METHOD AND DEVICE FOR RECOGNISING THE CONDITION OF VEHICLE OCCUPANTS

Abstract

The physical and/or mental condition of a vehicle occupant can be recognized on the basis of a BCG (ballistocardiograph) signal, which is obtained by means of a BCG sensor. The BCG sensor is an MEM sensor; a cross-correlation of the BCG signal with heartbeat parameters is carried out in an optimum filter, which heartbeat parameters are varied within predefined limits to find a maximum of the cross-correlation function; and probable peaks are located in a cross-correlation function found in this manner and the heart rate is calculated therefrom.

Claims

1-11. (canceled)

12. A method for identifying a condition of a vehicle occupant on the basis of ballistocardiograph (BCG) data, comprising: obtaining the BCG data from a BCG sensor, wherein the BCG sensor is a micro-electrical-mechanical (MEM) sensor; carrying out a cross-correlation of the BCG signal with heartbeat parameters in an optimum filter, wherein the heartbeat parameters are varied within predefined limits to find a maximum of the cross-correlation function; locating probable peaks in the cross-correlation function; and calculating the heart rate from the probable peaks.

13. The method of claim 12, wherein the heartbeat parameters include a plurality of heartbeat patterns that are generated by frequency variation of one or more predefined heartbeat patterns within natural heartbeat limits.

14. The method of claim 13, wherein the heartbeat parameters include a plurality of heartbeat patterns that are generated by frequency variation of a single predefined heartbeat pattern within natural heartbeat limits.

15. The method of claim 13, wherein a maximum of the cross-correlation function is found by short-term interval cross-correlation of the BCG signal with the generated heartbeat patterns.

16. The method of claim 12, wherein, after a maximum of the cross-correlation function has been found and before the probable peaks are located, the BCG signal is subjected to an adaptive window function.

17. The method of claim 12, wherein, after at least one of (a) the maximum of the cross-correlation function has been found and (b) an adaptive window function has been applied, and before the probable peaks are located, a separate parameter adaptation is carried out to optimize the peak amplitudes.

18. The method of claim 12, wherein located peaks are filtered to exclude unrecognized peaks from the calculation of the heart rate.

19. The method of claim 12, wherein the BCG sensor is a seat sensor.

20. The method of claim 12, further comprising determining a blood pressure of the vehicle occupant in addition to the heart rate.

21. The method of claim 12, wherein different seat damping and support coefficients are used to obtain the heart rate.

22. A system, comprising: a ballistocardiograph (BCG) sensor, wherein the BCG sensor is a micro-electrical-mechanical (MEM) sensor; and a computing device programmed to obtain the BCG data from the BCG sensor; carry out a cross-correlation of the BCG signal with heartbeat parameters in an optimum filter, wherein the heartbeat parameters are varied within predefined limits to find a maximum of the cross-correlation function; locate probable peaks in the cross-correlation function; and calculate the heart rate from the probable peaks.

23. The system of claim 22, the computing device further programmed to include in the heartbeat parameters a plurality of heartbeat patterns that are generated by frequency variation of one or more predefined heartbeat patterns within natural heartbeat limits.

24. The system of claim 23, the computing device further programmed to include in the heartbeat parameters a plurality of heartbeat patterns that are generated by frequency variation of a single predefined heartbeat pattern within natural heartbeat limits.

25. The system of claim 23, the computing device further programmed to find a maximum of the cross-correlation function by short-term interval cross-correlation of the BCG signal with the generated heartbeat patterns.

26. The system of claim 22, the computing device further programmed to, after a maximum of the cross-correlation function has been found and before the probable peaks are located, subject the BCG signal subjected to an adaptive window function.

27. The system of claim 22, the computing device further programmed to, after at least one of (a) the maximum of the cross-correlation function has been found and (b) an adaptive window function has been applied, and before the probable peaks are located, carry out a separate parameter adaptation to optimize the peak amplitudes.

28. The system of claim 22, the computing device further programmed to filter located peaks to exclude unrecognized peaks from the calculation of the heart rate.

29. The system of claim 22, wherein the BCG sensor is a seat sensor.

30. The system of claim 22, the computing device further programmed to determine a blood pressure of the vehicle occupant in addition to the heart rate.

31. The system of claim 22, the computing device further programmed to use different seat damping and support coefficients to obtain the heart rate.

