TECHNICAL FIELD: The present invention relates to a method and a device for use in connection with tests in motor vechicles. The invention provides a simple, cost-efficient, and quick analysis of functions in motor vehicles.

BACKGROUND OF THE INVENTION: The degree of complexity in contemporary vehicles is constantly increasing.

Concurrently with this, the demand for systems which continuously can monitor various functions and partial functions in a vehicle, so called diagnosis systems, is also increasing. Previously known diagnosis systems are principally intended for detecting faulty functions and, in connection with this, suggest and possibly take action.

A known diagnosis system is shown in US patent 5 396 422. The device which is shown in this document appears only to be intended for handling detected errors in a vehicle, and for localizing causes for errors which have been detected. The device can not give any information about conditions concerning the operation of the vehicle which do not have any connection with errors which have ocurred during operation.

Another known diagnosis system is shown in GB patent 1 496 893. One drawback as regards this system is that it appears to generate very large quantities of data during collection of information. Similarly, it does not appear to be able to sort out false data which have been obtained when the conditions which are necessary for a test have not prevailed.

Thus, one deficiency as regards known diagnosis systems is that they only can indicate afterwards, in other words after an error has occured and has been detected by the diagnosis system, and that they generate large quantities of data during collection of information. An additional deficiency as regards known diagnosis systems is that they can not indicate how often the conditions have prevailed which makes a certain test possible. This means that in order for the majority of the tests to be allowed to be carried out,

certain operating parameters in the vehicle must prevail, for example engine speed, speed, temperature, or a combination of such parameters.

Irrespective of whether a faulty function has been detected or not, it can be of great importance to know how many times during a certain time period or distance driven that a certain test has been carried out succesfully in a certain vehicle.

Thus, there is a demand for a diagnosis system which, even before the errors have occured, can predict which errors are going to occur, and at which point in time. In this connection, such a system should generate a mangeable quantity of data, and it should be able to give information about how often various tests have been carried out during a certain time period or distance driven. Of course, a system which in this manner predicts errors must have a very high degree of reliability.

SUMMARY OF THE INVENTION: Thus, the problem which is solved by means of the present invention is to develop a method and a device which make it possible to in a highly secure way predict if, when, and with which degree of probability, various faults in a vehicle are going to occur. An additional problem which is solved by means of the present invention is to make it possible to predict errors either during driving or in connection with repair shop visits, with generation of manageable quantities of data, and with the possibility of being informed about how often a certain test has been allowed to be carried out during a certain time period or distance driven.

The object of the invention is achieved by means of a method and a device for use in connection with tests in motor vehicles, where the result of a test which has been carried out, irrespective of the outcome of the test, is stored and processed for analysis. Thereafter, the result of the analysis can be utilized in order to predict if and when a certain error is going to occur.

Due to the fact that each test only can be carried out when certain conditions prevail, a central processing unit in the vehicle makes a decision of whether the conditions prevail which makes it possible to carry out the various tests.

Prefarably, the results of the tests which have been carried out are stored in

the central processing unit in the vehicle. In order to reduce the demands for storage for the test results, compression of data which are obtained during the tests takes place during the processing for analysis.

According to the invention, after a test has been carried out, the test result is furthermore transformed to a scale which is common for all test results, which makes it easier when the degree of seriousness is to be compared between various test outcomes.

Furthermore, the number of times that each test has been carried out since the last resetting to zero is stored by a counter, which is advantageous since it, irrespective of the test result, often is of great interest to know if and how often a test has been allowed to be carried out during a certain time period.

In an alternative method of the invention, the analysis can also comprise estimation of the degree of probability with which various errors are going to occur. In this case, an alarm signal can be emitted if the degree of probability exceeds a certain predetermined value.

BRIEF DESCRIPTION OF THE DRAWINGS: The invention will be further described in the following by means of examples of embodiments and with reference to the annexed drawing, Fig. 1, which shows a rough flowchart of the invention.

PREFERRED EMBODIMENTS: In Fig. 1, a general flowchart of a method according to the invention is shown. The flowchart also shows the primary functions of a device according to the invention.

In the following, the invention will be described in connection with the flowchart in Fig. 1, and with reference to a number of tests in a motor vehicle.

It should be realised that the invention can be applied to almost any optional test in a motor vehicle, the tests which in the following are used in order to describe the invention are only examples.

For all tests in a vehicle, it is required that certain conditions must prevail in order for test results to be obtained. For some tests, it is the simple condition that the engine is running, and as one example of tests, for the carrying out of which other conditions are necessary, test for ignition failure in the engine can be mentioned. This is a test which defines how many ignition failures which have occured during a certain number of revolutions, for example each 1000th revolution, which means that the result of the ignition failure test only can be obtained each 1000th revolution.

Furthermore, measurement of leakage in the vehicle's tank system can be mentioned as an additional example of a test for which certain conditions must prevail in order for any results of the test to be obtained. Leakage measurement is a test which is carried out once per so-called"driving cycle", which in this case is defined as the time which passes from the moment when the vehicle's ignition is turned on and the vehicle's engine starts to run, to the moment when the vehicle's ignition is switched off. Driving occasion is hereinafter termed DCY,"Driving Cycle".

According to the invention, a control unit in the vehicle examines (box 10), in an initial phase, if the conditions prevail which results in that the test result from a certain test should be received.

