Image sonar arc array wave beam digital correction method

Abstract

The invention relates to the technical field of image sonar signal processing, in particular to an image sonar arc array wave beam digital correction method. The image sonar signal arc array beam directivity correction method is achieved through a digital method, numerical compensation is carried out on IQ orthogonal signals of all elements of an arc array, beam directivity errors caused by sound velocity changes during beam forming operation are reduced, the method has certain universality, and the method is suitable for popularization and application. The method is especially suitable for the field of small high-frequency image sonar signal processing, and can effectively relieve the difficulty of correcting the orientation precision in the later period. Meanwhile, the correction method adopted by the invention is simple and effective, can be realized by using hardware with limited resources, and is high in portability; according to the design method, the influence of the sound velocity change on the arc array beam directivity is effectively reduced, so that the image sonar orientation error caused by the sound velocity change is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of image sonar signal processing, and in particular to a method for digitally correcting an image sonar arc array beam. Background Art

[0002] Imaging sonar typically uses beamforming to detect the direction and range of a target. Beamforming involves phase-compensating the output signals of each array element so that when a target echo is incident from a certain direction, the output signals from each element are superimposed in phase, maximizing the array output. This process can determine the direction of the signal source while suppressing noise from directions different from the target echo.

[0003] Image sonar signal preprocessing typically involves A / D conversion, quadrature demodulation, filtering, and decimation to generate orthogonal I / Q signals. Weighted superposition of these I / Q signals generates beamforming values. Beamforming methods can be categorized into time-domain and frequency-domain methods.

[0004] Without loss of generality, assume that the receiving sound array is a uniform semicircular array composed of N identical array elements, and the phase difference between each array element is By making corresponding time delay compensation for the signals output by various array elements, the signals of each array element can be in phase in the beam direction, so that the superimposed output signal in the target direction is maximized to achieve the purpose of directionality. At the same time, the superimposed output signal in other directions becomes smaller, which plays the role of spatial filtering. The arc array beamforming diagram is shown below. Figure 1 As shown, N is an even number.

[0005] For the Kth array element, the time delay τ relative to the center of the circle K or phase delay The calculation formula is:

[0006]

[0007] where θ i is the direction angle, R is the radius of the receiving arc array, λ is the wavelength, and v is the wave speed.

[0008] When forming a beam, side lobes are inevitably formed, which will affect the image sonar quality. The inherent characteristics of the arc array make the side lobe level higher, which brings more serious interference to the image quality. Usually, the amplitude weighting of the output signal of the receiving array element is used to reduce the side lobe level. Compared with uniform weighting, non-uniform weighting can more effectively suppress the side lobe level. Cosine square weighting is used to suppress side lobes. The weighting value W K The calculation formula is as follows.

[0009] W K =cos 2 [θ i-(k-1)·π / (N-1)],

[0010] By outputting the value x for each array element K The beam value y(t) can be obtained by weighted summation. The calculation formula is as follows:

[0011]

[0012] It can be seen from the above calculation formula that the beam steering accuracy is affected by the speed of sound v.

[0013] Common imaging sonar systems are relatively small and typically lack a sound velocity meter. Since the carrier carrying the imaging sonar cannot obtain real-time sound velocity information in the current operating environment, the imaging sonar's beam direction accuracy is inevitably affected by variations in sound velocity. Therefore, a solution is urgently needed to mitigate the impact of varying sound velocity on the directivity of the array beam, thereby reducing the imaging sonar's directionality errors caused by these variations. Summary of the Invention

[0014] The problem to be solved by the present invention is to reduce the influence of the sound velocity change on the beam orientation accuracy, thereby reducing the image sonar orientation error caused by the sound velocity change.

