Sonar beamforming method under element abnormal condition based on reliability evaluation

Abstract

This invention discloses a sonar beamforming method under array element anomaly conditions based on reliability assessment, belonging to the field of sonar array signal processing technology. The method includes: framing and frequency domain transformation of the hydrophone array signal; determining the set of neighboring array elements based on their spatial location; calculating the array element energy consistency reliability, neighboring element coherence reliability, and short-time stability reliability, and fusing them to obtain a comprehensive reliability; constructing and normalizing soft-weighted coefficients with lower-limit constraints based on the comprehensive reliability; introducing the normalized soft-weighted coefficients into broadband beamforming, performing soft-weighted beamforming at each frequency point and scanning azimuth, and averaging the results within the processing frequency band to obtain the broadband beam output; repeating the processing of all time frames to generate an enhanced azimuth history map. This invention assesses the reliability of array element channels through multi-feature fusion and uses soft weighting instead of traditional hard rejection, effectively suppressing the contamination of beamforming by abnormal array elements while retaining effective acoustic information, thus improving the clarity and stability of the azimuth history map.

Description

Technical Field

[0001] This invention belongs to the field of sonar array signal processing technology, specifically relating to a sonar beamforming method under array element anomaly conditions based on reliability assessment. Background Technology

[0002] In passive sonar array signal processing, conventional broadband beamforming typically assumes all array element channels are in normal working condition and uses all received signals from all elements for coherent superposition with equal weights. However, in real-world applications, some array element channels may be affected by localized flow noise, abnormal channel gain, short-term broadband interference, narrowband interference, or intermittent failures, leading to a degradation in the signal quality of individual elements. When abnormal element signals still participate in beamforming with fixed weights, noise and interference from the abnormal channels are introduced into the array output, causing issues such as increased background energy and more false bright fringes in the azimuth history map, thus affecting the accuracy of subsequent target detection and tracking.

[0003] The most similar technical solution to this invention is a fault detection method for underwater acoustic sensors based on root mean square cross rate. This method includes the following steps: First, the received array signal is processed in frames, and the root mean square value of each array element channel over a short period of time, and the cross-change of the root mean square value relative to the reference level or adjacent channel, are calculated. When an array element channel malfunctions or experiences significant abnormal interference, the relative relationship between its short-time root mean square value and that of a normal channel changes. Therefore, based on the difference between the statistical characteristics of each channel and the characteristics of the normal channel, it can be determined whether a particular array element channel is abnormal or faulty. Based on the fault determination result, the array element identified as faulty is muted or zeroed to prevent obviously faulty channels from further contaminating the processing results. Finally, beamforming processing is performed using the array data after the faulty channel has been removed. Some methods also incorporate time-domain statistics such as zero-crossing rate to assist in determining whether there are abnormal fluctuations or abnormal noise components in the channel signal. The core of this type of method is a hard-processing approach, that is, first dividing the array element channels into normal or faulty categories, and then muting, zeroing, or removing faulty channels.

[0004] Existing methods for handling abnormal channels based on sensor fault detection have the following technical shortcomings: First, single statistical features have limited ability to characterize complex channel anomalies, and fault diagnosis results are easily affected by environmental changes and signal fluctuations. In actual arrays, element channel anomalies do not always manifest as long-term stable complete failures, but may take various forms such as short-term broadband noise enhancement, narrowband interference superposition, channel gain fluctuations, local flow noise enhancement, or intermittent failures. If judgment is based solely on single features such as short-term energy, root mean square value, or zero-crossing rate, misjudgments and omissions are likely to occur under short-term anomaly or complex background conditions, resulting in abnormal channels not being suppressed in time, or normal channels being incorrectly identified as faulty channels.

[0005] Secondly, existing methods typically employ a hard decision-making process using binary representation of normal / faulty elements, removing those deemed faulty. While this approach can suppress the impact of obviously faulty channels, it relies on sufficiently accurate fault identification. If the identification is flawed, elements still containing valid target acoustic information are directly deleted, potentially causing main lobe broadening, sidelobe increase, decreased azimuth resolution, and reduced beam output stability. Especially in low signal-to-noise ratio and weak target scenarios, the loss of valid element information further degrades the display quality of target trajectories in the azimuth history map. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a sonar beamforming method based on reliability assessment under array element anomaly conditions.

[0007] This invention aims to alleviate the problem of increased background energy and more false peaks in the azimuth history map after beamforming when individual array element channels are affected by factors such as broadband noise, narrowband interference, abnormal gain, or intermittent failure. This is achieved through adaptive reliability assessment of array element channels and a dynamic soft weighting mechanism.

