A multi-static sonar underwater target attitude angle identification method

Abstract

The present application relates to a kind of underwater target attitude angle identification method based on multi-base sonar, belong to target parameter estimation field.Multi-base sonar system transmits signal by transmitting array, then the echo signal of target is received by multiple receiving arrays, the target distance-angle dictionary is constructed by echo signal, the distance-angle dictionary of different receiving arrays is combined to form distance-angle joint dictionary, three attitude angle estimation values are obtained by solving distance-angle joint dictionary, and the optimal weighted linear data fusion method is used to process it to obtain optimal solution.The present application increases the information amount of target, and combines sparse feature extraction method, so that the target distance attitude angle obtained is more accurate, makes up the shortcoming that single-base sonar detection range is small and concealment is poor.The angle information obtained by each receiving array is fused by the method of data fusion to obtain optimal solution, and the identification precision of target attitude angle is improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of target parameter estimation, and relates to a method for solving target distance and attitude angle through target echo signal processing. BACKGROUND

[0002] Most of the conventional sonars are single base sonars with transceiver collocated, which can easily expose the position of the sonar itself when detecting the position of the target by active sound emission. The multi-base sonar system receives the sound at a point far away from the emission point, so that the concealment is better than that of the collocated node.

[0003] In the prior art, single base sonar is used to identify the attitude angle of the target. However, due to the small size of the carrier vehicle, the aperture of the transceiver collocated sonar is small, which leads to a small detection range and low identification accuracy of the target attitude angle under low signal-to-noise ratio. The multi-base sonar detection system can identify the target attitude angle simultaneously using multiple receiving arrays, thereby improving the identification accuracy. SUMMARY

[0004] Technical problem to be solved

[0005] In order to solve the problem of low identification accuracy of the target attitude angle by the single base sonar in the prior art, the present application provides a multi-base sonar underwater target attitude angle identification method.

[0006] Technical scheme

[0007] The technical scheme adopted by the present application to solve the technical problem is to construct a distance-angle dictionary and use the basis pursuit algorithm to obtain the distance and attitude angle of the target. Since the transmitter and the receiver are configured separately, the system detection range is increased and the concealment of the system is improved. In addition, the use of multiple receiving arrays to receive target echoes can improve the identification accuracy of the target attitude angle.

[0008] A multi-base sonar underwater target attitude angle identification method, characterized in that the multi-base sonar includes one transceiver collocated array and two receiving arrays, the transceiver collocated array transmits signals, and each receiving array receives target echo signals of different distances and attitude angles; the steps are as follows:

[0009] Step 1: Construct a target distance-angle dictionary:

[0010] For the target echo received by the transceiver collocated array, the target is rotated around its geometric center, and the target attitude angle is divided into N s angles, denoted as {θ1, θ2, …, θ Ns}, each angle θ i (i = 1, 2, …, N s ) contains N ra distance, each angle and distance has an echo signal x i , i = 1, 2, …, (N s *N r ), construct a dictionary about distance and angle;

[0011] Assume that the target echo constitutes a distance-angle dictionary D1, which is an (L*n) x (N s *N r ) matrix, the i-th (1≤i≤N s *N r ) column of the matrix is

[0012]

[0013] The distance-angle dictionary of the target is That is

[0014]

[0015] Similarly, the distance-angle dictionaries of the other two receiving arrays can be obtained respectively;

[0016] Step 2: Merge the distance-angle dictionaries of different receiving arrays to form a distance-angle joint dictionary D c , which is specifically expressed as:

[0017]

[0018] Step 3: The transmitting array and the two receiving arrays receive target echoes with a relative attitude angle of θ i , i represents the i-th rotation angle, and the distance is C j , in the distance-angle joint dictionary D c Sparse representation obtains sparse vectors α, β and η;

[0019] When solving the sparse expression of the target echo signal x in the dictionary D c , the L1 norm minimization method in the convex optimization theory is used for solving, that is, solving

[0020]

[0021] Where γ represents the weight coefficient, which changes with noise, and α is the sparse representation vector; solve When α reaches the maximum value at the i x j element, it is determined that the estimated value of the distance attitude angle of the target is (θ i , C j ); Similarly, the estimated values of the target distance attitude angle measured by the two receiving arrays are obtained from the vectors β and η respectively;

[0022] Step 4: According to the angle of the receiving array relative to the transmitting-receiving array, the attitude angle obtained by the receiving array is processed to obtain the attitude angle estimation value of the target relative to the transmitting-receiving array; due to the existence of measurement error, the three sets of attitude angle estimation values obtained from the transmitting-receiving array and the two receiving arrays are not completely the same, therefore, the optimal weighted linear data fusion method is used for processing to obtain the optimal solution.