[0015] In view of the shortcomings of the prior art, the present invention solves the technical problems thereof by adopting a technical solution: a method for digitally correcting an arc array beam of an imaging sonar. According to the index conditions and performance parameters of the imaging sonar, a portable and easy-to-implement FPGA-based arc array beamforming implementation method is constructed, and the method is digitally corrected. The method comprises the following steps:

[0016] Step 1: Based on the imaging sonar's system indicators, the dynamic range of sound velocity, and the arc array parameters, the correction coefficient and weight of each array element are obtained through simulation calculation. The system indicators include the observation range and the number of beams.

[0017] Step 2: Writing the correction coefficients and weights into the FPGA internal memory BRAM in a pre-stored or dynamically configured manner;

[0018] Step 3: In the real-time arc array beamforming operation, the correction coefficients and weights are read from the memory BRAM, and weighted superposition operation is performed on the digitized IQ orthogonal signals of the array elements to generate multi-beam data in azimuth.

[0019] Preferably, the method for determining the correction coefficient and weight includes: dynamically adjusting the weight value through simulation calculation according to the difference in array elements on the left and right sides of the target beam azimuth, so that the number of array elements participating in the calculation on the left and right sides is approximately equal.

[0020] Preferably, for a beam in which the number of left array elements participating in the calculation is less than the number of right array elements, the weight values of the left array elements are increased; otherwise, the weight values of the right array elements are increased.

[0021] Preferably, the simulation calculation of the correction coefficient and weight further includes: within the range of dynamic change of sound speed, fine-tuning the theoretical weighting coefficient to approach the optimal weighting coefficient, thereby minimizing the beam directivity error.

[0022] Preferably, the memory BRAM is a dual-port memory built into the FPGA, supports a dynamic configuration mode, and the correction coefficients and weights are updated in real time through the FPGA interface and internal logic.

[0023] Preferably, the FPGA platform further includes a logic control module for scheduling the reading timing of the correction coefficients and weights, and synchronizing them with the beamforming operation.

[0024] Preferably, the method is applicable to an imaging sonar whose receiving array is an arc array, and the array element correction coefficient and weight capacity are determined by the beam number index of the imaging sonar.

[0025] Preferably, the beam directivity error after weighted superposition in the method is less than 0.5°, and the error reduction range is not less than 30% when the sound speed dynamic change range is ±5%.

[0026] The beneficial effects of the present invention are as follows:

[0027] To address the impact of sound velocity variations on the beam orientation of imaging sonar arrays, this application provides an improved method for reducing the impact of sound velocity variations on beam orientation accuracy. This application employs time-domain multi-beamforming and combines it with pulse ranging to achieve positioning and ranging for imaging sonar. First, the observation range, number of beams, array element parameters of the receiving array, and the dynamic range of sound velocity variations in the imaging sonar index system are used as input variables, and the theoretical weighting coefficients of each array element are obtained through simulation calculation. It is particularly noteworthy that there may be slight differences between the theoretical weighting coefficients and the optimal weighting coefficients, and simulation fine-tuning can be performed under stringent requirements; secondly, the generated array element correction coefficients and weights are written into the memory BRAM inside the sonar in a pre-stored or dynamically configured manner. When performing real-time beamforming operations, the array element correction coefficients and weights are read out from the memory BRAM in real time under logical control, and combined with the array element digitized IQ orthogonal signals to participate in the arc array beamforming operation; finally, compared with other methods, the method of the present application uses array element weighting to reduce the impact of sound speed changes on the arc array beam directional accuracy, effectively reducing the complexity of the correction method. In this method, there is no need to use the IP core provided by the manufacturer in engineering practice, and it has certain independent intellectual property rights. At the same time, this method requires relatively little additional hardware resources and can be widely used in the field of image sonars where the receiving array is an arc array, and has certain practical and scientific research value. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a schematic diagram of beamforming involved in the background technology (N is an even number);

[0029] Figure 2 It is a schematic flow diagram of the method of the present invention;

[0030] Figure 3 This is a schematic diagram of the arc array arrangement involved in the present invention. DETAILED DESCRIPTION

[0031] The present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments. The embodiments of the present invention are provided for purposes of illustration and description and are not intended to be exhaustive or to limit the invention to the disclosed forms. Many modifications and variations will be apparent to those skilled in the art. The embodiments are chosen and described to better illustrate the principles of the invention and its practical application, and to enable those skilled in the art to understand the invention and design various embodiments with various modifications suitable for specific applications.