[0008] The objective of this invention is achieved through the following technical solution: A sonar beamforming method under array element anomaly conditions based on reliability assessment, comprising the following steps: S1. Perform frame-segmentation processing on the multi-channel time-domain signals acquired by the hydrophone array to obtain the frame-segmented signals of each array element in each time frame. S2. Perform frequency domain transformation on each frame of array element signal to obtain the frequency domain signal of each array element at each frequency point. S3. Based on the actual spatial location of the array elements, determine the set of neighboring array elements for each array element; S4. Calculate the array element energy consistency reliability, neighboring array element coherence reliability, and short-time stability reliability of each array element in the current time frame. S5. The array element energy consistency reliability, neighboring array element coherence reliability and short-time stability reliability are fused to obtain the comprehensive reliability of each array element in the current time frame. S6. Construct the soft weighting coefficients for each array element in the current time frame based on the comprehensive reliability, and perform lower limit constraints and normalization on the soft weighting coefficients. S7. Introduce the normalized soft weighting coefficients into the broadband beamforming process, perform soft weighting beamforming at each frequency point and scanning azimuth, and obtain the soft weighting beam output. S8. Average the beam output energy at each frequency point within the selected processing frequency band to obtain the broadband beam output of each scanning direction in the current time frame. S9. Repeat the processing of all time frames to generate an enhanced orientation history map.

[0009] Furthermore, the framing process in step S1 uses windowed framing, and the frequency domain transformation in step S2 uses Fast Fourier Transform.

[0010] Furthermore, in step S3, the method for determining the neighboring array element set is as follows: select the K array elements that are spatially closest to the current array element as its neighboring array element set.

[0011] Furthermore, the method for calculating the energy consistency reliability of the array element in step S4 is as follows: calculate the short-time energy of the current array element, calculate the median of the short-time energy of each array element in its neighboring array element set as the reference energy, and calculate the energy consistency reliability based on the deviation between the current array element energy and the reference energy.

[0012] Furthermore, the method for calculating the coherent reliability of neighboring array elements in step S4 is as follows: within a selected frequency band, calculate the frequency domain coherence coefficient between the current array element and each array element in its neighboring array element set, and average the coherence coefficients to obtain the coherent reliability of the neighboring array element.

[0013] Furthermore, the method for calculating short-time stability reliability in step S4 is as follows: based on the short-time energy of the current array element in the current frame and several historical frames, calculate the median energy of historical frames, and calculate the short-time stability reliability based on the deviation between the energy of the current frame and the median energy of historical frames.

[0014] Furthermore, in step S5, the array element energy consistency reliability, neighboring array element coherence reliability, and short-time stability reliability are weighted and fused to obtain a comprehensive reliability. in, For the reliability of array element energy consistency, For the coherent reliability of neighboring array elements, For short-term stability and reliability, .

[0015] Furthermore, the method for constructing the soft-weighted coefficients in step S6 is as follows: based on the overall reliability... Construct the first Frame number The soft weighting coefficients of each array element ; in, For overall reliability, As the lower bound of the weight, satisfying This avoids the array element signals being directly set to zero or completely eliminated.

[0016] Furthermore, the expression for soft-weighted beamforming in step S7 is: in, These are the normalized soft-weighted coefficients. Let m be the frequency domain signal of the m-th array element. The complex conjugate of the guiding phase is represented by M, where M is the total number of array elements.

[0017] The beneficial effects of this invention are as follows: This invention addresses the problems of existing abnormal channel processing methods that rely on single features and are easily affected by signal fluctuations and environmental changes. It constructs a multi-feature fusion reliability calculation method composed of element energy consistency, neighboring element coherence, and short-time stability. Since energy consistency reflects the amplitude deviation between the current element and its neighboring elements, neighboring element coherence reflects whether the element is still suitable for coherent array superposition, and short-time stability reflects the smoothness of element state changes over time, this invention reduces the impact of misjudgments or omissions based on a single statistical feature, thus improving the stability of abnormal channel evaluation compared to fault detection methods that rely solely on a single statistical feature.

[0018] This invention does not directly zero out or remove suspected anomalous array elements. Instead, it represents the element state as a continuous reliability weight and embeds this weight into the broadband beamforming process. Therefore, array elements with higher reliability maintain a larger contribution, while the contribution of array elements with lower reliability is appropriately reduced, but not completely deleted. This approach suppresses the contamination of beamforming results by anomalous array elements while preserving potentially valid acoustic information from suspected anomalous elements, avoiding the loss of target information and impact on azimuth resolution caused by hard removal.

