test is evaluated as a switching criterion for the provisional control signal, and that after actuation of the actuator with the provisional control signal or after an error message, the memory for the switching state when the operating voltage is reapplied is deleted.

8. The method according to any one of claims 3 - 7, characterized in that a third section of the self-test is formed from a test routine of a monitoring circuit (watchdog) in which the actuating is again held in the defined reference position, and that the actuator for control signals of the computer is only released when the monitoring circuit has passed the test.

9. The method according to any one of claims 1-8, da¬ characterized in that the control signals are released at the transition of the computer from a self-test program into a main program.

10. The method according to claim 8 or 9, characterized ge indicates that the control signals are released in a main program after a transition of the computer from a test program via a reset operation to the initial state.

11. The method according to any one of claims 1-10, characterized in that the defined reference position of the actuator corresponds to a zero quantity and the position assumed by the provisional control signal corresponds to a starting quantity of diesel fuel for injection into a diesel engine.

12. Computer (36) coupled to an actuator (34), by means of which the actuator (34) can be used between ^ -

ner defined reference position by a reference signal (42) and a release position for control signals of the computer (36), and during a self-test of the computer (36) is defined in the defined reference position, characterized in that the actuator (34) by means of of the computer (36) can be switched to a third position in which it can be controlled by a preliminary control signal, and that a control signal generator (46) is provided, to which the actuator (34) is connected in this position.