An underwater acoustic target recognition method, device and computer-readable storage medium

Abstract

The present invention belongs to the technical field of target recognition, and relates to an underwater acoustic target recognition method, device and computer-readable storage medium. Multiple underwater acoustic signal samples of the same category are obtained from datasets corresponding to different sea areas; the prediction probabilities of each underwater acoustic signal sample belonging to each category are obtained; the prediction probabilities of each underwater acoustic signal sample belonging to each category are parameterized, and the uncertainty of the prediction probability confidence distribution of each underwater acoustic signal sample is output; the prediction probabilities of multiple underwater acoustic signal samples belonging to each category and the uncertainty of the prediction probability confidence distribution are fused to obtain a comprehensive basic probability assignment and a comprehensive probability distribution uncertainty, the prediction categories of multiple underwater acoustic signal samples are output, and the value of the underwater acoustic target recognition loss function is calculated; the convolutional neural network module and the basic probability assignment module are trained to obtain an underwater acoustic target recognition model. The underwater acoustic target recognition model of this solution has a high recognition accuracy for underwater acoustic signals in different sea areas.

Description

Technical Field

[0001] The present invention relates to the technical field of target recognition, and in particular to an underwater acoustic target recognition method, device and computer-readable storage medium. Background Art

[0002] Underwater acoustic target recognition refers to analyzing and processing underwater sound signals to identify whether there are target objects or phenomena in the water, which plays an important role in fields such as marine scientific research and marine resource development. Traditional and mature intelligent underwater acoustic target recognition technologies mainly focus on feature selection, feature extraction, and classifier design. Feature values that can represent the essence of the target are obtained through feature extraction, and then classification and recognition are carried out in combination with a classifier.

[0003] In the prior art, various intelligent underwater acoustic target recognition methods have been proposed. For example, the patent with the publication number CN118606804A proposes an underwater acoustic target recognition method based on the fuzzy support vector center. By collecting the ship radiated noise signals in the sea area to be recognized and extracting the statistical feature of the acoustic texture slices in different dimensions, the features are input into the fuzzy support vector machine for training. The trained fuzzy support vector is used to output the fuzzy support vector, and the distances from each statistical feature to the center of the fuzzy support vector are calculated to determine the membership assignment function. The new training set is generated by combining the membership assignment function with the statistical features in different dimensions, and the fuzzy support vector machine is trained again to obtain the optimal fuzzy support vector machine, so as to classify the new ship radiated noise signals and identify whether there are targets in the sea area to be recognized. The patent with the publication number CN118131200A discloses an underwater acoustic target recognition method based on metric learning. By collecting the sonar signals in the sea area to be recognized, calculating the time-frequency spectrum features of the sonar signals, and then generating an underwater acoustic signal data set, a depth residual network optimized by attention is constructed to achieve efficient signal amplitude and phase feature mining. Then, a loss function based on metric learning is used to mine the key recognition features in the underwater acoustic signals, so as to achieve high-accuracy and high-generalization target recognition for the sea area to be recognized. The patent with the publication number CN106250848A discloses a one-class underwater acoustic target recognition method based on multi-model fusion. The features of the underwater acoustic target are extracted by using a deep autoencoder neural network, and then the Adaboost algorithm is used for one-class classifier ensemble learning to obtain a one-class underwater acoustic target recognition model, so that the target features in the underwater acoustic signals can be directly extracted to judge whether there are targets in the sea area to be recognized.

[0004] Due to the large differences in the underwater acoustic propagation environment in different sea areas, for example, the water layer in shallow sea areas is relatively thin and the seabed topography is complex, underwater acoustic signals will be affected by reflection, scattering and absorption during the propagation process. In deep sea areas, the water layer is thick, the propagation distance is long, and the sound speed changes significantly with depth. Underwater acoustic signals are prone to distortion during the propagation process. Therefore, underwater acoustic signals in different sea areas are very different in characteristics such as frequency, amplitude, and phase. Existing underwater acoustic target recognition methods all use historical data of the sea area to be recognized to train the underwater acoustic target recognition model. Since the trained model can only specifically capture the characteristics of the underwater acoustic signals in the sea area to be recognized and cannot adapt to the unique characteristics of data in other sea areas, the accuracy of the recognition results of the model for underwater acoustic signals in other sea areas is relatively low.

[0005] Currently, there are two ways to solve this problem. One is to re-collect the underwater acoustic signal data of the sea area before target recognition in different sea areas and conduct targeted training on the model. However, this method will increase the computational resources and recognition time costs. The other is to directly fuse the data sets of different sea areas and input the samples in the fused data set into the model for training in sequence. However, the difference in the feature distribution of data in different sea areas will cause the update direction of the model parameters to change frequently during the training process, which not only increases the computational resources and time consumption, but also the model cannot fully learn the characteristics of various data, and even may occur the situation where the model cannot converge, that is, the model cannot effectively identify whether there is a target in the underwater acoustic signal.