When it has been decided that the result from a certain test should be received, this occurs (box 20). Due to the large amount of very differing tests in a vehicle, the test results will be on a large amount of different formats, from the simple format of error/not error to values on a scale with many possible outcomes. Thus, the value which is received is on a basic format which is specific for the test, hereinafter termed"Physical Ranking".

When"Physical Ranking"for a certain test has been recevied, the"Physical Ranking"is transformed (box 30) to a value on a scale which is common for all the tests in the vehicle. This is made in order to guarantee that one and the same type of equipment can be used for the further processing of the results of all the tests. The value which"Physical Ranking"is given on the common value is termed"Standard Ranking". If this common scale for example goes from 0 to 254,"error"would correspond to 254 and"not error"

would correspond to 0. On this scale, for example 128 could represent the level where a fault has reached such a level that action is required.

Irrespective of the outcome of a certain test, the test result is saved in the form of"Standard Ranking" (box 40), and is processed for analysis (box 50) in the vehicle. Suitably, the test results are stored and processed in a control unit or central processing unit in the vehicle. In order to make an analysis possible, each test result is time marked. The time indication is absolute, i. e. the time which has passed since the last power operation of the equipment or resetting to zero of the time metering is indicated, which can be seen as resetting to zero by a counter.

Due to the fact that the results from all the tests are processed and stored in the central processing unit, the results can also be used in connection with repair shop visits for prediction of which components in the car run the risk of breaking down before the next scheduled repair shop visit, which means that the components in question can be repaired.

In connection with repair shop visits, the results can also be read, and compilation can be made of the test results from a certain motor vehicle fleet, which provides the car manufacturer with excellent opportunities to take care of constructional defects.

In the above-mentioned processing of the test results, compression of data which have been obtained during the tests is included, the purpose of which is to keep the demands for storage capacity low. The compression is essentially made by obtaining certain statistical parameters from the test results.

Hereinafter, the parameters which are obtained from the test results for analysis are shown, as well as abbrevations for the parameters: -Latest Standard Ranking Value, LSR.

-The time when the latest LSR was received, LSRT.

-The highest Standard Ranking in this Driving Cycle, HSRdcy.

-The time when the latest HSRdcy was received, HSRT.

-The highest Standard Ranking since the last power operation of the equipment or resetting to zero of the device, WSR.

-The time when the last WSR was received, WSRT.

-Average Standard Ranking, SRLP.

-Standard Deviation, SRD.

-Trend in Average Standard Ranking, SRLPT.

-Time for the last estimated SRLPT, Trend in Standard Ranking, termed SRLPTT.

-Number of times that Standard Ranking Value for a certain test has been allowed to be received since the last power operation of the equipment or resetting to zero of the device, SRMRC.

All parameters are obtained after each obtained test result, separately for each test.

By means of these parameters, a large amount of information regarding the vehicle can be obtained before any fault has occured.

A limit value for SRLP can be statistically determined by means of SRD.

Together with SRLP, SRD provides a measurement of probability of the risk that a separate test will indicate faulty function.

As the name impies, SRLPT, Trend in Average Standard Ranking, indicates the trend in the calculated average values, SRLP. SRLPT is obtained by estimating the difference between the values on SRLP at certain predetermined test intervals. This means that SRLPT can be expressed as SRLP (t2)-SRLP (ti), where tu-tu is the interval that SRLPT is estimated over.

The interval can, for example, be each 100th test, each 200th, or any other optional value that is considered suitable. Of course, the interval can also be a time interval.

SRLPT, Trend in Average Standard Ranking, can be used in order to predict if and at which point in time a specific limit value for error indication will be exceeded, where the probability for a fault exceeds a defined risk. Thus, the estimations of SRLPT can also be used in order to, for example, emit an

alarm signal if the degree of probability that a certain error is going to occur exceeds a certain predetermined value (box 60).

The alarm signal, in turn, indicates to the driver or the repair shop mechanic that a certain component/partial function in the vehicle needs to be attended to within a certain time interval. In other words, the alarm signal does not need to be shown to the driver, but can possibly be made available only to the repair shop mechanics who have the equipment which is necessary for reading the test equipment. This can be utilized if the estimation of SRLPT indicates that a fault is going to occur, but with low probability and/or later on in the future.

Of course, even if SRLPT indicates that a limit value for error indication will not be exceeded for a certain test, SRLPT can be used in order to read the status of the function which corresponds to the test.

Preferably, SRD is used in order to define the statistical uncertainty in SRLP (box 70).

By means of SRMRC, it can be determined how many times a certain test has been allowed to be carried out, which can be very useful when attempts are made to establish why an error has occured, and when it was detected.

To have access to this parameter is of great value even if a test has resulted in approved value every time that the test has been carried out during a certain time period, which is due to the fact that the parameter also gives information about how many times the conditions which are necessary for the test have prevailed during the time period. For a large number of the above- described parameters, the same reasoning applies as well, that each parameter can provide a lot of information regardless of whether the respective test has been carried out with an approved result.

A device according to the invention can be realized in any optional combination of hardware and software, but in a preferred embodiment, the invention is accomplished by software in a control unit/central processing unit in a motor vehicle.

The invention is not limited to the embodiments which are described above, but may be freely varied within the scope of the appended claims.