The invention relates to a method of analyzing a frequency domain based signal.

By analyzing time varying signals in the time domain it is known to transform the time domain signal to the frequency domain, which can e.g. be performed by using a Fourier analysis of the time domain signal, since an improved implementation of several data improving signal processing algorithms is hereby achieved.

This is among other things the case if the target of the signal analyzing is to identify repetitive signal frequencies, which contain the fundamental frequencies and / or the harmonics of the fundamental frequencies.

Such repetitive signal frequencies appear e.g. in connection with physiological activity such as heart pulse in humans or animals.

Experience shows that it can be quite difficult to determine fundamental frequencies in such a repetitive signal sequence using the known methods of analysis in the time domain as well as in the frequency domain.

The cause of these difficulties is that it can be very varying if it is the fundamental frequency or one or several of the harmonic frequencies, which are the dominating frequency components, as a result of the time varying behaviour of the signal suppliers during several repetitions, as well as the variance of the transfer function between the signal supplier and the transducer, which detect the signal, is often difficult to analyse. This analysis becomes more difficult if there are unwanted noise components in the measured signal.

It is on this background an object of the invention to improve the identification and the interpretation of the fundamental frequencies of a repetitive signal, which are transformed from the time domain to the frequency domain.

The object of the invention is achieved by a method of the type stated in the introductory portion of claim 1 , which is characterized in that it includes the following steps:

a) a minimum and maximum frequency of the signal, which the analysis are made from, are chosen b) a 1st frequency F, which is the smallest fundamental frequency, which it is desired to analyse for, is chosen. c) between the minimum and maximum frequency as defined in a) there are made X samples by a number of whole multiples of F. d) the values of the X samples are multiplied by e.g. logarithmic addition to form a value T e) a new frequency for analysis is selected, such as by multiplying the existing analysis frequency by ΔF f) step c - e is repeated until the desired frequency range is analyzed g) the calculated values of T can be transferred to a display unit such as a co-ordinate system, or can be used as basis for further signal processing. In this way the fundamental frequency of the repetitive signal can clearly be identified e.g. by visualization in a co-ordinate system.

In order to further clarify the identification and/or the visualization of the fundamental frequencies it is advantageous, if, as stated in claim 2, the following step is performed after step d):

d1 ) for each X sample, an Y sample is made, where the difference in frequency between the X and Y points is Vz F or an odd multiple hereof such as +1 VzF, - 1 V-F, +2VzF, -2VzF etc.

d2) the values of the Y samples are multiplied in order to form a value N.

d3) a value R is created as a relation between T and N.

d4) the calculated values of step g) are replaced with the values R.

A further preferred embodiment is as stated in claim 4 that the display unit is composed of an visualization in a co-ordinate system.

As stated in claim 5, the interval in the frequency domain is 25 Hz and the interval is divided into 128 points, corresponding to an interval size of 0,195 Hz. Hereby precise results are achieved, if e.g. determination of the heart pulse of a human or an animal is desired.

As stated, the invention also relates to an application. This application is further described in claim 6. The invention will now be explained more fully with reference to the drawings, in which

fig. 1 shows a block diagram of an arrangement in which the principles of the invention are included.

fig. 2 shows a measured signal in the time domain.

fig. 3 shows the signal from fig. 2 in weighted form

fig. 4 shows the signal from fig. 3 exposed to a Fourier analysis, while

fig. 5 shows the signal from fig. 4 exposed to an analysis according to the invention.

In fig. 1 is with 1 indicated a block, which is supplied with a time varying signal X(t), which can e.g. originate from a sensor, which measures physiological activity such as heart pulse.

The output signal from this block is shown in fig. 2 and indicated with 5.

The signal has got three peaks, which are indicated with 6, 7 and 8.

These three peaks originate from the heartbeat of a human or an animal and express the heart pulse.

As it is shown on fig. 2 there are a different distance between peaks 6 and 7 and peaks 7 and 8, which expresses the variation in time between the three heartbeats, which is normal in humans and animals.

The signal 5 is transferred to block 2, where the signal is weighted in such a way that it becomes adjusted for further processing.

This weighting, which itself is known, can e.g. be carried out in such a way that the signal components over time can be amplified with varying factors.

The signal from the block 2 is shown in fig. 3, where the signal peaks 10, 1 1 and 12 correspond to the signal peaks 6, 7 and 8, but in fig. 3 they are weighted.

The signal on the output of block 2 is supplied to block 3, where the signal is transformed from the time domain to the frequency domain, e.g. by a Fourier analysis.

The signal from the output of block 3 is shown in fig. 4.

This signal contains fundamental - as well as harmonic frequencies, and is shown in fig. 4.

By visual inspection of fig. 4 it is obviously not easy to determine the fundamental frequency of the signal.

In order to determine the fundamental frequency the signal is transferred to block 4, where an algorithm according to the invention extracts the fundamental frequencies of the signal from the output of block 3.

The signal at the output of block 4 is shown in fig. 5.

This signal has got two visible peaks, which are named 16 and 17.

The connection between the time domain signal shown in fig. 2 and the signal in fig. 5 is that the distance between the peaks 6 and 7 in fig. 2 correspond to a frequency shown by the peak 16 in fig. 5, while the distance between peaks 7 and 8 in fig. 2 correspond to the frequency shown at peak 17 in fig. 5.

The calculated frequencies, which are shown at the peak 16, will be equal to 1/(the time distance in seconds between the peaks and 6 and 7)

Example: If it is desired to measure a pulse from an animal or a human in the range 48 - 150 heartbeats pr. minute, which corresponds to the frequency range from 0,8 Hz to 2,5 Hz, an input frequency range is initially selected from e.g. 0,8 Hz to 25 Hz.

Subsequently a start analysis frequency is chosen of e.g. 0,8 Hz and a multiplication factor ΔF. The multiplication factor can e.g. be selected on the basis of the equation ΔF=(2,5/0,8)Λ(1/128), equivalent to ΔF = 1 ,0089, whereby the whole frequency interval from 0,8 Hz to 2,5 Hz can be carried out during 128 steps.

Afterwards the T and N values are determined as well as the results T/N for the chosen frequency range.

The result, which clarifies the fundamental frequency can e.g. be represented graphically in a co-ordinate system or used for further signal processing.