[0006] To sum up, the existing underwater acoustic target recognition methods ignore the characteristic differences of underwater acoustic signals in different sea areas, resulting in the problem that it is impossible to balance the recognition efficiency and recognition accuracy when performing target recognition in different sea areas. Summary of the Invention

[0007] Therefore, the technical problem to be solved by the present invention is to overcome the problem that the existing underwater acoustic target recognition methods ignore the characteristic differences of underwater acoustic signals in different sea areas, resulting in the problem that it is impossible to balance the recognition efficiency and recognition accuracy when performing target recognition in different sea areas.

[0008] To solve the above technical problem, the present invention provides an underwater acoustic target recognition method, including:

[0009] Obtain multiple underwater acoustic signal samples with the same category from the data sets of different sea areas;

[0010] Input each underwater acoustic signal sample into the convolutional neural network module and output the prediction probabilities of each underwater acoustic signal sample belonging to each category;

[0011] Input the prediction probabilities of each underwater acoustic signal sample belonging to each category into the basic probability assignment module, perform parameterization operation on the prediction probabilities of each underwater acoustic signal sample belonging to each category by using the Dirichlet distribution, and output the uncertainty of the prediction probability confidence distribution of each underwater acoustic signal sample;

[0012] The prediction probabilities of multiple underwater acoustic signal samples belonging to various categories and the uncertainty of the prediction probability confidence distribution are respectively fused using the Dempster-Shafer evidence synthesis rule to obtain the comprehensive basic probability assignment and the comprehensive probability distribution uncertainty of multiple underwater acoustic signal samples belonging to various categories, thereby outputting the prediction categories of multiple underwater acoustic signal samples, and calculating the value of the underwater acoustic target recognition loss function;

[0013] Multiple underwater acoustic signal samples with the same category are newly obtained from datasets in different sea areas, and the convolutional neural network module and the basic probability assignment module are trained until the value of the underwater acoustic target recognition loss function is minimized, obtaining a trained underwater acoustic target recognition model.

[0014] Preferably, the construction process of the underwater acoustic target recognition loss function includes:

[0015] Construct the loss function of each underwater acoustic signal sample based on the prediction category and the true category of each underwater acoustic signal sample; the loss function is a mean square error loss function, a cross-entropy loss function or a likelihood value loss function;

[0016] Taking the expectation optimization of the loss function of each underwater acoustic signal sample with respect to its prediction probability confidence distribution as the goal, construct the recognition loss function of each underwater acoustic signal sample;

[0017] Based on the sum of the recognition loss functions of multiple underwater acoustic signal samples, obtain the underwater acoustic target recognition loss function.

[0018] Preferably, the prediction probability confidence distribution of each underwater acoustic signal sample is expressed as:

[0019] ,

[0020] where represents the prediction probability confidence distribution of the th underwater acoustic signal sample; represents the prediction probability vector of the th underwater acoustic signal sample belonging to various categories; represents the Dirichlet distribution parameter vector of the th underwater acoustic signal sample; represents the Dirichlet distribution normalization constant of the th underwater acoustic signal sample; represents the prediction probability that the th underwater acoustic signal sample belongs to the th category; represents the number of categories; represents the th underwater acoustic signal sample belonging to the Dirichlet prediction parameters for each category; For the set of prediction probabilities that the

[0021] th underwater acoustic signal sample belongs to each category;

[0022] ,

[0023] where represents the recognition loss function of the th underwater acoustic signal sample; represents the mean squared error loss function of the th underwater acoustic signal sample; represents the true probability vector that the th underwater acoustic signal sample belongs to each category; represents the true probability that the th underwater acoustic signal sample belongs to the th category; represents that the random variable follows a Dirichlet distribution with parameter vector ; represents the random variable; represents a Dirichlet distribution with parameter vector ; represents the expectation of the prediction probability that the th underwater acoustic signal sample belongs to the th category under the condition that the random variable follows a Dirichlet distribution with parameter vector ; represents the expectation of the square of the prediction probability that the th underwater acoustic signal sample belongs to the th category under the condition that the random variable follows a Dirichlet distribution with parameter vector ;

[0024] When the loss function is the cross - entropy loss function, the underwater acoustic target recognition loss function is:

[0025] ,

[0026] where represents the cross - entropy loss function of the th underwater acoustic signal sample;

[0027] When the loss function is the likelihood value loss function, the underwater acoustic target recognition loss function is:

[0028] ,

[0029] where Represents the likelihood value loss function of the th underwater acoustic signal sample.

[0030] Preferably, after obtaining the underwater acoustic target recognition loss function, it further includes:

[0031] Taking the minimization of the difference between the predicted probability confidence distribution of each underwater acoustic signal sample and the collapsed model distribution as the goal, constructing the KL divergence loss of each underwater acoustic signal sample; obtaining the target KL divergence loss function based on the sum of the KL divergence losses of multiple underwater acoustic signal samples;

[0032] Constructing an evidence collapse regularization term based on the target KL divergence loss function and the warm start parameter;

[0033] Using the evidence collapse regularization term to correct the underwater acoustic target recognition loss function, and then using the corrected underwater acoustic target recognition loss function to train the convolutional neural network module and the basic probability assignment module.

