Classifications

• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
  • G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
  • G01S7/52—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S15/00
  • G01S7/534—Details of non-pulse systems
  • G01S7/536—Extracting wanted echo signals

Definitions

• the present invention relates to methods which make it possible to subtract the noise specific to underwater vehicles in the acoustic reception systems fitted to these vehicles, in order to avoid drowning the acoustic signals received by the own noises generated by the vehicle.
• most torpedoes include a seeker comprising a sonar provided with an acoustic antenna formed by transducers intended to receive the noises propagating in the underwater environment.
• a seeker comprising a sonar provided with an acoustic antenna formed by transducers intended to receive the noises propagating in the underwater environment.
• the sonar works in passive mode, it detects the noise radiated by the target towards which the torpedo is heading, and when it works in active mode it detects the echoes of the pulses emitted by itself towards the target.
• the useful noises coming from the target are mixed with noises of various origins coming both from the underwater environment and from the torpedo itself. In fact, it radiates clean noises from the water, coming largely from engines and various electronic power equipment it contains. The level of these noises is such that they considerably hamper the operation of the sonar.
• French patent 82 15279 discloses a noise subtraction device which in principle makes it possible to create a zone of silence in a determined space by using a single noise sensor, microphone or accelerometer.
• this device works relatively poorly. This is precisely the case in torpedoes where the own noise comes from several sources of noise.
• the invention proposes a method for subtracting noise for an underwater vehicle, mainly characterized in that it comprises the following steps:
• FIG. 1 the partial diagram of a torpedo.
• This torpedo is propelled by a propeller 101 driven by a motor 102 powered by batteries 103.
• the motor 102 also drives a generator 104 which serves as a source of energy at 400 Hz for the other organs of the torpedo.
• This source supplies an acoustic base formed by a seeker 105 provided with a sonar antenna 106.
• the direction control signals delivered by the seeker are applied to servos 107 which make it possible to actuate the steering members of the torpedo.
• These servo drives are also powered by the generator 104.
• the sources of noise specific to the torpedo mainly come from the motors 102 and the generator 104, as well as from vibrations due to the excitation of certain parts by the magnetic fields at 400 Hz caused by the relatively strong currents which propagate in the connections. from the generator 104.
• means of measuring currents flowing are used as noise sensors, one on the power supply cable of the motor 102 and the other on the output cable of the generator 104.
• These measurement means are by example of current transformers similar to the ampere clamps used in electrical engineering.
• sensors can be supplemented by other sensors, such as microphones or accelerometers located at particular parts of the torpedo when the mechanical disturbances are the most important ⁇ .
• the signals from these other sensors improve the final result, provided that the correlation with the noise present in the signal picked up by the sonar hydrophones is sufficiently high.
• the treatment brings about a greater disturbance than the improvement that it is supposed to cause.
• the inventors noted during their experiments that, contrary to current opinion, the effect of the mechanical sensors was relatively weak and that most of the improvement, at least in the useful bandwidth for torpedoes, came from ammeter sensors.
• Figure 2 the block diagram of signal processing in the case where there is only one own noise sensor delivering a signal B (t) to be subtracted from the signal sonic reception signal S (t).
• the spectral lines obtained, f n to f, for S (t) and B (t) are separated in two devices 213 and 223 which can be simple series / parallel registers. Each spectral line is then correlated and subtracted with the corresponding spectral line of the other signal in a set of elementary correlation cells such as 201.
• a set of elementary correlation cells such as 201.
• spectral lines are recomposed in a device 202, for example of the parallel / series type, then subjected to an inverse Fourier transformation in 203 and finally finally converted into analog in a digital to analog converter 204 to obtain the signal S. (t) analog cleared of the noise component corresponding to B (t). If necessary, the digital signal could be used directly.
• Each elementary correlation cell operates in the manner shown in FIG. 3.
• a spectral line of index i there is at the output of the module 213 of FIG. 2 a signal S (fi) which represents the amplitude of the spectral line.
• S (fi) which represents the amplitude of the spectral line.
• a circuit 301 makes it possible to obtain the amplitude of the interspectrum by delaying one of the two signals by a duration equal to the duration of the sample subjected to the transform of
• the correlation product is then obtained by integration into a circuit 302 over this same duration.
• the noise spectral component is in turn normalized with respect to a value determined in a circuit 303 and then integrated in a circuit 304.