[0032] Imaging sonar uses a variety of environments, and the speed of sound in water exhibits complex variations depending on factors such as depth, temperature, and salinity. Temperature is a significant factor influencing the speed of sound in water. For example, when the temperature approaches 15°C, the speed of sound increases by 16 m / s for every 6°C increase, an increase of approximately 1%. If the imaging sonar's design range is 100 m, the dynamic range of sound speed can reach up to 50 m / s. The directional error of the edge beam of the observation range formed using the above algorithm can reach up to 0.1°, an unacceptable error in many imaging sonar applications.

[0033] For imaging sonars with an arc array receiving array, when the sound speed varies within a dynamic range of 50 m / s, the error in beam pointing is minimal near 90° of the receiving array normal, and increases toward the sides. The impact is greatest at the edges of the imaging sonar's observation sector. This is due to the different numbers of array elements used in different beams during the arc array beamforming operation. Figure 3 The figure shows the layout of the arc array, where N is the number of receiving array elements and the target S is located at the left edge of the sonar observation range.

[0034] The beam numbers are usually marked from the left to the right of the observation range, 1 # The beam is at the left edge of the observation range. Considering the opening angle of the receiving array element and the amount of beamforming calculation, some array elements are used to participate in the calculation during the arc array beamforming operation. # When forming a beam, array elements 1 to M are used to participate in the beamforming operation, so that 1 # The number of left and right array elements participating in the beam azimuth calculation is unequal, causing the beam azimuth to be non-perpendicular to the equivalent chords of elements 1 through M. For the middle beam, which is close to the normal to the arc array, the number of elements participating in the beam calculation on the left and right sides is essentially the same, minimizing the impact.

[0035] Based on the above analysis, the correction coefficients and weights of the elements with fewer numbers on both sides of each beam are increased to artificially make the number of elements on both sides approximately equal. The weighting coefficient of each element is calculated according to the dynamic range of sound speed, so as to effectively reduce the impact of sound speed changes on beam directivity.

[0036] As shown in the figure, the present invention provides a digital correction method for arc array beamforming of an imaging sonar. The method is applicable to an imaging sonar receiving array that is an arc array. The capacity of the array element correction coefficient and weight is determined by the number of beams of the imaging sonar. According to the index conditions and performance parameters of the imaging sonar, a portable and easy-to-implement FPGA-based arc array beamforming implementation method is constructed and digitally corrected. The FPGA platform also includes a logic control module for scheduling the reading timing of the correction coefficient and weight, and synchronizing it with the beamforming operation. The method includes the following steps:

[0037] Step 1: Based on the imaging sonar's system specifications, the dynamic range of sound velocity, and the arc array parameters, simulation calculations are used to determine the correction coefficients and weights for each array element. These system specifications include observation range and number of beams. The correction coefficients and weights are determined by dynamically adjusting the weights based on the difference in array elements on the left and right sides of the target beam's azimuth, ensuring that the number of array elements participating in the calculation on the left and right sides is approximately equal. For beams where the number of elements participating on the left side is less than the number on the right, the weights of the elements on the left are increased; conversely, the weights of the elements on the right are increased. Within the dynamic range of sound velocity, the theoretical weighting coefficients are fine-tuned to approach the optimal weighting coefficients, thereby minimizing beam directivity errors.

[0038] Step 2: Write the correction coefficients and weights into the FPGA internal memory BRAM in a pre-stored or dynamically configured manner; the memory BRAM is a dual-port memory built into the FPGA, supports dynamic configuration mode, and the correction coefficients and weights are updated in real time through the FPGA interface and internal logic.

[0039] Step 3: In the real-time arc array beamforming operation, the correction coefficients and weights are read from the memory BRAM, and weighted superposition operation is performed on the digitized IQ orthogonal signals of the array elements to generate multi-beam data in azimuth.