[0034] Preferably, the corrected underwater acoustic target recognition loss function is:

[0035] ,

[0036] where represents the corrected underwater acoustic target recognition loss function; represents the underwater acoustic target recognition loss function; represents the number of underwater acoustic signal samples; represents the warm start parameter; represents the target KL divergence loss function; represents the th KL divergence loss of the underwater acoustic signal sample; represents the new Dirichlet distribution parameter vector obtained by removing the Dirichlet prediction parameters of the categories with non-zero predicted probabilities from the Dirichlet distribution parameter vector of the th underwater acoustic signal sample; represents the th Dirichlet distribution parameter vector of the underwater acoustic signal sample is when the predicted probability confidence distribution; represents the all-ones Dirichlet distribution.

[0037] Preferably, the comprehensive basic probability assignment of multiple underwater acoustic signal samples belonging to each category is expressed as:

[0038] ,

[0039] where represents that multiple underwater acoustic signal samples belong to the Comprehensive basic probability assignment for each category; Indicates the predicted probability that the first underwater acoustic signal sample belongs to the th category; Indicates the predicted probability that the second underwater acoustic signal sample belongs to the th category; Indicates the predicted probability that the Nth underwater acoustic signal sample belongs to the th category; , Indicates the number of categories;

[0040] The uncertainty of the comprehensive probability distribution is expressed as:

[0041] ,

[0042] where, Indicates the uncertainty of the comprehensive probability distribution; Indicates the normalization coefficient; Indicates the uncertainty of the confidence distribution of the predicted probability of the first underwater acoustic signal sample; Indicates the uncertainty of the confidence distribution of the predicted probability of the second underwater acoustic signal sample; Indicates the uncertainty of the confidence distribution of the predicted probability of the Nth underwater acoustic signal sample.

[0043] Preferably, inputting each underwater acoustic signal sample into the convolutional neural network module includes:

[0044] Framing, windowing, and pre-emphasizing each underwater acoustic signal sample to obtain the target underwater acoustic signal sample;

[0045] Performing short-time Fourier transform on each target underwater acoustic signal sample to obtain the spectrogram of the target underwater acoustic signal sample;

[0046] Inputting the spectrogram of each target underwater acoustic signal sample into the convolutional neural network module.

[0047] Preferably, the steps of underwater acoustic recognition using the underwater acoustic target recognition model include:

[0048] Obtain the underwater acoustic signal to be recognized, input the underwater acoustic signal to be recognized into the convolutional neural network module in the underwater acoustic recognition model, and output the predicted probabilities that the underwater acoustic signal to be recognized belongs to each category;

[0049] Input the predicted probabilities that the underwater acoustic signal to be recognized belongs to each category into the basic probability assignment module in the underwater acoustic recognition model, and perform parameterization operation on the predicted probabilities that the underwater acoustic signal to be recognized belongs to each category using the Dirichlet distribution, and output the uncertainty of the confidence distribution of the predicted probability of the underwater acoustic signal to be recognized;

[0050] Based on the prediction probabilities of the underwater acoustic signals to be recognized belonging to various categories and the uncertainty of the confidence distribution of the prediction probabilities, it is determined whether there is a target in the underwater acoustic signals to be recognized.

[0051] The present invention also provides an underwater acoustic target recognition device, including:

[0052] A data acquisition module, configured to acquire multiple underwater acoustic signal samples with the same category from datasets in different sea areas;

[0053] A category probability prediction module, configured to input each underwater acoustic signal sample into a convolutional neural network module and output the prediction probabilities of each underwater acoustic signal sample belonging to various categories;

[0054] An uncertainty acquisition module, configured to input the prediction probabilities of each underwater acoustic signal sample belonging to various categories into a basic probability assignment module, perform a parameterization operation on the prediction probabilities of each underwater acoustic signal sample belonging to various categories by using a Dirichlet distribution, and output the uncertainty of the confidence distribution of the prediction probabilities of each underwater acoustic signal sample;

[0055] An evidence fusion module, configured to respectively fuse the prediction probabilities of multiple underwater acoustic signal samples belonging to various categories and the uncertainty of the confidence distribution of the prediction probabilities by using the Dempster-Shafer evidence synthesis rule, obtain the combined basic probability assignments and the combined probability distribution uncertainties of multiple underwater acoustic signal samples belonging to various categories, thereby output the prediction categories of multiple underwater acoustic signal samples, and calculate the value of the underwater acoustic target recognition loss function;

[0056] A model training and acquisition module, configured to re-acquire multiple underwater acoustic signal samples with the same category from datasets in different sea areas, train the convolutional neural network module and the basic probability assignment module until the value of the underwater acoustic target recognition loss function is minimized, and obtain a trained underwater acoustic target recognition model.

[0057] The present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned underwater acoustic target recognition method are implemented.