[0019] Therefore, the present invention can improve the stability of broadband beamforming results under abnormal channel conditions, reduce background energy rise and false peaks in azimuth history diagrams, and improve the display quality of azimuth history diagrams. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art or ordinary skills, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart of the method of the present invention.

[0022] Figure 2 The figures show the experimental results of additive broadband noise anomalies, where (a) is the azimuth history diagram under normal array element conditions, (b) is the azimuth history diagram after adding anomalies, (c) is the azimuth history diagram after processing using the method of this invention, and (d) is the array element reliability diagram.

[0023] Figure 3 The figures show the results of the gain anomaly experiment, where (a) is the azimuth history diagram under normal array element conditions, (b) is the azimuth history diagram after adding anomalies, (c) is the azimuth history diagram after processing using the method of this invention, and (d) is the array element reliability diagram.

[0024] Figure 4 The figures show the results of the intermittent failure anomaly experiment, where (a) is the azimuth history diagram under normal array element conditions, (b) is the azimuth history diagram after adding anomalies, (c) is the azimuth history diagram after processing using the method of this invention, and (d) is the array element reliability diagram.

[0025] Figure 5 The figures show the experimental results of narrowband interference anomalies, where (a) is the azimuth history diagram under normal array element conditions, (b) is the azimuth history diagram after anomalies are added, (c) is the azimuth history diagram after processing using the method of this invention, and (d) is the array element reliability diagram. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0027] like Figure 1As shown, this invention first performs frame-based processing and frequency-domain transformation on the multi-channel time-domain signals acquired by the hydrophone array; determines the set of neighboring array elements for each array element based on the spatial location of the array elements; calculates the array element energy consistency reliability, neighboring array element coherence reliability, and short-time stability reliability respectively; fuses the above reliability to obtain the comprehensive reliability of each array element in the current time frame; constructs array element soft weighting coefficients based on the comprehensive reliability, and performs lower limit constraints and normalization processing on the soft weighting coefficients; finally, introduces the array element soft weighting coefficients into the broadband beamforming process to obtain the enhanced broadband beamforming result and generate an azimuth history map.

[0028] Furthermore, the passive sonar broadband beamforming enhancement method based on soft weighting of array element channel reliability provided in this embodiment of the invention includes the following specific implementation steps: S1. Perform frame segmentation processing on the multi-channel time-domain signals acquired by the hydrophone array. Assume the hydrophone array contains... The array element, the first The time-domain signal received by each array element is ,in , This represents the sampling point number. The signal of each array element is divided into frames to obtain the sampling point number. Frame number The frame signal of each array element .

[0029] S2. Perform frequency domain transformation on the array element signal of each frame. Windowing and Fast Fourier Transform are applied to the signals of each array element in the frame to obtain the frequency domain signals of each element at different frequency points. ,in Indicates the frequency point. The processing frequency band is selected according to the sonar processing requirements. Subsequent reliability calculations and broadband beamforming are all performed within this frequency band.

[0030] S3. Determine the set of neighboring array elements based on their actual spatial locations. Read the spatial coordinates of each array element and select the nearest one. Each array element is considered as a set of its neighboring array elements, denoted as . .

[0031] S4. Calculate the reliability of array element energy consistency. Frame number The short-term energy of each array element Calculate the first The set of neighboring array elements of each array element Median of short-time energy of each array element The reference energy is used as the reference energy. Based on the deviation between the current element energy and the reference energy, the energy consistency reliability is calculated: in, To prevent extremely small positive numbers with a denominator of zero.

[0032] S5. Calculate the coherence reliability of neighboring array elements. For the first... Each array element and its neighboring array elements Calculate the frequency domain coherence coefficient within the selected frequency band: in, This represents the set of frequency points within the selected processing frequency band, for the th The coherence reliability of a neighboring array element is obtained by averaging the coherence coefficients of each array element and the elements within its neighboring array element set. .

[0033] S6. Calculate short-time stability reliability. According to the... Calculate the short-time stability reliability of each array element within the current frame and several historical frames. The array element in the first The median energy of the previous historical frames is Then the short-time stability reliability is: S7. Integrate multiple reliability characteristics to obtain the comprehensive reliability of the array elements. The comprehensive reliability is obtained by weighted fusion of the array element energy consistency reliability, the coherent reliability of neighboring array elements, and the short-time stability reliability. in, .

[0034] S8. Based on overall reliability Construct the first Frame number The soft weighting coefficients of each array element : in, As the lower bound of the weight, satisfying This avoids the array element signals being directly set to zero or completely eliminated.

