Weight based occupant classification system

• US 7,475,903 B2
• Filed: 04/08/2005
• Issued: 01/13/2009
• Est. Priority Date: 04/08/2005
• Status: Active Grant

• Current Assignee: Robert Bosch Corporation (Robert Bosch GmbH), Robert Bosch GmbH
• Original Assignee: Robert Bosch GmbH
• Inventors: Watts, Russell C.
• Primary Examiner(s): Fleming; Faye M.
• Application Number:
  US11/102,054
• Publication Number:
  US 20060226641A1
• Time in Patent Office: 1,376 Days
• Field of Search: 280/735, 180/271, 180/273
• US Class Current: 280/735
• CPC Class Codes: B60R 21/01516 using force or pressure sensing means
• CPC Class Codes: B60R 21/0155 sensing belt tension

Abstract

A system and method of classifying an occupant in a seat. The system can include a plurality of sensors, and each sensor can be configured to produce an output related to a weight of an object in a seat of the vehicle. A controller can be configured to receive the output of each of the sensors when an occupant is in the seat, and perform a maximum likelihood ratio test to provide a signal indicative of the classification of the occupant. The maximum likelihood ratio test can incorporate the output of each sensor relating to the occupant, and can further incorporate data sets related to the outputs of each of the sensors in response to a plurality of test cases, for a specific class of occupants.

Claims

1. An occupant classification system for a vehicle, the system comprising:

a plurality of sensors, each sensor configured to produce an output related to a characteristic of an occupant in a seat of the vehicle; and

a controller configured to receive the output of each of the sensors and to perform a maximum likelihood ratio test to provide a signal indicative of the classification of the occupant, wherein the maximum likelihood ratio test incorporates the output of each individual sensor relating to the occupant, and further incorporates a mean and a covariance of each of a pair of data sets, wherein each data set is related to a specific class and includes the outputs of each of the sensors in response to a plurality of test cases.

2. The occupant classification system of claim 1, wherein the output signal of the sensors is indicative of a weight of the occupant in the seat.

3. The occupant classification system of claim 1, wherein the plurality of test cases includes various belt tension conditions and the maximum likelihood ratio test accounts for the various belt tension conditions on the occupant.

4. The occupant classification system of claim 1, wherein the plurality of test cases includes various belt anchor point positions and the maximum likelihood ratio test accounts for the various belt anchor point position conditions on the occupant.

5. The occupant classification system of claim 1, wherein the plurality of test cases includes various seat positions and the maximum likelihood ratio test accounts for the various seat position conditions on the occupant.

6. The occupant classification system of claim 1, wherein the controller is configured to utilize a threshold, said threshold filtering information from subsequent controller operations under predetermined conditions.

7. The occupant classification system of claim 1, wherein the controller is configured to be in communication with a vehicle system.

8. The occupant classification system of claim 7, wherein the vehicle system is an airbag deployment system.

9. The occupant classification system of claim 1, wherein the signal indicative of the classification of the occupant is used to distinguish occupants of test cases that have a class overlap.

10. A method for classifying an occupant in a seat of a vehicle, comprising:

providing a plurality of sensors, wherein each sensor provides an output related to a characteristic of an occupant in the seat;

obtaining a first data set including the outputs from each sensor under a plurality of test cases representing a first class of occupants, and further computing from the first data set a covariance and a mean;

obtaining a second data set including the outputs from each sensor under a plurality of test cases representing a second class of occupants, and further computing from the second data set a covariance and a mean;

obtaining the outputs from each individual sensor; and

using a maximum likelihood ratio test to classify the occupant, wherein the maximum likelihood ratio test incorporates the outputs from each individual sensor, the covariance and the mean of the data set relating to the first class of occupants, and the covariance and the mean of the data set relating to the second class of occupants.

11. The method of claim 10, wherein the covariance and mean relating to the first class, the covariance and mean relating to the second class, and the outputs from each individual sensor relating to the occupant are expressed as matrices.

12. The method of claim 11, wherein if the number of sensors is p, then the covariance matrix for the data set for the first class has dimensions of p by p, and the mean matrix for the data set for the first class has dimensions of 1 and p.

13. The method of claim 10, wherein the plurality of test cases representing a first class of occupants includes various belt tension conditions and the maximum likelihood ratio test accounts for the various belt tension conditions on the occupant.

14. The method of claim 10, wherein the plurality of test cases representing a first class of occupants includes various belt anchor point positions and the maximum likelihood ratio test accounts for the various belt anchor point position conditions on the occupant.

15. The method of claim 10, wherein the plurality of test cases representing a first class of occupants includes various seat positions and the maximum likelihood ratio test accounts for the various seat position conditions on the occupant.

16. The method of claim 10, further comprising filtering information from subsequent controller operations under predetermined conditions.

17. The method of claims 10, further comprising delivering classification information to an airbag deployment system.

18. An occupant classification system for a vehicle, the system comprising:

a plurality of sensors, each sensor configured to produce an output related to a weight of an occupant in a seat of a vehicle;

a controller configured to receive the output of each of the sensors and to arrange the outputs into a matrix, the controller further configured with a plurality of matrices of pre-stored data, at least a portion of the pre-stored data related to mechanical loading caused by occupant restraints, and to perform a maximum likelihood ratio test using the matrix of outputs of the plurality of sensors and the matrices of pre-stored data;

wherein the matrices of pre-stored data include a mean matrix and a covariance matrix related to a plurality of classes of occupants.

19. The occupant classification system of claim 18, wherein at least a portion of the pre-stored data is related to mechanical loading caused by belt tensioning systems.

20. The occupant classification system of claim 18, wherein at least a portion of the pre-stored data is related to mechanical loading caused by the configuration of seat belt anchor points.

21. The occupant classification system of claim 18, wherein at least a portion of the pre-stored data is related to mechanical loading caused by varying seat positions.

22. The occupant classification system of claim 18, wherein the controller is configured to utilize a threshold, said threshold filtering information from subsequent controller operations.

23. The occupant classification system of claim 18, wherein the controller is configured to be in communication with a vehicle system.

24. The occupant classification system of claim 23, wherein the vehicle system is an airbag deployment system.

25. The occupant classification system of claim 18, wherein the maximum likelihood ratio test is used to provide a signal indicative of the classification of the occupant.

26. The occupant classification system of claim 25, wherein the signal indicative of the classification of the occupant is used to distinguish occupants of test cases that have a class overlap.

28. The method of claim 22, wherein the ratio test provides a classification that distinguishes occupants of test cases that have a class overlap.

27. A method for classifying an occupant in a seat of a vehicle, comprising:

providing a plurality of sensors, wherein each sensor provides an output related to a weight of an occupant in the seat;

obtaining a plurality of data sets, the plurality of data sets including the outputs from each sensor under a plurality of test cases representing a plurality of classes of occupants;

obtaining the outputs indicative of the weight of the occupant from each individual sensor; and

using a maximum likelihood ratio test to classify the occupant, wherein the maximum likelihood ratio test incorporates the outputs from each individual sensor relating to the weight of the occupant, at least a portion of the data sets relating to the weights of the occupants of the plurality of test cases, and a mean matrix and a covariance matrix for each of the portion of the data sets related to the plurality of classes of occupants.