CN102183741B - A Method for High Frame Rate Unambiguous Positioning in Long Baseline Asynchronous Underwater Acoustic Positioning System - Google Patents

Abstract

The invention provides a method for realizing high frame rate unambiguous positioning in a long baseline asynchronous acoustic positioning system; a signal is in a combined pulse string form at a signal emitting end; a CW signal and a LFM signal are alternatively used by the combined pulse string so as to reduce interference among pulses without reducing the frame rate since the frequencies of the sequent pulses are different; a processing method of the received signal comprises the following steps: performing the detection, time delay and frequency estimation of a CW pulse and a LFM pulse respectively by using a parallel multi-channel self-adaptive wave trap set and a parallel copy correlator set at a signal receiving end. The frame rate can be improved to 0.1s order by using the combined pulse string signal form so that the positioning system can be greatly improved in comparison with the present mature technology at home and abroad in terms of the track measurement frame rate, the positioning precision, the use reliability and convenience. The corresponding signal processing method can be used for effectively inhibiting the channel string leakage and reducing the influence of the Doppler effect.

Description

A Method for High Frame Rate Unambiguous Positioning in Long Baseline Asynchronous Underwater Acoustic Positioning System

technical field

The invention relates to an underwater acoustic positioning technology, in particular to a distributed buoy array high frame rate underwater acoustic positioning method for the trajectory of an underwater non-cooperative high-speed moving target.

Background technique

Restricted by many unfavorable factors such as narrow bandwidth of underwater acoustic channel, slow signal propagation speed, multiple interfaces, and sound ray bending, it is much more difficult to locate and navigate underwater targets with high precision than land and air targets. The spatially distributed buoy array underwater acoustic positioning system provides the necessary technical means for high-precision three-dimensional positioning of underwater targets.

For underwater cooperation targets, the existing underwater acoustic positioning systems mostly adopt synchronous working mode ([1] THOMSON D, ELSONS. New generation acoustic positioning systems. Oceans'02MTS/IEEE, 2002, 3: 1312-1318), namely Before each measurement, the clock of the measured sound source (installed on the moving target) and the receiving system need to be synchronized. Although synchronous time synchronization can achieve very high precision, the synchronization system has high requirements on the stability of the sound source and the receiving system clock. The limited clock stability will not only lead to the accumulation of clock deviation, but also easily cause clock disturbance deviation due to external random strong pulse interference, which will affect the positioning accuracy of the system. The asynchronous working mode does not require accurate timing of the sound source and the receiving system, and can achieve positioning accuracy comparable to that of the synchronous system within the array ([2] KOZICK R J, SADLER B M. Source localization with distributed sensor arrays and partial spatial coherence. IEEE Transactions on Signal Processing, 2004, 52(3): 601-616), so the scope of application is wider.

For underwater targets with high-speed and complex movements, the positioning system must ensure a high enough trajectory sampling rate to describe its trajectory correctly. But limited by range ambiguity ([3] Liang Guolong, Yang Chun, Wang Yan. Application analysis of software anti-range ambiguity technology in asynchronous underwater acoustic tracking and positioning. Applied Acoustics, 2005, 24(6): 359-363), usually The trajectory sampling rate of the traditional positioning system is low (on the order of seconds), which limits the ability to track high-speed targets. Combining the two positioning methods of long baseline and ultra-short baseline ([4] QUAZI A H. An overview on the time delay estimation in active and passive systems for target localization. IEEE Trans Acoust Speech Signal Process, 1981, 29(3): 527- 533), the measurement frame rate can be appropriately increased without causing distance ambiguity, but the structure of a single buoy primitive is complex, and it is inconvenient to deploy and recover at sea.

Contents of the invention

The purpose of the present invention is to provide a long baseline asynchronous underwater acoustic positioning system, which can realize high frame rate trajectory sampling without generating distance ambiguity. Methods.