[0035] S9. Normalize the soft-weighted coefficients of the pair element in different time frames to obtain... .

[0036] S10, Perform soft-weighted broadband beamforming for array element channels. Frame, First Frequency points and scanning orientation Next, the normalized array element soft-weighting coefficients are introduced into the beamforming process to obtain the soft-weighted beam output: in, This represents the complex conjugate of the guiding phase.

[0037] S11, in the selected processing frequency band Within, the output energy of each frequency beam is averaged to obtain the first... Frame scan orientation Broadband beam output.

[0038] S12. Repeat the processing of all time frames to generate an enhanced orientation history map.

[0039] The effects of the present invention will be further illustrated below with reference to experiments: To verify the effectiveness of the method of this invention, experiments were conducted using publicly available data from the SWellEx-96 HLA South array. The array originally had 32 elements; 27 normal elements retained from the publicly available data were used in the experiment. The sampling rate was set to 3276 Hz, the velocity of sound to 1500 m / s, the FFT points to 2048, and the processing frequency band to 20 Hz to 1000 Hz.

[0040] Under the aforementioned unified parameters, four types of abnormal array element scenarios were constructed. The abnormal array elements were designated as elements 5, 14, and 23. The first type was additive broadband noise anomaly, where broadband random noise was added to the abnormal array element between the 25th and 30th minute, with the noise amplitude being twice the root mean square value of the corresponding array element. The second type was gain anomaly, where the abnormal array element signal was multiplied by a gain factor of 3 between the 25th and 30th minute. The third type was intermittent failure anomaly, where the abnormal array element signal was reset to zero between the 27th and 28th minute. The fourth type was narrowband interference anomaly, where 300 Hz narrowband interference was added to the abnormal array element between the 25th and 30th minute, with the interference amplitude being twice the root mean square value of the corresponding array element.

[0041] For each type of abnormal data, the results of conventional broadband beamforming, array element reliability, and reliability soft-weighted beamforming are calculated and compared with the results of conventional beamforming under all normal array element conditions.

[0042] The experimental results are as follows: Appendix Figure 2 The results of the additive broadband noise anomaly experiment are shown, including the attached... Figure 2 (a) is the azimuth history diagram under all normal array element conditions; (attached) Figure 2 (b) shows the azimuth history of array elements 5, 14, and 23 after adding broadband noise. It can be seen that the background energy significantly increased during the abnormal time period, and a strong horizontal interference band appeared, locally obscuring the target trajectory; (See attached image) Figure 2 (c) This is the azimuth history map processed using the method of this invention. It can be seen that the abnormal background is suppressed and the main trajectory of the target is recovered; Appendix Figure 2 (d) is the array element reliability diagram. It can be seen that the reliability of abnormal array elements is significantly reduced during abnormal time periods, indicating that the present invention can effectively identify array elements contaminated by broadband noise.

[0043] Appendix Figure 3 Results of the gain anomaly experiment. (Attached) Figure 3 (a) is a diagram showing the orientation history of all normal array elements; attached. Figure 3 (b) shows the azimuth history diagrams of array elements 5, 14, and 23 after gain amplification, revealing local energy enhancement; (See attached diagram) Figure 3 (c) This is the result after processing according to the present invention. The main trajectory of the target remains continuous, and the local impact caused by abnormal gain is reduced. Figure 3 (d) is the array element reliability diagram. It can be seen that the reliability of abnormal array elements decreases within the set abnormal time period, which is consistent with the location and time of the injected abnormality.

[0044] Appendix Figure 4 These are the results of an intermittent failure anomaly experiment. (Attached) Figure 4 (a) is a diagram showing the orientation history of all normal array elements; attached. Figure 4 (b) shows the azimuth history of elements 5, 14, and 23 after they failed between 27 and 28 minutes. Due to the small number of failed elements and the short duration, the overall change in the conventional azimuth history is not very significant; see attached. Figure 4 (c) is the orientation history map after processing according to the present invention, where the target trajectory remains continuous; Appendix Figure 4 (d) is the array element reliability diagram. It can be seen that the reliability of the corresponding abnormal array element decreases significantly during the failure period.

[0045] Appendix Figure 5 These are the experimental results for narrowband interference anomalies. (Attached) Figure 5 (a) is a diagram showing the orientation history of all normal array elements; attached. Figure 5 (b) is the azimuth history diagram of array elements 5, 14, and 23 after adding 300 Hz narrowband interference. Local background elevation and interference fringes can be observed during the anomalous time period. (See attached diagram.) Figure 5 (c) shows the result after processing according to the present invention, where the background interference is suppressed to a certain extent; Appendix Figure 5 (d) is the array element reliability diagram. It can be seen that the reliability of abnormal array elements decreases during the interference period, indicating that the present invention also has the ability to detect and suppress narrowband interference anomalies.