Description

[0022] A description of exemplary embodiments on the basis of the drawings follows. In the figures:

[0023] FIG. 1 shows example signals which are acquired, generated, processed and/or calculated in the scope of the ballistocardiographic heartbeat recognition.

[0024] FIG. 2 shows a further example of signals which are acquired, generated, processed and/or calculated in the scope of the ballistocardiographic heartbeat recognition.

[0025] FIG. 3 shows a further example of signals which are acquired, generated, processed and/or calculated in the scope of the ballistocardiographic heartbeat recognition.

[0026] FIG. 4 shows a further example of signals which are acquired, generated, processed and/or calculated in the scope of the ballistocardiographic heartbeat recognition.

[0027] FIG. 5 shows a further example of signals which are acquired, generated, processed and/or calculated in the scope of the ballistocardiographic heartbeat recognition.

[0028] FIG. 6 shows a further example of various signals which are acquired, generated, processed and/or calculated in the scope of the ballistocardiographic heartbeat recognition.

[0029] FIG. 7 shows a block diagram to explain the ballistocardiographic heartbeat recognition.

[0030] The signals shown in FIGS. 1 to 6 are the following signals, which are simulated on the basis of realistic values:

[0031] FIG. 1 shows typical seat acceleration noise in a vehicle seat, having a mean amplitude of approximately 0.1 g during travel and over a time of 10 seconds.

[0032] FIG. 2 shows the amplitude of a heartbeat signal, as it arrives at a BCG seat sensor, having a maximum amplitude of approximately 0.0005 g and over a time of 1 second. The signal has the form of a damped cosine function, wherein the form of the damping is not relevant here.

[0033] FIG. 3 shows the sum of the signals acquired by the BCG seat sensor from FIGS. 1 and 2 over a time of 10 seconds.

[0034] FIG. 4 shows the autocorrelation function of the seat signal from FIG. 1, i.e. the correlation of the signal with itself, with the time and/or phase shift on the abscissa.

[0035] FIG. 5 shows the autocorrelation function of the heartbeat signal from FIG. 2.

[0036] FIG. 6 shows the absolute value of the cross-correlation of the sum of seat noise and heartbeat signal acquired by the BCG seat sensor.

[0037] The amplitude in the recognition process is dependent on two influences, namely firstly the noise procedure during the seat measurement (inherent noise and noise induced by the road and the driver) and secondly the length and amplitude (energy) of the heartbeat signal.

[0038] FIG. 7 is a block diagram to explain the ballistocardiographic heartbeat recognition in a motor vehicle. In block 1, a BCG signal is continuously detected by a seat sensor and supplied to a block 2 in intervals which correspond to short time intervals.

[0039] Block 2 represents an optimum filter, which carries out a short-term interval cross-correlation (CCF) of the BCG signal using various heartbeat patterns 3a, . . . , 3b, . . . , 3c, which correspond to different heart rates and which are generated by frequency variation of a predefined heartbeat pattern 5 within natural heartbeat limits. For this purpose, the optimum filter 2 also receives the respective adaptation frequency of the frequency variation performed in block 4.

[0040] The cross-correlation function obtained in block 2 is subjected in block 6 to an adaptive window function depending on the length of the heartbeat pattern and the measured BCG signal, in order to limit the computing effort.

[0041] After carrying out the adaptive window function, the absolute value of the cross-correlation of the measured sum of seat noise and heartbeat signal is obtained, as shown in block 7 and in FIG. 6.

[0042] For the signal form shown in block 7, a separate parameter adaptation is also carried out in block 8 to optimize the peak amplitudes, before, in block 9, the peaks are located in the signal form (according to maximum probability) and the time interval between adjacent peaks is ascertained, as illustrated in block 10.

[0043] The peak frequency, which represents the heart rate of the vehicle occupant, then results therefrom in block 11.

[0044] This heart rate is not a smooth signal and therefore requires a further filter (not shown) to eliminate outliers, which result from nonrecognition of peaks in the cross-correlation function. Such a filter can be a mean value filter or a Kalman filter with residual regulation, which are both capable of eliminating atypical measurements in real time.

[0045] On the basis of the smoothed heart rate of the vehicle occupant, his physical and/or mental condition can now be concluded, as is known per se.

[0046] The above-described method contains the two following essential method steps: firstly the frequency variation of a predefined heartbeat pattern and the maximization of peak amplitudes by means of correlation of different heartbeat patterns with a measured BCG signal; and secondly peak identification, peak location, and ascertainment of the heart rate.