Data processing: vehicles – navigation – and relative location – Vehicle control – guidance – operation – or indication – Vehicle diagnosis or maintenance indication
Reexamination Certificate
2002-05-09
2004-07-20
Zanelli, Michael J. (Department: 3661)
Data processing: vehicles, navigation, and relative location
Vehicle control, guidance, operation, or indication
Vehicle diagnosis or maintenance indication
C701S035000, C701S029000
Reexamination Certificate
active
06766232
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a method for detecting errors in a motor vehicle, performance characteristics and information for characterizing the performance quantities in a motor vehicle being recorded over a preestablished period of time. The present invention also relates to a diagnostic device for the predictive detection of errors in a motor vehicle.
BACKGROUND INFORMATION
From the related art, it is known to carry out preventive maintenance on a motor vehicle based on the driving performance or the hours of operation. For this purpose, in the motor vehicle specific performance characteristics (e.g., driving performance or the operational life) are recorded over a preestablished period of time and are stored in memory. If the performance characteristics reach a preestablished value, then a check or an exchange of certain parts, components, and/or operational equipment of the motor vehicle is carried out. In the known preventive maintenance, reliance is placed on empirical values, which parts, components, and/or operational equipment need to be checked or exchanged, if certain performance characteristics have reached a preestablished value. These empirical values can deviate from the actual situation in the motor vehicle in some cases significantly. Thus, by way of example, it can happen that the effective parts, components, and/or operational equipment were not checked or exchanged, because the corresponding performance characteristics had not yet reached a preestablished value. The consequences are a defective motor vehicle, an unscheduled garage visit, and possibly even other subsequent errors due to the error arising in the motor vehicle. On the other hand, in preventive maintenance, the case can also arise that completely intact parts, components, and/or operational equipment are checked or exchanged, only because the corresponding performance characteristics reached a preestablished value. This leads to additional unnecessary labor and costs.
From German Published Patent Application No. 198 49 328, it is known to record performance characteristics in a motor vehicle over a preestablished period of time, and to store them. On the basis of the stored performance characteristics, after an error has occurred in the motor vehicle, it is possible to localize the error. However, this method only makes it possible to diagnose an error after it has already occurred. A predictive diagnosis, i.e., detecting an error even before it has actually occurred, is not possible using this method. Using the known method, an unscheduled garage visit and possible secondary errors due to the error that has occurred cannot be avoided in the motor vehicle.
From U.S. Pat. No. 5,528,516, a method is known for detecting errors in a complex system on the basis of observable performance characteristics. Cited as complex systems in which the known method can be used are, inter alia, complex vehicles such as a spaceship; its use in motor vehicles is not mentioned. In the exemplary embodiments, the known method for detecting errors in a computer network and in a satellite system is described. It is also noted to use the known method for the medical diagnosis of a patient's symptoms. In the described method, performance characteristics of the complex system are recorded and stored over a preestablished period of time. In response to the occurrence of a specific error, a performance characteristics model, to which the error is assigned, is generated from the recorded performance characteristics. Redundant or unnecessary information is eliminated from the performance characteristics model. On the basis of the reduced performance characteristics model, it is possible then to identity and localize an error that is occurring in the complex system. Error prediction is not possible using the known system.
Furthermore, the known method is used in each case for an individual complex system. No thought has been given to combining the performance characteristics models generated from a plurality of complex systems. This has the disadvantage that the performance characteristics models must be generated for each individual system to be diagnosed, and the results cannot be transferred to other complex systems in a simple manner.
From the described disadvantages of the related art, the objective of the present invention arises to make possible a predictive detection of errors in a motor vehicle, enjoying a high degree of reliability.
To achieve this objective, from a baseline of the method for detecting errors in a motor vehicle of the type cited above, the present invention proposes a method which is characterized by the following steps:
from the performance characteristics recorded in the motor vehicle before the onset of a specific error, a performance characteristics model is generated which is assigned to the error;
the performance characteristics model is described in a suitable form; and
the currently recorded performance characteristics are compared during the operation of the motor vehicle, with the descriptions of the performance characteristics models that are characteristic of the errors.
SUMMARY OF THE INVENTION
In a motor vehicle, performance characteristics are recorded over a specific period of time, which can be different from motor vehicle to motor vehicle. By performance characteristics are meant all the information which describes the condition of the motor vehicle and its environment. For example, this denotes the signals from sensors located in the motor vehicle. In addition, information is recorded for characterizing the performance characteristics, for example, the condition of systems, including error codes that arise, as well as date, time and/or location of the performance characteristics record. The recorded performance characteristics and information can be stored for the purpose of subsequent retrieval. The recorded performance characteristics are stored, for example, in the form of vectors, the individual vector elements corresponding to the values of the performance characteristics at specific points in time.
When a specific error has arisen in the motor vehicle, the error is identified. It can be, for example, the failure of a specific component or an unusual signal from a specific sensor. The identification of the error that has arisen takes place on the basis of the recorded performance characteristics and the recorded information for characterizing the performance characteristics in a manner that is generally known from the related art. From the performance characteristics recoded before the occurrence of the error, a so-called performance characteristics model is generated, which is assigned to the identified error,
Processing the recorded performance characteristics in order to identify the error can take place either in the context of an on-board diagnosis within the motor vehicle or outside of the motor vehicle in a garage. The performance characteristics model is stored, for example, in the form of a matrix, the individual matrix elements corresponding to the values of different performance characteristics at specific points in time. In particular, the time points before the occurrence of the error and those performance characteristics that are influenced by the error are observed.
Outside the motor vehicle, the performance characteristics model is then described using appropriate rules and/or mathematical functions (e.g, convolution). The description of the performance characteristics model aids in simplification and therefore in saving memory space and computing resources in a computer of the motor vehicle. Finally, the descriptions of the performance characteristics model are transmitted to the motor vehicle and there, during the operation of the motor vehicle, they are compared with the currently recorded performance characteristics.
During the operation of the motor vehicle, for the predictive diagnosis of errors in the motor vehicle, currently recorded performance characteristics are compared with the previously determined descriptions o
Robert & Bosch GmbH
Zanelli Michael J.
LandOfFree
Method for recognition of faults on a motor vehicle does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for recognition of faults on a motor vehicle, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for recognition of faults on a motor vehicle will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3221030