The purpose of the present invention is achieved like this:

At the signal transmitting end, the signal form is a combined pulse train; the combined pulse train uses CW and LFM signals interleaved, and the frequencies of the front and back pulses are different, so as to reduce the inter-pulse interference without reducing the frame rate;

At the signal receiving end, the processing method for the received signal is to use a parallel multi-channel adaptive notch filter group and a parallel copy correlator group for CW pulse and LFM pulse detection, time delay and frequency estimation respectively.

The core technical content of the present invention lies in the waveform design of the signal and the corresponding signal processing method.

The signal waveform adopts combined pulse trains, and CW and LFM signals are used interleavedly, and the frequency bands of different types of pulses do not overlap each other, or the frequency bands overlap but the polarity of the FM slope is opposite. There are N pulses in a pulse train, let Pi (i=0, 1, . . . , N) be the serial number of the pulse type. P0 is a CW pulse, and P1, P2, ... PN are narrow-band LFM pulses. The pulse P0 is used to measure the Doppler frequency shift generated by the high-speed movement of the target in real time, which is used to adjust the reference signal related to the LFM pulse in real time. The pulses P1 and P2 are used as the parity identification code of the pulse subsequence, and are used for the temporal and spatial correlation of the pulse signal information measured between different buoys. The measurement frame rate can be increased to the order of 0.1s without generating distance ambiguity and without increasing system complexity.

For this signal form, the parallel multi-channel adaptive notch filter group and the parallel copy correlator group are used for signal processing of CW pulse and LFM pulse detection, time delay and frequency estimation respectively, which can effectively suppress cross-channel leakage and reduce the effects of the Doppler effect.

In addition, in the case of non-consonic velocity profiles, the use of sound ray correction can obtain higher-precision depth positioning results.

Description of drawings

Figure 1 is a working situation diagram of the buoy array asynchronous underwater acoustic positioning system;

Figure 2 is a composition diagram of the measurement array of the buoy array asynchronous underwater acoustic positioning system;

Fig. 3 is several optional sequence structures of combined bursts;

Fig. 4 is a block diagram of a signal processing method;

Fig. 5 is a group of parallel multi-channel adaptive notch filters;

Figure 6 is a parallel copy correlator bank.

Detailed ways

Below in conjunction with accompanying drawing and specific embodiment this paper is described in further detail:

1. Composition of positioning system

As shown in Figure 1 and Figure 2, a basic positioning system consists of 4 radio-controlled hydroacoustic positioning buoys, including 2 vertical double hydrophone buoys and 2 single hydrophone buoys, forming a 6-element three-dimensional measurement array. The buoys are synchronized by GPS, and the buoy receiver receives the pulse signal emitted from the target sound source in an asynchronous manner, and records the time when the pulse front arrives after detecting the signal. GPS measures the geodetic coordinate position of the buoy in real time, and transmits parameters such as the position of the buoy and signal propagation delay to the display and control platform of the survey ship in real time by means of a radio communication link for positioning calculation. A basic formation can carry out real-time three-dimensional tracking and monitoring of targets within the range of 3km×3km×300m. If the measurement range needs to be increased, it can be realized by increasing the number of cascaded buoys, and the system expansion is convenient.

2. Signal waveform: combined pulse train

In the underwater acoustic channel, the upper and lower boundaries will cause multi-path delay expansion, and the delay expansion in the shallow water channel can reach as much as 300-400ms. The multi-path delay extension will cause mutual interference between the preceding and following pulses, which increases the difficulty of detection and estimation. Increasing the pulse spacing will reduce this interference, but will reduce the system frame rate.

A combined pulse train adopted in the present invention interleaves pulse signals of different frequency bands, and the frequencies of the preceding and following pulses are different, so that the inter-pulse interference can be reduced without reducing the frame rate.

Considering that the available frequency band of the underwater acoustic channel is narrow, the combined pulse train is designed to be composed of 12 pulses and a maximum of 7 types of pulses, and the combined pulse train repetition period is 2s. The different types of pulse frequency bands do not overlap each other, or the frequency bands overlap but the polarity of the FM slope is opposite. Let Pi (i=0, 1, . . . , 6) be the serial number of the pulse type. P0 is the CW pulse, and P1, P2, . . . P6 are narrow-band LFM pulses. The pulse P0 is used to measure the Doppler frequency shift generated by the high-speed movement of the target in real time, so as to adjust the reference signal related to the LFM pulse in real time. Pulses P1 and P2 are used as parity codes of pulse subsequences to facilitate the temporal and spatial correlation of pulse signal information measured between different buoys.