[0058] The underwater acoustic target recognition method provided by this application has the following beneficial effects:

[0059] 1. Parametrize the prediction probabilities of each underwater acoustic signal sample of the same category from different sea areas using the Dirichlet distribution, and output the uncertainty of the prediction probability confidence distribution of each underwater acoustic signal sample. Use this uncertainty to describe the uncertainty of the prediction probability of each underwater acoustic signal sample, considering the uncertainty that may be brought about by the differences in data characteristics in different sea areas. Then, take the prediction probabilities of each underwater acoustic signal sample belonging to each category as the evidence amount, and take the uncertainty of its prediction probability confidence distribution as the uncertainty information of the evidence amount. Introduce the Dempster-Shafer evidence combination rule to fuse the evidence amounts and uncertainty information of multiple underwater acoustic signal samples respectively, and obtain the comprehensive probability assignment and comprehensive probability distribution uncertainty of multiple underwater acoustic signals of the same category in different sea areas belonging to each category. Since the comprehensive probability assignment represents the prediction probability that multiple underwater acoustic signal samples jointly belong to each category, and the comprehensive probability distribution uncertainty represents the prediction probability uncertainty caused by integrating the data characteristic differences in multiple sea areas, the evidence fusion strategy is used to fuse the characteristic information of underwater acoustic signal samples in different sea areas and the prediction result uncertainty caused by the characteristic information differences. Finally, directly output the prediction categories of multiple underwater acoustic signal samples based on the comprehensive probability assignment and comprehensive probability distribution. Iteratively train the model by constructing a loss function, so that the model can capture the common characteristics of underwater acoustic signals of the same category in different sea areas during the training process. The model obtained after one training has high target recognition accuracy when dealing with underwater acoustic signals in different sea areas.

[0060] 2. In a conventional classification model, the prediction probability distribution output by the classification model is a function of the classification model parameters. Therefore, directly optimizing the output prediction probability distribution using the mean square error loss function, cross entropy loss function, or likelihood value loss function can complete the optimization of the classification model parameters. In this application, a basic probability assignment module is introduced, and the output layer of the convolutional neural network is parameterized using the Dirichlet distribution. Therefore, during the training process, an underwater acoustic target recognition loss function is constructed by optimizing the expectation of the mean square error loss function, cross entropy loss function, or likelihood value loss function with respect to the Dirichlet distribution, which can guide the model to focus on specific features and overall distribution rules in underwater acoustic signal samples, so that the model can fully learn the common probability distribution characteristics of underwater acoustic signal samples of the same category from different sea areas, and further improve the accuracy of the model in underwater acoustic target recognition in different sea areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to make the content of the present invention easier to be clearly understood, the following further details the present invention according to specific embodiments of the present invention in combination with the drawings, where:

[0062] Figure 1 is the flowchart of the underwater acoustic target recognition method provided by this application;

[0063] Figure 2 Schematic diagram of parameter dependency relationship of underwater acoustic target recognition model provided by this application

[0064] Figure 3 Schematic diagram of principle of underwater acoustic target recognition provided by this application

[0065] Figure 4 Schematic diagram of structure of underwater acoustic target recognition device provided by this application Specific embodiments

[0066] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, so that those skilled in the art can better understand the present invention and be able to implement it, but the embodiments cited are not intended to limit the present invention

[0067] Please refer to Figure 1 , Figure 1 as shown in the flowchart of the underwater acoustic target recognition method provided by this application. The method specifically includes

[0068] S10: Obtain multiple underwater acoustic signal samples with the same category from datasets in different sea areas

[0069] Specifically, multiple underwater acoustic signal samples with the same category refer to multiple underwater acoustic signal samples that all belong to the category with a target or all belong to the category without a target

[0070] S20: Input each underwater acoustic signal sample into the convolutional neural network module and output the prediction probability of each underwater acoustic signal sample belonging to each category

[0071] Exemplarily, the convolutional neural network module in the embodiments of this application is the ResNet18 deep residual network. The ResNet18 deep residual network solves the problems of gradient disappearance and performance saturation in deep networks by introducing residual connections. Each residual block contains multiple convolutional layers and a skip connection. This design enables the network to maintain good feature extraction performance while deepening. At the same time, the ResNet18 deep residual network has 18 convolutional layers, which enables it to extract rich features from underwater acoustic signals. Combining the characteristic that the residual connection allows the network to learn the identity mapping and the gradient disappearance, and finally using batch normalization to stabilize the training process and improve the convergence speed of the network. In addition, compared with other deep neural networks, the ResNet18 deep residual network is relatively lightweight and can be better applied to resource-constrained devices

[0072] S30: Input the prediction probabilities of each underwater acoustic signal sample belonging to various categories into the basic probability assignment module, perform parametric operations on the prediction probabilities of each underwater acoustic signal sample belonging to various categories using the Dirichlet distribution, and output the uncertainty of the prediction probability confidence distribution of each underwater acoustic signal sample.

[0073] S40: Use the Dempster-Shafer evidence synthesis rule to fuse the prediction probabilities and the uncertainty of the prediction probability confidence distribution of multiple underwater acoustic signal samples belonging to various categories respectively, obtain the comprehensive basic probability assignments and the comprehensive probability distribution uncertainties of multiple underwater acoustic signal samples belonging to various categories, thereby output the prediction categories of multiple underwater acoustic signal samples, and calculate the value of the underwater acoustic target recognition loss function.