[0023] Further technical solutions of the application: the transmitting-receiving array and the two receiving arrays are circumferentially distributed, and the angle between each other is 120 degrees.

[0024] Further technical solutions of the application: the optimal weighted linear data fusion method specifically is:

[0025]

[0026] Among them, The target attitude angle result is, The three sets of attitude angle estimation values are respectively (SNR) i , (SNR) j , (SNR) k The signal-to-noise ratios of the transmitting-receiving array and the two receiving arrays are respectively.

[0027] A computer system, characterized by comprising: one or more processors, a computer readable storage medium for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors realize the above-mentioned method.

[0028] A computer readable storage medium, characterized by storing computer executable instructions, the instructions being executed to realize the above-mentioned method.

[0029] Advantages

[0030] The multi-static sonar underwater target attitude angle identification method provided by the application, the multi-static sonar system transmits signals through a transmitting array, and then receives echo signals of a target through multiple receiving arrays, the echo signals carry a large amount of information reflecting the essential characteristics of the target, the information amount of the target is increased, and the sparse feature extraction method is combined, so that the obtained target distance attitude angle is more accurate, and the shortcomings of small detection range and poor concealment of the single-static sonar are compensated.

[0031] The application is based on a multi-static sonar system, uses multiple receiving arrays to receive target echo signals, takes the sparse expression of the echo signals in a dictionary domain as target attitude angle information, then applies a data fusion method to fuse the angle information obtained by each receiving array to obtain an optimal solution, and improves the identification accuracy of the target attitude angle. BRIEF DESCRIPTION OF DRAWINGS

[0032] The accompanying drawings are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings:

[0033] Figure 1 Flow chart of the method of the application;

[0034] Figure 2 Coherent characteristics of the range-angle target dictionary of different receiving arrays: (1) target range-angle dictionary of different receiving arrays: (a) T / R transceiving array; (b) R1 receiving array; (c) R2 receiving array; (2) coherent characteristics side view of the target range-angle dictionary of different receiving arrays: (a) T / R transceiving array; (b) R1 receiving array; (c) R2 receiving array; (3) coherent characteristics contour map of the target range-angle dictionary of different receiving arrays: (a) T / R transceiving array; (b) R1 receiving array; (c) R2 receiving array;

[0035] Figure 3 Coherent characteristics of the range-angle joint dictionary of different receiving arrays: (a) three-dimensional coherent characteristics diagram; (b) coherent characteristics side view; (c) coherent characteristics -3dB contour map;

[0036] Figure 4 Results of the target range-pose angle recognition of different receiving arrays: (a) T / R transceiving array; (b) R1 receiving array; (c) R2 receiving array;

[0037] Figure 5 Target pose angle recognition results obtained by using the data fusion algorithm;

[0038] Figure 6 Target pose angle recognition results curve with signal-to-noise ratio. DETAILED DESCRIPTION

[0039] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application. In addition, the technical features involved in the various embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0040] The embodiment of the present application adopts three base sonars, including a transmitting-receiving combined array and two receiving arrays. The transmitting-receiving combined array transmits signals, and each receiving array receives echo signals of targets at different distances and attitude angles to form a dictionary. Assuming that the transmitting-receiving combined array is taken as a reference, the attitude angle of the target is θ, and the echo signals of the target received by different receiving arrays are respectively obtained by using a basis pursuit algorithm under the dictionary to obtain distance and attitude angle information of the target. Since the transmitting-receiving combined array and the receiving array of the multi-base sonar system can provide measurement data, some redundant information is increased, and therefore information fusion technology is used to obtain a final target attitude angle Figure 1 The flow chart of the method of the present application is shown.

[0041] Step one: target distance-angle dictionaries D1, D2 and D3 of the transmitting-receiving combined array T / R and the receiving arrays R1 and R2 are respectively constructed.

