JPS6020058Y2 - Stable frequency signal removal device for frequency analyzer - Google Patents

Description

[Detailed Description of the Invention] The present invention relates to a processing device for frequency analysis results in a frequency analysis device.

Conventionally, as a device for extracting a periodic target signal from an underwater acoustic signal, there is a device that outputs a level for each frequency to obtain each power spectrum by frequency-analyzing underwater acoustic signals of the same color at regular intervals. Ta.

As shown in Figure 1, the signals obtained by frequency analysis include a signal that appears for a certain period of time and whose frequency or level is unstable, and a signal a that lasts for a long time and whose frequency level is stable. Generally, the latter includes many periodic and unnecessary signals such as electric power sources included in the environment.

Since conventional devices output the power spectrum of the frequency analysis result as is, two types of signals coexist, and unnecessary signals had to be removed from the power spectrum by human identification.

An object of the present invention is to provide a device that automatically removes long-lasting, frequency-stable signals that are assumed to be unnecessary signals from frequency analysis results.

In order to achieve the above objective, the present invention removes frequency-stable signals that last for a long time by normalizing using a power spectrum integral calculation circuit and a multiplier that frequency-analyzes underwater acoustic signals.

Embodiments of the present invention will be described below with reference to the drawings.

When the power spectrum level of frequency 11 time j obtained by frequency analysis is aij, and the integral calculation result of frequency 11 time j is Sij, Sij is generally expressed by equation (1).

5ij=Si, j-1・K+aij(1-K)
”・(1) 0<K<1
(2) Now, let's focus on a signal whose frequency is f and whose power spectrum level is time-related (af is constant).

At this time, equation (1) is expressed as equation (3), where k corresponds to the integration time.

If this signal continues for an infinite length of time, the integral calculation result will be expressed by equation (4) and converge to a constant value af.

Sf, j = Σaf (1-K) -Kj-'
-...(3)-1 Sf, oo=af (I K)! number of animals X K”=af
(4) J k=1 The effect of the integral operation is shown in FIG. 3, and the time constant for asymptotic to af is determined by the way the constant is taken.

Now, the frequency analysis results, ai, and j are integrated into the integral calculation result S.
If i, j are removed and the result is Ai, j, then the formula (5) "t 3 = ai * J /Siv j"・
(5) Applying this to a signal with a constant level af at frequency f, from equations (4) and (5), Afj = 1 ・
...(6).

However, it is assumed that the integration time is sufficiently long taking K into account.

Therefore, a frequency-stable signal that lasts for a long time has a level of 1.

The above configuration will be explained using figures.

An example of the power spectrum of the frequency analysis result of a certain band is shown in the first example.
It will look like the figure.

Regarding the power spectrum in FIG. 1, when the calculation of equation 51 is performed, as shown in FIG. 2, both the background noise and the signal at the constant level af reach a level close to l, and can be removed.

For target signals that are not constant in frequency and time, Sij takes a smaller value than ai and j, so Ai.

j is a level greater than 1.

FIG. 4 is a block diagram showing an embodiment of the present invention.
The converter is composed of a frequency analyzer 2, an integral calculation circuit 3, a reciprocal converter 4, and a multiplier 5.

An input signal C is converted into a digital signal by an A/D converter 1.

The output thereof becomes an input to the A/D conversion signal station wave number analyzer 2, and is subjected to frequency analysis.

The power spectrum e, which is the frequency analysis result, is a level at each frequency, and is integrated by an integral calculation circuit for each frequency according to equation (1).

The integral circuit g, which is the integration result, is an integral value of the power spectrum level for each frequency, and is converted by a reciprocal converter into an integral reciprocal signal, which is the reciprocal of the integral value.

This integral reciprocal signal is the reciprocal of the integral value of the power spectrum level for each frequency, and is input to the multiplier, (5)
It is multiplied by the power spectrum e as shown in the formula.

This multiplication is performed on values of the same frequency, and output signal 1, which is a value for each frequency, is output.

Here, by determining the constant K so that the integral operation result (integral value) does not converge to a certain level within the time period during which the target signal can maintain stability in terms of frequency, a frequency-stable signal that lasts for a long time can be obtained. can only be removed.

Figure 2 is a diagram showing the improvement effect of the present invention.
The stable sustained signal of frequency f shown in the figure is removed.

As explained above, the present invention removes a frequency-stable signal that lasts for a long time from the power spectrum of the frequency analysis result by multiplying the frequency analysis result by the reciprocal of the value obtained by integrating the frequency analysis result. effective.

[Brief explanation of drawings]

Figure 1 shows an example of a power spectrum obtained by frequency analysis using a conventional device, and Figure 2 shows the result of removing a long-lasting, frequency-stable, unnecessary signal from the example in Figure 1 using the present invention. FIG. 3 is a diagram showing the effect of integral calculation, and FIG. 4 is a block diagram showing an embodiment of the present invention. ■... A/D converter, 2... Frequency analyzer, 3... Integral calculation circuit, 4... Reciprocal converter, 5... ...multiplier.

Claims (1)

[Scope of utility model registration request] Frequency analysis means 2 that frequency-analyzes an input signal and outputs a power spectrum; Integration means 3 that integrates the power spectrum signal obtained by the frequency analysis means for each frequency component; For a frequency analyzer, comprising: reciprocal conversion means 4 for outputting a reciprocal component signal of an integral signal; and multiplication means 5 for multiplying the reciprocal component signal and the power spectrum signal for each frequency component. Stable frequency signal removal device.