[0074] S50: Re-obtain multiple underwater acoustic signal samples with the same categories from the datasets in different sea areas, train the convolutional neural network module and the basic probability assignment module until the value of the underwater acoustic target recognition loss function is minimized, and obtain the trained underwater acoustic target recognition model.

[0075] Furthermore, the specific implementation steps of inputting each underwater acoustic signal sample into the convolutional neural network in step S20 include:

[0076] S200: Perform frame segmentation, windowing, and pre-emphasis processing on each underwater acoustic signal sample to obtain the target underwater acoustic signal sample.

[0077] S201: Perform short-time Fourier transform on each target underwater acoustic signal sample to obtain the spectrogram of the target underwater acoustic signal sample.

[0078] S202: Input the spectrogram of each target underwater acoustic signal sample into the convolutional neural network module.

[0079] By selecting appropriate windows and window lengths to perform preprocessing such as frame segmentation, windowing, and pre-emphasis on the underwater acoustic signal samples, the data quality can be ensured; by using the short-time Fourier transform to process the preprocessed target underwater acoustic signal samples, the spectrogram of each target underwater acoustic signal sample can be obtained. Since the spectrogram contains rich time-frequency information, therefore, using the spectrogram as the input of the convolutional neural network module is beneficial for the convolutional neural network to better extract the feature information of the underwater acoustic signal samples, and thus more accurately output its prediction probabilities belonging to different categories.

[0080] Due to the complex and ever-changing marine environment, sound propagation is severely affected by the complex marine environment. Therefore, the predicted probabilities output by the convolutional neural network often have uncertainties. In practical application scenarios, if the uncertainties of the output results cannot be scientifically evaluated, it will seriously affect the accuracy of the recognition results. Therefore, after obtaining the predicted probabilities of each underwater acoustic signal sample belonging to each category, this application assumes that the predicted probability distribution of the underwater acoustic signal sample follows a Dirichlet distribution prior, and outputs the posterior distribution of the predicted probability by maximizing the probability likelihood of the observed data, so as to learn the uncertainty of the predicted probability for use in measuring the uncertainty of underwater acoustic target recognition. This application uses the Dirichlet distribution to parameterize the output layer of the convolutional neural network module, uses the parameterized result as the evidence quantity, and realizes the estimation of the uncertainty of the predicted probability, so as to effectively assist underwater acoustic target recognition using a scientific uncertainty measurement method.

[0081] Furthermore, after obtaining the predicted probabilities of each underwater acoustic signal sample belonging to each category and the uncertainty of the predicted probability confidence distribution, this application introduces an evidence fusion strategy to fuse the evidence quantities and uncertainties of underwater acoustic signal samples from different sources, enabling the model to learn the common features of underwater acoustic signal samples of the same category from different sources, thereby improving the accuracy of the model for underwater acoustic target recognition in different sea areas.

[0082] Evidence theory is an imprecise reasoning theory, also known as the Dempster-Shafer evidence combination rule (D-S evidence theory), belonging to the category of artificial intelligence, with the ability to process uncertain information. Its main feature is that it satisfies conditions weaker than Bayesian probability theory and has the ability to directly express "uncertainty" and "don't know". By using D-S evidence theory to fuse the predicted probabilities of underwater acoustic signal samples from different sources belonging to different categories and the uncertainty of their predicted probability confidence distributions, considering the conflict in the predicted probability distributions between different underwater acoustic signal samples, and then using a normalization factor to adjust to obtain the final comprehensive probability distribution, enabling the model to determine the predicted categories of multiple underwater acoustic signal samples based on the fused comprehensive probability distribution.

[0083] Specifically, the comprehensive basic probability assignment of multiple underwater acoustic signal samples belonging to each category is expressed as:

[0084] ,

[0085] where, represents the comprehensive basic probability assignment of multiple underwater acoustic signal samples belonging to the th category; represents the predicted probability of the first underwater acoustic signal sample belonging to the th category; represents the predicted probability of the second underwater acoustic signal sample belonging to the th category; Indicates the predicted probability that the Nth underwater acoustic signal sample belongs to the th category; , Indicates the number of categories;

[0086] The comprehensive probability distribution uncertainty is expressed as:

[0087] ,

[0088] where, Indicates the comprehensive probability distribution uncertainty; Indicates the normalization coefficient; Indicates the uncertainty of the predicted probability confidence distribution of the first underwater acoustic signal sample; Indicates the uncertainty of the predicted probability confidence distribution of the second underwater acoustic signal sample; Indicates the uncertainty of the predicted probability confidence distribution of the Nth underwater acoustic signal sample.