[0040] During engineering implementation, the optimal weighting coefficient for each array element, namely the array element correction coefficient and weight, can be calculated and simulated based on the number of beams formed by the imaging sonar, the arc array formation parameters, and the dynamic range of sound speed. These can be pre-stored or dynamically configured in real time and written to the sonar's internal memory (BRAM). When participating in arc array beamforming operations, the array element correction coefficient and weight are read out in real time according to the control logic and combined with the array element's digitized IQ orthogonal signals in the BRAM memory to achieve real-time calculation and output of multi-beam data.

[0041] The array element digital IQ orthogonal signal refers to the IQ orthogonal signal obtained by converting the echo signal received by the imaging sonar receiving arc array into a digitized echo signal data stream, and further performing orthogonal demodulation, filtering, and extraction on the digitized echo signal. The IQ orthogonal signal is used to participate in the beamforming operation.

[0042] The array element correction coefficient and weight are mainly determined by the array element difference between the left and right sides of a beam orientation when forming a certain beam, and the weight value is determined according to the array element difference.

[0043] Sometimes there are slight differences between the theoretical weight values and the optimal weight values, requiring fine-tuning through simulation. Under the same conditions, for each beam, the weight value of each array element must be fine-tuned through simulation calculation.

[0044] The weighting method is universal in the field of image sonar signal processing where the receiving array is an arc array.

[0045] This application discloses a digital correction method for an imaging sonar array beam. First, the observation range, number of beams, and element parameters of the receiving array, as well as the dynamic range of sound velocity variations, in the imaging sonar index system are used as input variables. The theoretical weighting coefficients for each element are then obtained through simulation calculation. It is particularly important to note that there may be slight differences between the theoretical weighting coefficients and the optimal weighting coefficients, and simulation fine-tuning can be performed under stringent requirements. Secondly, the generated weighting coefficients are written to the sonar's internal memory in a pre-stored or real-time configurable manner. During real-time beamforming operations, the element correction coefficients and weights are read from the memory BRAM in real time under logical control to participate in the array beamforming operations. Finally, compared to other methods, the method of this application uses element weighting to reduce the impact of sound velocity variations on the array beam directional accuracy, effectively reducing the complexity of the correction method.

[0046] Finally, the method described in this paper does not require the use of vendor-provided IP cores in engineering practice, and possesses independent intellectual property rights. Furthermore, this method requires minimal additional hardware resources and can be widely applied to imaging sonars using arc arrays, demonstrating both practical and scientific value.

[0047] This application provides a method for digital correction of arc array beams for imaging sonar, preferably implemented in software within an FPGA. The weighting coefficients of each array element are obtained by simulation calculation based on parameters such as the observation range, number of beams, dynamic range of sound velocity, and arc array formation parameters. It should be noted that there are sometimes slight differences between the theoretical weighting coefficients and the optimal weighting coefficients. The obtained weighting coefficients can be written into the FPGA's built-in memory BRAM in a pre-stored manner or written into the FPGA's built-in dual-port memory BRAM in a dynamic configuration manner.

[0048] When performing arc array beamforming operations, the FPGA reads out the correction coefficients in the dual-port memory BRAM in real time under the control of logic to participate in the operation, thereby obtaining multi-beam values in real time.

[0049] To facilitate understanding, a specific example implementation will be used to analyze the basic principles and implementation key points of each link. The basic parameters of the imaging sonar used in this implementation example are: a central operating frequency of 450kHz, a horizontal observation range of 90°, 180 channels, 512 beams, a horizontal beam angle of 1°, an operating range of 100m, and a dynamic range of sound speed of 50m / s.

