Method for determining a periodic pattern in a line spectrum

• US 5,436,847 A
• Filed: 05/26/1994
• Issued: 07/25/1995
• Est. Priority Date: 05/26/1993
• Status: Expired due to Fees

• Current Assignee: Deutsche Aerospace AG (Mercedes-Benz Group AG)
• Original Assignee: Deutsche Aerospace AG (Mercedes-Benz Group AG)
• Inventors: Schmid, Johannes, Bauer, Christoph, Schroer, Wolfgang
• Primary Examiner(s): Ramirez, Ellis B.
• Assistant Examiner(s): Assouad, Patrick J.
• Application Number:
  US08/249,937
• Time in Patent Office: 425 Days
• Field of Search: 364/485, 395/900, 324/76.12, 324/76.19
• US Class Current: 702/76
• CPC Class Codes: G01R 23/16 Spectrum analysis; Fourier analysis
• CPC Class Codes: G01S 7/415 Identification of targets based on measurements of movement associated with the target
• CPC Class Codes: Y10S 706/90 Fuzzy logic

Abstract

The invention relates to a method for determining a periodic pattern in an amplitude spectrum, for example, in a Doppler spectrum of a radar echo. A residual spectrum is formed using Fuzzy Set Theory, and the periodic pattern is determined for each residual spectrum by histogramming the original spectrum. The method can be used for identifying aircraft, for example helicopters, that have periodically-moving components, e.g., rotors, that generate a periodic pattern in a Doppler spectrum.

Claims

1. A method for determining a periodic pattern in a line spectrum comprising the steps of:

forming an original amplitude line spectrum having a reference line and a plurality of spectral lines which are contiguous to the reference line;

forming a plurality of intervals in the line spectrum by combining a first predetermined number of contiguous spectral lines of the original amplitude line spectrum;

determining a mean value for each respective interval based on an amplitude of each spectral line combined to form the respective interval;

filtering the amplitude of each spectral line of the original amplitude line spectrum with a predetermined factor;

normalizing the filtered amplitude of each spectral line of each respective interval with the associated mean value of the respective interval;

categorizing the normalized amplitudes of the spectral lines into a plurality of overlapping normalized amplitude categories;

determining a periodic complement to each respective spectral line which is a continuation line that is at least contiguous to the reference line for each spectral line and forms a line pair with the respective spectral line;

categorizing each line pair in a relevance category as a function of a predetermined combination of the normalized amplitude categories, each relevance category having a predetermined relevance value range;

determining a relevance value for each line pair;

determining a maximum relevance value for all line pairs associated with one spectral line;

associating the determined maximum relevance value with the spectral line;

forming a residual amplitude line spectrum from spectral lines having an associated relevance value which is greater than a predetermined threshold value;

histogramming the original amplitude line spectrum to determine the number of continuation lines associated with each respective spectral line of the residual amplitude line spectrum;

determining the spectral line having the greatest number of continuation lines from the histogram; and

determining a modulation frequency of the periodic pattern based on the difference in frequency between the spectral line having the greatest number of continuation lines and the reference line.

2. A method according to claim 1, wherein the original amplitude line spectrum is digitized.

3. A method according to claim 1, further comprising the steps of thinning the original amplitude line spectrum by combining a second predetermined number of respectively contiguous spectral lines into a single spectral line in the line spectrum, and assigning each single spectral line an amplitude which equals a maximum amplitude of the second predetermined number of spectral lines combined to form the respective single spectral line.

4. A method according to claim 1, wherein the original amplitude line spectrum has a frequency resolution that is finer than the modulation frequency of the periodic pattern, and wherein the original amplitude line spectrum is represented by a linear scale.

5. A method according to claim 1, wherein the amplitude categories have an association function which is linear in a changeover region.

6. A method according to claim 1, wherein the relevance classes have an association function which is linear in a changeover region.

7. A method according to claim 1, wherein the amplitude classes in which the spectral lines of a line pair are categorized are combined by an AND operator, and a result of the combination by the AND operation is assigned to a predetermined relevance class.

8. A method according to claim 1, wherein the periodic pattern is determined based on Fuzzy Set Theory.

9. A method according to claim 1, wherein the step of histogramming includes multiplying amplitude values of the spectral lines by a weighing coefficient.

10. A method according to claim 1, wherein only spectral lines of the residual spectrum having a simple distance from the reference line and/or simple fractions of a distance from the reference line are histogrammed.

11. A method according to claim 1, wherein the periodic pattern is formed by an aircraft having periodically-moving components which modulate a radar signal.

12. A method according to claim 1, wherein the periodic pattern is formed by a helicopter.