[0089] Taking two underwater acoustic signal samples A and B from different sea areas as an example, where the categories of underwater acoustic signal samples include the presence of a target and the absence of a target, the comprehensive basic probability assignment for it belonging to the category of the presence of a target is expressed as:

[0090] ,

[0091] The comprehensive basic probability assignment for it belonging to the category of the absence of a target is expressed as:

[0092] ,

[0093] The comprehensive probability distribution uncertainty is expressed as:

[0094] ,

[0095] where, Indicates the predicted probability that the underwater acoustic signal sample A belongs to the category of the presence of a target; Indicates the predicted probability that the underwater acoustic signal sample B belongs to the category of the presence of a target; Indicates the predicted probability that the underwater acoustic signal sample A belongs to the category of the absence of a target; Indicates the predicted probability that the underwater acoustic signal sample B belongs to the category of the absence of a target; Indicates the uncertainty of the predicted probability confidence distribution of the underwater acoustic signal sample A; Indicates the uncertainty of the predicted probability confidence distribution of the underwater acoustic signal sample B; Indicates the normalization coefficient, used to ensure , indicating the complete distribution of belief and uncertainty.

[0096] Furthermore, Figure 2The figure shows the parameter dependency diagram of the underwater acoustic recognition model provided by the present application. In conventional classification models, the predicted probability distribution output by the classification model is a function of the classification model parameters. Therefore, the classification model parameters can be optimized by directly optimizing the output predicted probability distribution using the mean square error loss function, cross entropy loss function, or likelihood loss function. However, the present application introduces a basic probability allocation module and uses the Dirichlet distribution to parameterize the output layer of the convolutional neural network. Therefore, during the training process, the model parameters are optimized by optimizing the mean square error loss function, cross entropy loss function, or likelihood loss function with respect to the expectation of the Dirichlet distribution. Specifically, the construction process of the underwater acoustic target recognition loss function includes:

[0097] A loss function is constructed for each underwater acoustic signal sample based on the predicted category and the true category of each underwater acoustic signal sample; the loss function is a mean square error loss function, a cross entropy loss function, or a likelihood loss function.

[0098] Taking the expected optimization of the loss function of each underwater acoustic signal sample with respect to its predicted probability confidence distribution as the goal, the recognition loss function of each underwater acoustic signal sample is constructed.

[0099] The underwater acoustic target recognition loss function is obtained by summing the recognition loss functions of multiple underwater acoustic signal samples.

[0100] Specifically, the predicted probability confidence distribution of each underwater acoustic signal sample is expressed as:

[0101] ,

[0102] in, Indicates the The predicted probability confidence distribution of underwater acoustic signal samples; Indicates the The predicted probability vector of each underwater acoustic signal sample belonging to each category; Indicates the Dirichlet distribution parameter vector of underwater acoustic signal samples; Indicates the The normalization constant of the Dirichlet distribution of underwater acoustic signal samples; Indicates the The underwater acoustic signal sample belongs to The predicted probability of each category; Indicates the number of categories; Indicates the The underwater acoustic signal sample belongs to Dirichlet prediction parameters for each class; For the The predicted probability of underwater acoustic signal samples belonging to each category is a set, The sum of all predicted probabilities in is 1.

[0103] When the loss function is the mean squared error loss function, the recognition loss function for each underwater acoustic signal sample is:

[0104] ,

[0105] where represents the recognition loss function of the -th underwater acoustic signal sample; represents the mean squared error loss function of the -th underwater acoustic signal sample; represents the true probability vector of the -th underwater acoustic signal sample belonging to each category; represents the true probability that the -th underwater acoustic signal sample belongs to the -th category; represents that the random variable follows a Dirichlet distribution with parameter vector ; represents the random variable; represents the Dirichlet distribution with parameter vector ; represents the expectation of the predicted probability that the -th underwater acoustic signal sample belongs to the -th category under the condition that the random variable follows a Dirichlet distribution with parameter vector ; represents the expectation of the square of the predicted probability that the -th underwater acoustic signal sample belongs to the -th category under the condition that the random variable follows a Dirichlet distribution with parameter vector ;

[0106] When the loss function is the cross-entropy loss function, the underwater acoustic target recognition loss function is:

[0107] ,

[0108] where represents the cross-entropy loss function of the -th underwater acoustic signal sample.

[0109] When the loss function is the likelihood value loss function, the underwater acoustic target recognition loss function is:

[0110] ,

[0111] where represents the likelihood value loss function of the -th underwater acoustic signal sample.

[0112] Furthermore, during the model training process, we hope that when an underwater acoustic signal sample is misrecognized, the uncertainty of its predicted probability confidence distribution is relatively high, so as to reflect that the current prediction result is unreliable. Therefore, based on the underwater acoustic target recognition loss function, this application introduces an evidence collapse regularization term. Since in the actual application scenario, underwater acoustic signals usually clearly belong to the target-present class or the target-absent class, and the collapsed model distribution is more in line with this actual situation. Therefore, after using the evidence collapse regularization term to correct the underwater acoustic target recognition loss function, it can not only fit the consistency between the predicted probability and the label, but also make the predicted probability confidence distribution output by the model close to the collapsed model distribution, enabling the model to output a more definite predicted category for the underwater acoustic signal sample, rather than an ambiguous probability distribution, reducing the uncertainty of the recognition result.