[0050] According to the example parameters, it is necessary to generate 512 beams with a horizontal angle less than 1° within the observation range of 45° to 135°. Through simulation calculations, in order to achieve a horizontal beam width of 1°, at least 84 array elements are required to generate each beam. At the same time, in order to make full use of the receiving array elements and take into account the array element opening angle and calculation amount, 117 array elements are used to form the beam, that is, 1 to 117 array elements generate 1 # ~8 # Beam, 2 to 118 elements generate 9 # ~16 # Beam, and so on, the last 64 to 180 array elements generate 497 # ~512 # Ideally, the above method can meet the requirement that the horizontal beam angle is less than 1°.

[0051] When the sound speed varies within a dynamic range of 50 m / s, the beam pointing error is minimal (0.002°) near 90°, the normal to the arc array. The error increases toward the sides of the observation range, reaching its maximum impact at the edges, where the error is greatest (0.1°) at 45° and 135°. This is due to the dynamic changes in sound speed and the different numbers of array elements used on either side of the beam during the arc array beamforming operation.

[0052] For example, in generating 1 # When forming a beam, 1 to 117 array elements are selected to participate in the arc array beamforming operation. # The azimuth of the beam is 45° from the left edge of the observation sector, and is between elements 45 and 46 of the arc array. The left elements used in the beamforming operation are 1 to 45, while the right elements are 46 to 117. The former has a total of 45 elements, and the latter has a total of 72 elements. The right side has 28 more elements than the left side. The pre-formed beam direction is not perpendicular to the equivalent chord of elements 1 to 117. 32 to 148 elements are used to generate the middle 256 # When beamforming, the desired beam pointing is 89.912° (between 90 and 91 primitives), with 59 and 58 elements on both sides, respectively. The beam direction is essentially perpendicular to the equivalent chord of 32 to 148 elements.

[0053] Based on the above analysis, by increasing the weight of the element correction coefficient corresponding to the side with fewer elements when forming a beam, the number of elements on both sides is artificially made approximately equal, thereby reducing the error in beam pointing. # When beamforming, multiplying the weighting coefficient of one array element by 28 reduces the beam pointing error from 0.1° to 0.006° when the sound speed varies within a dynamic range of 50 m / s. Theoretically, the weighting value is proportional to the difference in the number of array elements forming the beam.

[0054] In engineering practice, there may be slight differences between the theoretical value and the optimal value. # When beamforming, the expected value is 46.23°. The number of array elements on both sides is 46 and 71, respectively. Theoretically, the weighting coefficient of one array element should be multiplied by 25. However, simulation found that the beam directivity error is 0.010° when the sound speed varies within a dynamic range of 50 m / s. Adjusting the weighting coefficient of one array element to multiply by 23 reduces the beam directivity error to 0.006° under the same conditions. Therefore, for all beam values, the weight value of each array element needs to be fine-tuned through simulation to achieve the optimal effect.

[0055] The correction algorithm is used to first calculate the 117 weighting coefficients for each beam value according to the above method, and then modify the corresponding array element weight value. # ~256 # Beam, multiply each weighting coefficient by the corresponding weight value, for 257 # ~512 # Beam, multiply the last weighting coefficient by the corresponding weight value.

[0056] Once the weighting coefficients and weight values are calculated, they can be stored in a dynamically configured dual-port BRAM format within the pre-stored FPGA internal memory. During the arc-array beamforming operation, the memory, under the logic's scheduling, participates in the arc-array beamforming operation, outputting 512 beam values in real time.

[0057] In summary, the present application provides a method for digital correction of image sonar array beams, the processing flow of which is as follows: Figure 2 As shown, the array element digital IQ orthogonal signals, array element correction coefficients and weights in the process are all stored in the FPGA internal memory BRAM. The use of pre-storage or dynamic configuration effectively improves the flexibility and versatility of the method in different image sonars. When performing real-time arc array beamforming operations, the array element correction coefficients and weights are read out from the memory BRAM in real time under logical control and participate in the arc array beamforming operation. This method reduces the image sonar directional error caused by changes in sound speed. Finally, compared with other methods, the method of this application has the advantages of strong real-time performance, small computational complexity, and low resource consumption. It can effectively reduce the workload of software transplantation for similar image sonar products and has a wider applicability.