[0046] In summary, the four sets of experiments demonstrate that the present invention has the ability to detect anomalies in various array elements and can effectively enhance the performance of azimuth history maps.

[0047] In this embodiment of the invention, a weighted summation method is used to fuse different reliability features to obtain the comprehensive reliability of the array elements. In other embodiments, product fusion, weighted geometric average, threshold segmentation fusion, or adaptive weight fusion can also be used to obtain the comprehensive reliability. As long as the final output is a continuous weight or soft decision result used to represent the reliability of the array element channels, it is an equivalent alternative of this invention.

[0048] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A sonar beamforming method under an element abnormal condition based on reliability evaluation, characterized in that, Includes the following steps: S1. Perform frame-segmentation processing on the multi-channel time-domain signals acquired by the hydrophone array to obtain the frame-segmented signals of each array element in each time frame. S2. Perform frequency domain transformation on each frame of array element signal to obtain the frequency domain signal of each array element at each frequency point. S3. Based on the actual spatial location of the array elements, determine the set of neighboring array elements for each array element; S4. Calculate the array element energy consistency reliability, neighboring array element coherence reliability, and short-time stability reliability of each array element in the current time frame. S5. The array element energy consistency reliability, neighboring array element coherence reliability and short-time stability reliability are fused to obtain the comprehensive reliability of each array element in the current time frame. S6. Construct the soft weighting coefficients for each array element in the current time frame based on the comprehensive reliability, and perform lower limit constraints and normalization on the soft weighting coefficients. S7. Introduce the normalized soft weighting coefficients into the broadband beamforming process, perform soft weighting beamforming at each frequency point and scanning azimuth, and obtain the soft weighting beam output. S8. Average the beam output energy at each frequency point within the selected processing frequency band to obtain the broadband beam output of each scanning direction in the current time frame. S9. Repeat the processing of all time frames to generate an enhanced orientation history map.

2. The sonar beamforming method in the presence of an element anomaly condition based on reliability evaluation according to claim 1, characterized in that: In step S1, the frame segmentation process uses windowed frame segmentation, and in step S2, the frequency domain transformation uses Fast Fourier Transform.

3. The method of claim 1, wherein: The method for determining the neighboring array element set in step S3 is as follows: select the K array elements that are closest to the current array element in space as its neighboring array element set.

4. The method of claim 1, wherein: The method for calculating the energy consistency reliability of array elements in step S4 is as follows: calculate the short-time energy of the current array element, calculate the median of the short-time energy of each array element in its neighboring array element set as the reference energy, and calculate the energy consistency reliability based on the deviation between the current array element energy and the reference energy.

5. The method of claim 1, wherein: The method for calculating the coherent reliability of neighboring array elements in step S4 is as follows: within a selected frequency band, calculate the frequency domain coherence coefficient between the current array element and each array element in its neighboring array element set, and average the coherence coefficients to obtain the coherent reliability of the neighboring array element.

6. The method of claim 1, wherein: The method for calculating short-time stability reliability in step S4 is as follows: based on the short-time energy of the current array element in the current frame and several historical frames, calculate the median energy of historical frames, and calculate the short-time stability reliability based on the deviation between the energy of the current frame and the median energy of historical frames.

7. The method of claim 1, wherein: In step S5, the energy consistency reliability of array elements, the coherence reliability of neighboring array elements, and the short-time stability reliability are weighted and fused to obtain the comprehensive reliability. wherein, is the array element energy consistency reliability, is the adjacent element coherence reliability, is the short-term stability reliability, .

8. The sonar beamforming method under array element anomaly conditions based on reliability assessment according to claim 1, characterized in that: The method for constructing the soft-weighted coefficients in step S6 is as follows: based on the comprehensive reliability... Construct the first Frame number The soft weighting coefficients of each array element ; in, For overall reliability, As the lower bound of the weight, satisfying This avoids the array element signals being directly set to zero or completely eliminated.

9. The sonar beamforming method under array element anomaly conditions based on reliability assessment according to claim 1, characterized in that: The expression for soft-weighted beamforming in step S7 is: in, These are the normalized soft-weighted coefficients. Let m be the frequency domain signal of the m-th array element. The complex conjugate of the guiding phase is represented by M, where M is the total number of array elements.

Images