[0113] Specifically, after obtaining the underwater acoustic target recognition loss function, it further includes:

[0114] Taking the minimization of the difference between the predicted probability confidence distribution of each underwater acoustic signal sample and the collapsed model distribution as the goal, constructing the KL divergence loss of each underwater acoustic signal sample; obtaining the target KL divergence loss function based on the sum of the KL divergence losses of multiple underwater acoustic signal samples.

[0115] Constructing an evidence collapse regularization term based on the target KL divergence loss function and the warm start parameter.

[0116] Using the evidence collapse regularization term to correct the underwater acoustic target recognition loss function, and then using the corrected underwater acoustic target recognition loss function to train the convolutional neural network module and the basic probability assignment module.

[0117] Specifically, the corrected underwater acoustic target recognition loss function is:

[0118] ,

[0119] Where, represents the corrected underwater acoustic target recognition loss function; represents the underwater acoustic target recognition loss function; represents the number of underwater acoustic signal samples; represents the warm start parameter; represents the target KL divergence loss function; represents the KL divergence loss of the th underwater acoustic signal sample; represents the new Dirichlet distribution parameter vector obtained by removing the Dirichlet prediction parameters of the categories with non-zero predicted probabilities from the Dirichlet distribution parameter vector of the th underwater acoustic signal sample; The Dirichlet distribution parameter vector of a hydroacoustic signal sample is The predicted probability confidence distribution when Indicates the Dirichlet distribution of all 1s.

[0120] Such as Figure 3 As shown in the schematic diagram of the hydroacoustic target recognition principle provided by this application, it specifically includes three parts: data preprocessing, deep evidence learning structure, and decision fusion strategy based on evidence theory. By preprocessing multiple hydroacoustic signal samples from different sources, obtaining their spectrograms and inputting them into a convolutional neural network respectively, the predicted probability of each hydroacoustic signal sample belonging to each category is output. Then, through the basic probability assignment module, the uncertainty of the predicted probability confidence distribution of each hydroacoustic signal sample is obtained. The predicted probability of each hydroacoustic signal sample belonging to each category is used as the evidence quantity, and the uncertainty information of the predicted probability confidence distribution of each hydroacoustic signal sample is used as the uncertainty information of the evidence quantity. Finally, an evidence fusion strategy is introduced to fuse the evidence quantities and uncertainty information of multiple hydroacoustic signal samples respectively to obtain the comprehensive evidence quantity and uncertainty information, so as to obtain the predicted categories of multiple hydroacoustic signal samples. Since the model continuously learns the common characteristics of hydroacoustic signals of the same category in different sea areas during the training process, the trained model has high target recognition accuracy when dealing with complex marine environments, providing new ideas and technical means for solving the problems existing in underwater target recognition and having high application value.

[0121] The embodiments of this application also provide specific experimental examples to verify the effectiveness of the above method. Specifically, actual ship radiated noise data under actual sea conditions collected in 3 different marine environments 3 times in a hydroacoustic database is obtained. Among them, the data in 3 different marine environments collected in the same batch is a data set, which altogether includes three data sets A, B, and C. This application randomly selects data with the same sample number in the three data sets A, B, and C it collected. The specific data description is shown in Table 1:

[0122] Table 1

[0123]

[0124] Based on the above data, the embodiments of this application conduct specific target recognition experiment tests, and the specific steps are as follows:

[0125] 1. Draw a hydroacoustic signal spectrogram data set, specifically including:

[0126] 1) Split all long hydroacoustic signals into audio signals of one second per segment;

[0127] 2) Perform short-time Fourier transform and dimensional transformation on each audio signal to obtain the spectrogram corresponding to each audio signal. The size of the spectrogram is 496*369*3. Taking the audio signal per second as the unit, perform Fourier transform using a 256-point window, and set the number of overlapping samples and the number of points for Fourier transform to 128 and 256 respectively;

[0128] 3) Label the data in the dataset as having a target and not having a target respectively;

[0129] 4) Divide the dataset into a training set and a test set. Specifically, for the data with and without a target, randomly select 80% as the training set and the remaining 20% as the test set.

[0130] 2. Model construction set probability prediction fusion, specifically including:

[0131] 1) Use the spectrogram of the underwater acoustic signal in each dataset as the input, take the deep neural network DNN as the convolutional neural network module, and output the prediction probability that each underwater acoustic signal belongs to each category, that is, the evidence vector e. Connect the evidence vector e to the parameters of the Dirichlet distribution in the basic probability assignment module through subjective logic, calculate its prediction probability confidence distribution and its uncertainty as subjective opinions, and finally use the Dempster-Shafer evidence synthesis rule to fuse the prediction probabilities and the prediction probability confidence distribution uncertainties of the underwater acoustic signals of the same category in different datasets to obtain a fused opinion;

[0132] 2) Adjust the traditional mean square error loss function to the form of integral on the Dirichlet distribution to obtain the underwater acoustic target recognition loss function, incorporate the KL divergence term into the underwater acoustic target recognition loss function, and regulate the prediction probability distribution by punishing the evidence generated by the wrong label;

[0133] 3) Obtain the value of the underwater acoustic target recognition loss function through the forward propagation of the model, measure the matching degree between the model prediction result and the actual result, continuously correct the model parameters to minimize the value of the underwater acoustic target recognition loss function, and update the weights of each module in the model through the Adam optimizer.