[0058] The calculation method of the array element correction coefficient has certain applicability, and its function is to reduce the directional error of the imaging sonar caused by the change of sound speed in different application environments.

[0059] The array element correction method is applicable to an imaging sonar whose receiving array is an arc array, and is particularly applicable to a compact imaging sonar in a scenario where operations such as signal preprocessing and beamforming are performed at the wet end.

[0060] The array element correction coefficient and weight calculation method uses an FPGA hardware platform during engineering implementation. The FPGA hardware platform has the advantages of strong real-time performance, small calculation amount, and low resource consumption.

[0061] The array element correction coefficient and weight calculation method described above can be pre-calculated and stored in a BRAM memory, or can be calculated and configured in real time to increase software portability. The parameters in the array element correction coefficient and weight calculation method should be determined based on the imaging sonar's observation range, the dynamic range of sound velocity, and the arc array parameters. The capacity of the correction coefficient is determined by the imaging sonar's beam count indicator. This method can effectively reduce the impact of sound velocity variations on the directional accuracy of imaging sonars with arc array receiving arrays, and the implementation method is simple. The beam directivity error after weighted superposition is less than 0.5°, and the error reduction is no less than 30% when the sound velocity dynamic range is ±5%.

Claims

1. A digital correction method for imaging sonar array beamforming, characterized by: According to the index conditions and performance parameters of the imaging sonar, a portable and easy-to-implement FPGA-based arc array beamforming implementation method is constructed and digitally corrected. The method includes the following steps: Step 1: Based on the imaging sonar's system indicators, the dynamic range of sound velocity, and the arc array parameters, the correction coefficient and weight of each array element are obtained through simulation calculation. The system indicators include the observation range and the number of beams. Step 2: Writing the correction coefficients and weights into the FPGA internal memory BRAM in a pre-stored or dynamically configured manner; Step 3: In the real-time arc array beamforming operation, the correction coefficients and weights are read from the memory BRAM, and weighted superposition operation is performed on the digitized IQ orthogonal signals of the array elements to generate multi-beam data in azimuth.

2. The imaging sonar arc array beamforming digital correction method according to claim 1, characterized in that: The method for determining the correction coefficient and weight includes: dynamically adjusting the weight value through simulation calculation according to the difference in array elements on the left and right sides of the target beam azimuth, so that the number of array elements participating in the calculation on the left and right sides is approximately equal.

3. The digital correction method for arc array beamforming of imaging sonar according to claim 2, characterized in that: For beams where the number of left array elements participating in the calculation is less than that of right array elements, the weight values of the left array elements are increased; otherwise, the weight values of the right array elements are increased.

4. The digital correction method for arc array beamforming of imaging sonar according to claim 2, characterized in that: The simulation calculation of the correction coefficient and weight further includes: within the range of dynamic change of sound speed, fine-tuning the theoretical weighting coefficient to approach the optimal weighting coefficient, thereby minimizing the beam directivity error.

5. The digital correction method for imaging sonar arc array beamforming according to claim 1, characterized in that: The memory BRAM is a dual-port memory built into the FPGA, supports a dynamic configuration mode, and the correction coefficients and weights are updated in real time through the FPGA interface and internal logic.

6. The imaging sonar arc array beamforming digital correction method according to claim 1, characterized in that: The FPGA platform also includes a logic control module for scheduling the reading timing of the correction coefficients and weights and synchronizing them with the beamforming operation.

7. The digital correction method for arc array beamforming of imaging sonar according to claim 1, characterized in that: The method is applicable to an image sonar receiving array that is an arc array, and the array element correction coefficient and weight capacity are determined by the beam number index of the image sonar.

8. The digital correction method for arc array beamforming of imaging sonar according to claim 1, characterized in that: The beam directivity error after weighted superposition by the method is less than 0.5°, and the error reduction range is not less than 30% when the sound speed dynamic change range is ±5%.

Images