• the signal leaving the integrator 302 is divided in a circuit 305 by the signal leaving the integrator 304, to obtain the correlation ratio of S (fi) with B (fi).
• B (fi) is then multiplied in a multiplier 306 by this ratio and applied to a subtractor 308 via a door 307.
• This door 307 is opened by a signal from the divider 305, which authorizes the passage of the signal coming from the multiplier 306 if the correlation ratio is greater than a fixed threshold and prohibits this passage otherwise.
• the subtractor 308 makes it possible to subtract B (fi) from S (fi) to obtain the signal S, (fi) corrected at least partially with respect to the initial noisy signal. For this subtraction, B (fi) was brought back to a correct level compared to its influence in S (fi) by the multiplier 306 itself controlled by the previous circuits which determine this percentage (which represents the autocorrelation index ).
• the routes are classified arbitrarily in principle, but in practice we manage to start with the most important route. This is the least disturbed by the others, which is easy to spot empirically.
• the other routes are then classified according to this same criterion.
• the first channel B_, (fi) is treated in a device called self-normalizing spectrum, abbreviated to A. N. S. We get the output of this first A. N. S. a signal U_. (fi), which has little correlation with the others.
• the following channel B Cred(fi) is first applied to a device 402, called correlation filter, abbreviated CF. , which also receives the output signal from the first AN S. 401.
• the signal The output of this correlofilter is subtracted from the signal B mecanic(fi) in a subtractor 403.
• the result of this subtraction is applied to a second AN S. 404 which delivers the decorrelated output signal U 2 (fi) corresponding to B fatigue(fi).
• the following channels are treated in the same way, i.e. the input signal B. (fi) is applied to a C F. which also receives B 1 (fi) and whose output signal is subtracted from B. (fi).
• the signal obtained from this subtractor is applied to another C F. which receives the signal U, - (fi) and the output of which is subtracted from the output signal of the previous subtractor.
• This process is iterated step by step using the output of the previous subtractor to process it in a C F. with the next signal U (fi) and subtract the result from the output signal of the previous subtractor.
• a last step we have subtracted the previous signal U, - (fi) duly processed in an F., we apply the output signal of the last subtractor to an AN S. which delivers the signal U. (fi).
• This tree structure therefore makes it possible to obtain the set of signals U. (fi) to U (fi) corresponding to the proper noise paths decorrelated therebetween. Expressed in another way, we thus obtain a set of own noise signals orthogonal to each other.
• the signal E. is delayed by a duration substantially equal to that of the sampling of the Fourier transform and this delayed signal is multiplied by the non-delayed input signal.
• a third circuit 503 allows it to divide the input signal E ⁇ by the estimation of the spectral density obtained at the output of circuit 502, in order to obtain the output signal S-, which is a normalized value of E_. .
• FIG. 6 represents the diagram of the correlofilter 402.
• a signal S- of the type obtained by the spectrum orthonormor described above, is applied at the input of this correlofilter, and a signal E nieof the type of the non-standard spectrum signal corresponding to one of the signals B. in Figure 4.
• a first circuit 601 makes it possible to obtain the interspectrum between S-. and E Trust, that is to say the cross spectral density corresponding to the product of the spectral density of E. by the conjugate complex spectral density of E.
• This circuit functions substantially like the circuit 501 of FIG. 5, by producing the product of the two signals S- and E ⁇ instead of the product of S- with itself delayed.
• S- is also applied to a filter 602 whose complex transfer function is equal to the cross spectral density determined in the circuit 601. The output of this filter is a signal S notebook.
• a set of correlo-filters such as that defined above is used, which it can be seen that it performs exactly the function necessary for this subtraction.
• FIG. 7 This diagram comprises a set of Correlofilter s 702 cascaded.
• the signal S (fi) and the signal U 1 (fi) are applied to the first correlofilter.
• the output of the first correlation filter 702 and the signal U effet(fi) is then applied to the second correlation filter 703, and so on until the last correlation filter which receives the output signal of the penultimate correlation filter and the signal U ⁇ fi).
• the latter correlofilter then delivers the signal S, (fi) which is applied to the circuit 202 intended to supply the inverse Fourier transformer 203.

Landscapes

• Engineering & Computer Science (AREA)
• Computer Networks & Wireless Communication (AREA)
• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Radar, Positioning & Navigation (AREA)
• Remote Sensing (AREA)
• Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=9410879

Family Applications (1)

Country Status (2)

Cited By (1)

Citations (5)

• 1991
  • 1991-03-19 FR FR9103306A patent/FR2674346A1/fr active Granted
• 1992
  • 1992-02-28 WO PCT/FR1992/000190 patent/WO1992016853A1/fr unknown

Patent Citations (5)

Also Published As

Similar Documents

Legal Events