The sequence structure of the pulse train can be selected from three structures shown in Figure 3 respectively. These sequence structures can increase the frame rate by 12 times, and the trajectory sampling period is reduced to 0.166ms. Among them, the number of independent channels required by structure 1 is the least (only 5), but the anti-multipath capability is slightly worse than structure 2 and structure 3. When the channel bandwidth is relatively sufficient and the multi-path delay extension is serious, the pulse sequence structure can be changed to structure 2 or structure 3, and the structure 3 can resist multi-path extension on the order of seconds.

3. Signal processing algorithm

For the signal waveform mentioned in 2, the corresponding signal processing method is shown in Figure 4. For CW signals, the first step is to process the received signal using a parallel multi-channel adaptive notch filter bank, and the second step is to use an instantaneous frequency variance detector for CW pulse detection, time delay, and frequency estimation. For LFM signals, this process also consists of two steps: the first step is to process the received signal using a bank of replica correlators, and the second step is to search for the presence of LFM pulses and their arrival times in the output of the bank of replica correlators. At the same time, the reference sample of the copy correlator is corrected in real time according to the CW pulse frequency offset measured by the adaptive filter, which can avoid the deviation of the measured LFM arrival time due to the Doppler effect.

(1) Detection, delay and frequency estimation of CW pulse

In principle, the single-frequency CW pulse signal P ₀ can be detected and parameter estimated only by using a single-channel adaptive notch filter. However, although the LFM pulse signals P ₁ , P ₂ , ... P ₆ do not overlap with the P ₀ frequency band, when the LFM signal strength is high, the cross-leakage of the sideband spectral components will still make the envelope of the notch filter output fluctuate greatly , which can easily lead to false alarms. Add multiple notch channels to the single-channel adaptive notch filter, corresponding to the center frequencies of several LFM pulse signals respectively, to form a parallel multi-channel adaptive notch filter group, which can remove the interference caused by the LFM pulse and avoid the envelope fluctuations, thereby significantly improving the detection capability and frequency estimation accuracy of CW pulse signals.

The parallel multi-channel adaptive notch filter group is composed of multiple single-frequency notch filters connected in parallel, as shown in Figure 5. The quadrature reference input of each channel can be expressed as:

$\left\{\begin{array}{c}{x}_{\mathrm{ci}}\left(k\right)=A\cos \left({\mathrm{\ω}}_{i}k\right),\\ x_{\mathrm{the\; si}}\left(k\right)=A\sin \left({\mathrm{\ω}}_{i}k\right).\end{array}\right.$ i=0, 2, . . . N. (1)

Where: frequency ω _i is the center frequency of pulse signal P _i .

After the filter output of each channel is accumulated, it is differenced with the expected signal to obtain a residual, which is used to adjust each orthogonal weight.

The adaptive iterative process is:

u _si (k+1)＝u _si (k)+με(k)sin(ω _i k) (2)

u _ci (k+1)＝u _ci (k)+με(k)cos(ω _i k) (3)

$\mathrm{<span\; class="notranslate">\; \&epsiv;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

y _i (k)=u _si (k)sin(ω _i k)+u _ci (k)cos(ω _i k) (5)

Among them: u _si (k) and u _ci (k) are the orthogonal weights, μ is the step size, ε(k) is the residual, d(k) is the desired output, and y _i (k) is the actual output. The envelope and frequency of the CW signal can be calculated using weights after the adaptive filter reaches a steady state:

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; A</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; N</span>}}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; no</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

$\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; N</span>}}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; no</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; tg</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\frac{{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; tg</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\frac{{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

(2) Detection and delay estimation of LFM pulse

A parallel copy correlator bank is used for the detection and delay estimation of multi-channel LFM pulses. Typically a copy correlator takes a sample of the transmitted signal as a reference signal. In addition to additive white noise, the signal received by the hydrophone often has Doppler, multi-path and other interference, so it is difficult to achieve a true match between the reference signal and the input signal. In addition, the CW pulse signal may also leak into adjacent frequency bands, causing false alarms in the detection of LFM signals in the corresponding frequency band. Aiming at the above problems, countermeasures are taken respectively in the design of the copy correlator.

