Method for determining the locations of at least two impacts
View Patent ↗A method for determining the locations of at least two impacts F 1 and F 2 on a surface using one or more sensors S i , i=1 to n, with n being the number of sensors, and the impacts F 1 and F 2 generating a signal being sensed by the one or more sensors, wherein each sensor provides a sensed signal s i (t), i=1 to n, with n being the number of sensors. To be able to determine simultaneous impacts of different amplitudes the method includes identifying the location x of one impact, and determining a modified sensed signal s i ′(t) for each sensor in which the contribution due to the identified impact is reduced and which is based on a comparison, in particular a correlation, of each of the sensed signals s i (t) and a predetermined reference signal r ij (t) corresponding to a reference impact R j at location j. The method can also be based on couples of sensed signals.
1. Method for determining the locations of at least two impacts F 1 and F 2 on a surface using one or more sensors S i , i=1 to n, with n being the number of sensors, the impacts F 1 and F 2 generating a signal being sensed by the one or more sensors, wherein each sensor provides a sensed signal s i (t), i=1 to n, with n being the number of sensors; and comprising:
a) identifying the location x of one impact, and
b) determining a modified sensed signal s i ′(t) for each sensor in which the contribution due to the identified impact is reduced and which is based on a comparison of each of the sensed signals s i (t) and a predetermined reference signal r ij (t) corresponding to a reference impact R j at a reference location j.
2. Method according to claim 1 , wherein the locations of the at least two impacts on the surface define a trajectory of a continuous impact on the surface.
3. Method according to claim 2 , wherein the first and second impact correspond to successive positions on the trajectory.
4. Method according to claim 1 , wherein the sensor is an acoustic sensor.
5. Method according to claim 1 , wherein identifying the location x of one impact is of the strongest impact.
6. Method according to claim 5 , wherein the comparison is a correlation.
7. Method according to claim 6 wherein b) comprises:
b1) correlating each of the sensed signals s i (t) with the reference signal r ix (t), which is the reference signal, out of a plurality of reference signals r ij (t) corresponding to a series of reference impacts provided at a certain number of reference positions R j , which corresponds to a reference impact in the vicinity of the location x of a strongest impact,
b2) fitting each correlation product with a predetermined function and
b3) subtracting the fitted function from the correlation product.
8. Method according to claim 7 , wherein the correlating of each of the sensed signals s i (t) with the reference signal r ix (t) is performed by using the Fourier transform S i (ω) of the sensed signal s i (t) and the Fourier transform R ix (ω) of the reference signal r ix (t).
9. Method according to claim 8 , wherein b) further comprises, multiplying the results obtained with the Fourier transform R ix (ω) of the reference signal r ix (t) to thereby obtain a modified Fourier transform S i ′(ω) corresponding to a modified sensed signal s i ′(t).
10. Method according to claim 9 , and further comprising c) identifying the location of the next weaker second impact using the modified sensed signals s i ′(t) or its Fourier transform S i ′(ω), wherein c) comprises:
determining the correlations of the modified sensed signals s i ′(t) of each sensor with each reference signal r ij (t), and averaging the correlation products over the sensors for each reference signal r ij (t).
11. Method according to claim 10 , wherein a) to c) are repeated to identify the location of the next weaker impact, wherein in each time in a) the location of the impact which has been identified in the previous run is taken as strongest location x and in b) a new modified sensed signal is determined out of the determined modified sensed signal of the previous run.
12. Method according to claim 10 , wherein determining the correlation is performed by using the Fourier transforms S i ′(ω) and R ij (ω).
13. Method according to claim 10 , wherein averaging the correlation product is performed in the time domain.
14. Method according to claim 7 , wherein the reference impact is closest to the location x of the strongest impact.
15. Method according to claim 7 , wherein the predetermined function is a complex exponential function.
16. Method according to claim 6 , wherein b) comprises:
b1) correlating each of the Fourier transforms S i (ω) of the sensed signals s i (t) with the Fourier transform R ix (ω) of the reference signal r ix (t), which is the reference signal, out of a plurality of reference signals r ij (t) corresponding to a series of reference impacts provided at a certain number of reference positions R j , which corresponds to a reference impact in the vicinity of the location x of a strongest impact, thereby obtaining n correlation products S i (ω)_R ix *(ω), with R ix *(ω) being the complex conjugate of the Fourier transform R ix (ω),
b2) averaging the correlation products S i (ω)_R ix *(ω) over the n sensors,
b3) determining the phase of the average of the n correlation products S i (ω)_R ix *(ω),
b4) multiplying each of the correlation products S i (ω)_R ix *(ω) with the complex conjugate of the phase of the average to thereby obtain modified correlation products,
b5) subtracting the real part from the modified correlation products, and
b6) multiplying each of the remaining part of the modified correlation products with the phase of the average.
17. Method according to claim 16 , wherein the reference signal is closest to the location x of the strongest impact.
18. A method for determining the locations of at least two impacts F 1 and F 2 on a surface using one or more sensors S i , i=1 to n, with n being the number of sensors, and the impacts F 1 and F 2 generating a signal being sensed by the one or more sensors, wherein each sensor provides a sensed signal s i (t), i=1 to n, with n being the number of sensors and comprising:
a) identifying the location x of one impact, and
b) determining a modified signal based on the sensed signal s i (t) in which the contribution of the identified impact is reduced and which is furthermore based on a comparison of pairs of sensed signals s i (t) and s i′ (t) and pairs of corresponding predetermined reference signals r ij (t) and r i′j (t) corresponding to a reference impact R j at a reference location i.