[0042] The target and the center of each device of the transmitting-receiving transducer are located in the same horizontal plane. The target rotates 360° with the geometric center as the origin, starting from the 0° rotation attitude of the normal side of the transmitting-receiving combined array T / R, and each attitude angle contains N r distance, and the target echo of each attitude angle and distance received by the transmitting-receiving combined array T / R is x i (i=1, 2, …, (N t *N r )), and the target distance-angle dictionary is constructed from the target echo, which is an (L*n)×(N s *N r ) matrix. The ith (1≤i≤N s *N r ) column of the matrix is

[0043] ψ i =[x1(1, θ i ,C j ), …, x1(n, θ i ,C j ), x1(1, θ i ,C j ), …, x1(n, θ i ,C j ), …, x L (1, θ i ,C j )…, x L (n, θ i ,C j ) T (1-1)

[0044] Then, the target distance-angle dictionary received by the transmitting-receiving combined array T / R is that is,

[0045]

[0046] The target echo signals are also received by the receiving arrays R1 and R2, and the target range-angle dictionaries D2 and D3 are formed respectively from the echoes as follows:

[0047]

[0048]

[0049] Figure 2 is the target range-angle dictionary, and its coherence characteristic analysis. Since the two-dimensional dictionary is large, in order to reflect the characteristics of the two-dimensional dictionary, only a part of it is taken for display.

[0050] It can be seen from FIG. (1) that the range-angle dictionaries formed by each receiving array are different, which shows that the target echoes received by different receiving arrays from different directions are different. It can be seen from FIGS. (2) and (3) that the main lobe of the target dictionary coherence characteristic of the submarine model is very narrow, and the side lobe is not very high, which shows that the similarity of the target dictionary is very low, which meets the requirements of constructing the dictionary.

[0051] Step two: the dictionaries formed by the transmitting-receiving array T / R, the receiving array R1 and the receiving array R2 are combined to form a range-angle joint dictionary D c , which is specifically expressed as formula (1-5)

[0052]

[0053] Figure 3 is the coherence characteristic diagram of the range-angle joint dictionary constructed by the three different receiving arrays. It can be seen from the diagram that the coherence characteristic of the range-angle joint dictionary is narrower than the main lobe of the coherence characteristic diagram of the single dictionary, the side lobe is lower, and the -3dB contour is thinner, which shows that the resolution of the range-angle joint dictionary is higher than the resolution of the dictionary formed by the single array.

[0054] Step three: based on the range-angle joint dictionary, the target range and attitude angle corresponding to different receiving arrays are obtained by using the basis pursuit method.

[0055] Taking the transmitting-receiving array as an example, the array receives the target echo from a distance C j , and an attitude angle θ i (representing the i-th rotation angle), and a sparse vector is obtained by sparse representation of the target echo in the dictionary. When solving the sparse expression α of the target echo signal x in the dictionary, the L1 norm minimization method in the convex optimization theory is used for solving, that is, solving

[0056]

[0057] wherein γ represents a weight coefficient, which changes with noise, and α is a sparse representation vector. The solution is When the alpha reaches the maximum value at the ith x j element, then the estimated value of the distance and attitude angle of the target is determined as (C j , θ i ).

[0058] Similarly, the receiving arrays R1 and R2 receive the echo signals of the target transmitted by the transceiving array, and the sparse vectors are obtained by sparse representation in the distance-angle joint dictionary, and the estimated values of the distance and attitude angle of the target are (C' j , θ' i ), (C" j , θ" i ).

[0059] Figure 4 are the distance and attitude angle results of the submarine identified by the T / R transceiving array, R1 and R2 receiving arrays according to the received echo signals of the target. As can be seen from the figure, the estimated distance and attitude angle results of the target by the three receiving arrays will have errors, and the error sizes estimated by different receiving arrays are different, mainly because the distances between the receiving arrays and the target are different, resulting in different signal-to-noise ratios.

[0060] Step four: using the best linear weighted data fusion algorithm to fully fuse the target echo information received by the three receiving arrays.

[0061] The principle of the best linear weighted data fusion is: let x i and x j be two observations of the parameter θ, and the two measurements of θ are unbiased, but have different measurement standard deviations σ i and σ j . The purpose of data fusion is to find a new estimator x ij = f(x i , y j ) such that its estimation of θ is unbiased and has a smaller estimation error than σ i and σ j . The simplest case is to consider the linear weighting of x i and x j , i.e. x ij = ax i + bx j , find a and b such that the standard deviation of x ij is minimized. Using the Lagrange multiplier method, the extreme value of the standard deviation of x ij can be obtained: Therefore, the best linear weighted data fusion is the weighting of two variables, and the size of the weighting parameter is inversely proportional to the standard deviation of the component, and the greater the error, the smaller the weighting.

[0062] The standard deviation of x ij is From the formula, it can be concluded that the error of the estimation quantity obtained by the optimal linear weighted data fusion based on the minimum mean square error criterion is not greater than the original estimation error of each component before fusion, that is, the final result obtained by the algorithm is relatively accurate.