The Doppler effect will shift the peak position of the copy-related output and reduce the peak value, thereby affecting the detection performance and timing accuracy. Therefore, in the signal waveform design, in addition to the LFM signal with high timing accuracy, a CW pulse P ₀ is also added. It can not only be used as a timing pulse, but its frequency can also give the size of Doppler frequency deviation, which can be used as the basis for modifying the frequency of the reference signal of the copy correlator, so that the reference signal and the received signal have the same Doppler frequency deviation. This not only improves timing accuracy, but also helps to improve related processing.

Compared with the water surface, the underwater medium tends to have greater sound absorption. At the same time, the buoy is closer to the water surface, and the reflected sound from other channels has little influence. Therefore, the primary reflection on the water surface is the most important multi-channel interference, and other reflected sounds can be ignored temporarily. consider. Generally, the water surface reflection coefficient is close to -1, so the copy correlation peak of the water surface reflection sound is negative, while the copy correlation peak of the direct sound is positive, so when selecting the peak, you only need to select the largest positive peak, and the corresponding signal should be direct Acoustic pulses, which can effectively eliminate reflected sound interference.

In order to reduce the influence of CW pulse on LFM signal detection, the correlator channel of CW pulse (copy correlator 0) is added in the parallel copy correlator group. The output of each correlator performs peak comparison after peak selection. If the channel with the largest peak value is a copy correlator 0, it is a cross leak; if the channel with the largest peak value is another channel and the peak value is higher than the correlation peak detection threshold, it is determined that the pulse is valid. , record the pulse type and time delay.

The bank of parallel copy correlators is shown in Figure 6.

Claims (2)

1. realize that at long baseline asynchronous acoustic positioning system high frame per second do not have the method for fuzzy location for one kind, it is characterized in that:

At the signal transmitting terminal, signal form is the assembled pulse string; Said assembled pulse string intersects and uses CW and LFM signal, and the frequency of front and back pulse is different, under the condition that does not reduce frame per second, reduces inter-pulse interference; Said assembled pulse string forms assembled pulse string repetition period 2s by 12 pulses, maximum 7 types of pulse combined; Dissimilar pulse band non-overlapping copies, or band overlapping but chirp rate polarity is opposite; Pi is pulse pattern sequence number, i=0 wherein, 1 ..., 6, P0 is the CW pulse, P1, P2 ... P6 is arrowband LFM pulse, the Doppler shift that utilizes the real-time measurement target high-speed motion of pulse P0 to produce, and the reference signal of relevant treatment is made in adjustment and LFM pulse in real time; Pulse P1, P2 are as pulse subsequence odd even identification code;

At signal receiving end, disposal route to the received signal is respectively applied for detection, time delay and the Frequency Estimation of CW pulse and LFM pulse for adopting parallel multi-channel adaptive notch filter group and parallel copy correlator bank; For the CW signal, the first step is to use parallel multi-channel adaptive notch filter group to handle the signal that receives, and second step was to use instantaneous frequency variance detecting device to be used for detection, time delay and the Frequency Estimation of CW pulse; For the LFM signal, also comprise two steps: the first step is to use the copy correlator bank to handle the signal that receives, and second step was LFM pulse and the due in thereof that search exists in the output of copy correlator bank; Simultaneously, the reference sample of copy correlator is revised according to the CW pulsed frequency side-play amount that sef-adapting filter records in real time.

2. according to claim 1ly realize that at long baseline asynchronous acoustic positioning system high frame per second do not have the method for fuzzy location; It is characterized in that: in the detection of CW pulse, time delay and the Frequency Estimation; The single channel adaptive notch filter is set up a plurality of trap passages; The centre frequency of corresponding several LFM pulse signals constitutes parallel multi-channel adaptive notch filter group respectively.

Images