19. Method according to claim 18 , wherein identifying the location x of one impact is of the strongest impact.
20. Method according to claim 19 , and wherein the comparison is a correlation.
21. Method according to claim 20 , wherein b) comprises:
b1) correlating the Fourier transforms S i (ω) and S i′ (ω) of two sensed signals s i (t) and s i′ (t) thereby obtaining a first correlation product S i (ω)_S i′ *(ω), with S i′ *(ω) being the complex conjugate of the Fourier transform S i′ (ω),
b2) correlating the Fourier transforms R ix (ω) and R i′x (ω) of the corresponding reference signals r ix (t) and r i′x (t), r ix (t) and r i′x (t) being the reference signals which correspond to a reference impact in the vicinity of the strongest impact x, thereby obtaining a second correlation product R ix *(ω) R i′x (ω) with R ix *(ω) being the complex conjugate of the Fourier transform R ix (ω),
b3) correlating the first and second correlation products to obtain a third correlation product S i (ω) S i ′*(ω) R ix *(ω) R ix ′(ω),
b4) averaging the real part of the third correlation product over all pairs of sensor couples and subtracting this average from the third correlation products to obtain P ii′x (ω).
22. Method according to claim 21 , further comprising c) identifying the location of the next weaker second impact, comprising:
multiplying P ii′x (ω) with the sum of R ix (ω) R im *(ω)+_R i′x *(ω) R i′m (ω) for all reference impacts R m , m=1 to P, with P being the number of reference impacts at reference locations m, and for all sensor couples S i and S i′ , with R im (ω) and R i′m (ω) being the Fourier transforms of reference signal r im (t) and r i′m (t), and R im *(ω) and R i′x *(ω) being the complex conjugates of Fourier transforms R im (ω)+R i′x (ω),
Fourier transforming each one of the products, and
summing the negative time part to the positive time part of each Fourier transform.
23. Method according to claim 22 , wherein c) further comprises averaging the obtained sums for all sensor couples S i and S i′ .
24. Method according to claim 23 , wherein the averaging of the obtained sums is performed in the time domain.
25. Method according to claim 21 , wherein in the vicinity of the strongest impact is closest to the location x of the strongest impact.
26. Method according to claim 25 , wherein a) to c) are repeated to identify the location of the next weaker impact, wherein each time in a) the location of the impact which has been identified in c) of the previous run is taken as strongest location x and in b) a new modified signal is determined out of the determined modified signal of the previous run.
27. Method according to claim 21 , wherein a) comprises
correlating the Fourier transforms S i (ω) and S i′ (ω) of two sensed signals s i (t) and s i′ (t) thereby obtaining the first correlation product S i (ω) S i′ (ω)*,
correlating the Fourier transforms R im and R i′m (ω) of two corresponding reference signals r im (t) and r i′m (t), m=1 to P, with P being the number of reference impacts at reference locations m, thereby obtaining second correlation products R im *(ω) R i′m (ω), R im *(ω) being the complex conjugate of Fourier transform R im (ω),
correlating the first and second correlation products to obtain the third correlation product S i (ω) S i′ *(ω) R im *(ω) R i′m (ω), and
averaging the correlation products over the couples of sensors S i and S i′ .
28. Method according to claim 27 , wherein the averaging of the correlation products is performed in the time domain.
29. Method according to claim 20 , wherein a) comprises:
determining the correlations of the sensed signal s i (t) of each sensor with each reference signal r ij (t), and
averaging the correlation products over the sensors for each reference signal r ij (t).
30. Method according to claim 29 , wherein determining the correlations is performed using the Fourier transforms S i (ω) and R ij (ω).
31. Method according to claim 29 , wherein the averaging of the correlation products is performed in the time domain.
32. Method according to claim 18 , wherein the locations of the at least two impacts on the surface define a trajectory of a continuous impact on the surface.
33. Method according to claim 18 , wherein the sensor is an acoustic sensor.
34. Computer program product, comprising one or more non-transitory computer readable media having computer-executable instructions adapted to perform the following:
a) identify the location x of one impact, and
b) determine a modified sensed signal s i ′(t), i=1 to n, with n being the number of sensors, for each sensor in which the contribution due to the identified impact is reduced and which is based on a comparison of each of the sensed signals s i (t) and a predetermined reference signal r ij (t) corresponding to a reference impact R j at location j.
35. A device for determining the locations of at least two impacts F 1 and F 2 on a surface comprising:
a processing unit, and
one or more sensors configured and arranged to transmit a sensed signal being the result of the at least two impacts to the processing unit,
wherein the processing unit is configured and arranged to:
a) identify the location x of one impact, and
b) determine a modified sensed signal s i ′(t), i=1 to n, with n being the number of sensors, for each sensor in which the contribution due to the identified impact is reduced and which is based on a comparison of each of the sensed signals s i (t) and a predetermined reference signal r ij (t) corresponding to a reference impact R j at location j.