[0063] Step five: different receiving arrays obtain the corresponding target distance and attitude angle estimation values according to step three, subtract their own relative T / R angles to obtain the target relative T / R attitude angle estimation values, and then obtain the influence degree of the target attitude angle recognition results of different receiving arrays, that is, the weight coefficient, according to the optimal linear weighted data fusion optimization algorithm in step four, and finally obtain the target attitude angle result

[0064] In order to be more convenient, the concept of signal-to-noise ratio is introduced to replace the measurement error to obtain the final measurement result, that is, The formula clearly shows the relationship between the optimal linear data fusion and the signal-to-noise ratio of each component.

[0065] Figure 5 The final result of the target attitude angle estimated by the data fusion algorithm of the different receiving arrays. As can be seen from the figure, the target information obtained by the three receiving arrays is combined by using the algorithm, so that the angle recognition result is more accurate.

[0066] Figure 6 The result of the target attitude angle recognition combined with the data fusion algorithm under different signal-to-noise ratios. The Monte Carlo experiment is performed 1000 times, the signal-to-noise ratio is-30-30dB, and the step is 5dB. As can be seen from the figure, when SNR≥-25dB, the algorithm can accurately recognize the target attitude angle, and as the signal-to-noise ratio increases, the error decreases, and the recognition accuracy correspondingly increases.

[0067] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any skilled person in the art can easily think of various equivalent modifications or replacements within the technical range disclosed by the present application, and these modifications or replacements should be covered within the protection scope of the present application.

Claims

1. A method for identifying the attitude angle of underwater targets based on multi-base sonar, characterized in that... Multistatic sonar consists of a combined transceiver array and two receiver arrays. The transceiver array transmits signals, and each receiver array receives echo signals from targets at different distances and attitude angles. The steps are as follows: Step 1: Construct a target distance-angle dictionary: For receiving target echoes using a combined transmit and receive array, the target is rotated around its geometric center, and the target attitude angle is divided into... An angle, expressed as Every angle All include For each distance, angle, and distance, there is an echo signal. Construct a dictionary of distances and angles; Assume the range-angle dictionary formed by the target echo is as follows: , it is Matrix, the first Listed as (1-1) The target distance-angle dictionary ,Right now (1-2) Simultaneously receiving array and The target echo signal was also received, and the target range-angle dictionary was constructed from the echo. and ,as follows: （1-3） （1-4） Step 2: Combine the range-angle dictionaries of different receiving arrays to form a joint range-angle dictionary. Specifically, it is expressed as: (1-5) Step 3: Based on this range-angle joint dictionary, use the basis pursuit method to obtain the target range and attitude angles corresponding to different receiver arrays; For a combined transmit and receive array, the array receives signals from a distance of [distance missing]. The attitude angle is The target echo is represented in a dictionary sparse representation to obtain a sparse vector; Solving for the target echo signal Sparse representation in the dictionary When using convex optimization theory The norm minimization method is used to solve the problem. (1-6) in, The representation weight coefficient changes with noise. Given a sparse representation vector; solve for ,when In the When the maximum value is reached at each element, the estimated value of the target's range attitude angle is determined to be ( ). Similarly, the receiving array and The target echo signals transmitted by the transmitting and receiving arrays are received respectively. A sparse vector is obtained by representing the target echo signals using a joint range-angle dictionary. The estimated values ​​of the target's range and attitude angles are then determined as follows: (), ); Step 4: Utilize the optimal linear weighted data fusion algorithm to fully fuse the target echo information received by the three receiving arrays; The optimal linear weighted data fusion method uses the following formula: in, , , , These are three sets of attitude angle estimates. , , These are the signal-to-noise ratios of the combined transmit / receive array and the two receiver arrays, respectively. Step 5: For each receiving array, obtain the estimated target range and attitude angle according to Step 3, and subtract its own relative angle from the estimated target range and attitude angle. T / R From the angle at which the target is located, we can obtain the relative position of the target. T / R The estimated attitude angle is then used to determine the influence of different receiver arrays on the target attitude angle identification result, i.e., the weighting coefficients, based on the optimal linear weighted data fusion optimization algorithm in step 4. Finally, the target attitude angle result is obtained. .

2. The method for identifying underwater target attitude angles based on multi-base sonar according to claim 1, characterized in that: The aforementioned transceiver array and two receiver arrays are arranged in a circular pattern, with an angle of 120 degrees between each array.

3. A computer system, characterized in that... include: One or more processors, a computer-readable storage medium for storing one or more programs, wherein, when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method of claim 1.

4. A computer-readable storage medium, characterized in that... The device stores computer-executable instructions, which, when executed, are used to implement the method of claim 1.

Images