Method and device for sensor monitoring, especially for ESP system for motor vehicles

• US 6,427,102 B1
• Filed: 08/01/2001
• Issued: 07/30/2002
• Est. Priority Date: 02/01/1999
• Status: Expired due to Term

• Current Assignee: Continental Teves AG & Co. OHG (Continental AG)
• Original Assignee: Continental Teves AG & Co. OHG (Continental AG)
• Inventors: Ding, Eve Limin
• Primary Examiner(s): Camby, Richard M.
• Application Number:
  US09/890,531
• Time in Patent Office: 363 Days
• Field of Search: 701/34, 701/29, 701/33, 701/32, 701/35, 701/76, 701/99, 701/101, 702/34, 702/185, 702/183
• US Class Current: 701/34
• CPC Class Codes: B60T 2270/413 Plausibility monitoring, cross check, redundancy
• CPC Class Codes: B60T 8/1755 Brake regulation specially adapted to control the stability of the vehicle, e.g. taking into account yaw rate or transverse acceleration in a curve road vehicle drive control systems for control of driving stability otherwise than by controlling a particular sub-unit B60W30/02
• CPC Class Codes: B60T 8/885 using electrical circuitry
• CPC Class Codes: G05B 9/02 electric

Abstract

The present invention relates to a method and a device for monitoring sensors that sense in each case individual reference variables or measured variables (A, B, C) relating to a process (32), which are especially suited for an electronic stability program (ESP) for vehicles. A particularly high degree of reliability is achieved by monitoring the course of the output signals of the individual sensors in a cyclical and sequential fashion in that analytical redundancies are produced from process reference variables and process measured variables (A, B) which are currently not being monitored by means of a multi-process model (31) for a normal operational mode, and a residuum (33) is formed of the redundancies along with the output signal being currently monitored. After evaluation of the residuum (36) and comparison with a threshold value (35), a fault signal (F) is generated when the residuum reaches the threshold value.

Claims

1. Method of monitoring a plurality of vehicle sensors which sense respectively associated reference variables or measured variables relating to a driving-dynamics process, comprising the steps of:

monitoring the course of an output signal from each vehicle sensor in said the plurality of vehicle sensors wherein said monitoring is conducted in a cyclical and sequential manner,

providing analytical redundancies for the process reference variables or process measured variables being monitored from currently non-monitored process reference variables and/or process measured variables of a current process on the basis of a multi-process model for a normal operational mode,

producing residua by subtraction of the provided redundant analytical redundancies from the process reference variable or process measured variable being currently monitored,

evaluating the residua by means of a residua evaluating function, and

comparing the evaluated residuum with a predetermined threshold value and producing an error message when the residuum reaches the threshold value for at least one predetermined monitoring time.

2. Method as claimed in claim 1, wherein at least one of reference variables and measured variables is a yaw rate, a transverse acceleration, or a steering angle.

3. Method as claimed in claim 2, wherein the method runs in parallel to a electronic driving stability program and deactivates the driving stability program when the fault message is generated.

4. Method as claimed in claim 1, wherein the multi-process model is formed from a number of partial models, by means of which the process reference variable or process measured variable being respectively monitored is reproduced on the basis of physical interrelationships of other physical quantities.

5. Method as claimed in claim 1, wherein the residua evaluating function in a transient process condition distinguished by a high rate of model inaccuracy is produced by an algorithm according to a ‘

minimum of all’

principle, wherein the smallest of all residua produced is compared with the threshold value, and wherein in a stationary process condition, the residua evaluating function is distinguished by a low rate of model inaccuracy, wherein an algorithm according to a ‘

majority principle’

is provided, wherein a mean value of the analytical redundancies is formed, and that analytical redundancy is used for producing the residuum to be compared with the threshold value which is the medium value of three redundancies with the least discrepancies from the mean value.

6. Method as claimed in claim 5, wherein in the case where the number of the valid models is smaller than three, the residuum is produced according to the principle ‘

minimum of all’ and

, in all other respects, the residuum is produced according to the ‘

majority principle’

.

7. Method as claimed in claim 1, further including the step of:

calculating and adapting the threshold values in dependence on a process condition by way of the process reference variables and process measured variables not being monitored and the wheel rotational speeds and the vehicle speed, wherein when a fault message is triggered when an unplausible sensor signal appears or a spurious fault message appears due to high rate of model inaccuracy.

8. Method as claimed in claim 1 of the output signals is conducted using monitoring times that are conformed adaptively to the current process condition and selected such that a brief failure of a sensor is tolerated.

9. Device for monitoring vehicle sensors which sense individual reference variables or measured variables relating to a driving-dynamics process, comprising:

a first device for calculating analytical redundancies for a currently monitored process reference variable or process measured variable from currently unmonitored process reference variables or process measured variables of a current process by means of a multi-process model for a normal operational mode,

a second device for producing residua by subtraction of the calculated, redundant analytical redundancies from the currently monitored process reference variables or process measured variable,

a third device for evaluating the residua with a residua evaluating function,

a fourth device for producing a threshold value, and

a fifth device for comparing the evaluated residuum with the threshold value and for producing an error message when the residuum reaches the threshold value for at least one predetermined monitoring time.