[0134] 3. For the 3 datasets, perform deep evidence fusion pairwise to obtain multiple fused datasets, and use each fused dataset to train and test the model respectively.

[0135] In the embodiments of this application, the ordinary ResNet18 deep residual network and the underwater acoustic target recognition model provided in this application are respectively used to perform recognition experiments on the data in the fused dataset. The recognition result accuracies of the two models and the uncertainty of the recognition results provided by the model in this application are shown in Table 2:

[0136] Table 2

[0137]

[0138] As can be seen from the data in Table 2, for the underwater acoustic signal datasets integrated with data from different sea areas, the accuracy of the recognition results of the underwater acoustic recognition model provided by this application is higher than 98.5% for all, and the highest reaches 99.87%. However, the accuracy of the underwater acoustic target recognition results of the traditional recognition model for the merged dataset is relatively low. This indicates that the underwater acoustic target recognition method provided by this application improves the ability to scientifically evaluate the uncertainty of the model recognition results by introducing the amount of evidence and uncertainty information, and by fusing the amount of evidence and uncertainty of underwater acoustic signals from different sources, the model can fully learn the common features of data of the same category in different sea areas, improving the accuracy and reliability of underwater acoustic target recognition.

[0139] Based on the underwater acoustic target recognition method provided in the above embodiments, an embodiment of this application also provides an underwater acoustic target recognition device, as Figure 4 shown. The device specifically includes:

[0140] A data acquisition module 10, configured to acquire multiple underwater acoustic signal samples with the same category from datasets in different sea areas.

[0141] A category probability prediction module 20, configured to input each underwater acoustic signal sample into a convolutional neural network module and output the prediction probability of each underwater acoustic signal sample belonging to each category.

[0142] An uncertainty acquisition module 30, configured to input the prediction probability of each underwater acoustic signal sample belonging to each category into a basic probability assignment module, perform a parameterization operation on the prediction probability of each underwater acoustic signal sample belonging to each category using the Dirichlet distribution, and output the uncertainty of the prediction probability confidence distribution of each underwater acoustic signal sample.

[0143] An evidence fusion module 40, configured to use the Dempster-Shafer evidence combination rule to fuse the prediction probabilities and the uncertainties of the prediction probability confidence distributions of multiple underwater acoustic signal samples belonging to each category respectively, obtain the comprehensive basic probability assignments and the comprehensive probability distribution uncertainties of multiple underwater acoustic signal samples belonging to each category, thereby output the prediction categories of multiple underwater acoustic signal samples, and calculate the value of the underwater acoustic target recognition loss function.

[0144] A model training and acquisition module 50, configured to re-acquire multiple underwater acoustic signal samples with the same category from datasets in different sea areas, train the convolutional neural network module and the basic probability assignment module until the value of the underwater acoustic target recognition loss function is the smallest, and obtain the trained underwater acoustic target recognition model.

[0145] The embodiments of the present application also provide a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned underwater acoustic target recognition method are implemented.

[0146] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0147] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0148] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0149] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Therefore, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0150] Obviously, the above embodiments are merely examples given for clear illustration and are not limitations on the implementation manners. For those of ordinary skill in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to enumerate all implementation manners here. And the obvious changes or modifications derived therefrom still fall within the protection scope of the present invention.

Claims

1. An underwater acoustic target recognition method, characterized in that, Including: Obtain multiple underwater acoustic signal samples of the same category from datasets in different sea areas; Input each underwater acoustic signal sample into a convolutional neural network module and output the predicted probabilities of each underwater acoustic signal sample belonging to various categories; Input the predicted probabilities of each underwater acoustic signal sample belonging to various categories into a basic probability assignment module, and perform a parameterization operation on the predicted probabilities of each underwater acoustic signal sample belonging to various categories using the Dirichlet distribution, and output the prediction probability confidence distribution uncertainty of each underwater acoustic signal sample; Use the Dempster-Shafer evidence combination rule to fuse the predicted probabilities and prediction probability confidence distribution uncertainties of multiple underwater acoustic signal samples belonging to various categories respectively, obtain the comprehensive basic probability assignments and comprehensive probability distribution uncertainties of multiple underwater acoustic signal samples belonging to various categories, thereby output the predicted categories of multiple underwater acoustic signal samples, and calculate the value of the underwater acoustic target recognition loss function; Re-obtain multiple underwater acoustic signal samples of the same category from datasets in different sea areas, and train the convolutional neural network module and the basic probability assignment module until the value of the underwater acoustic target recognition loss function is minimized, and obtain a trained underwater acoustic target recognition model, which specifically includes: Taking the minimization of the difference between the prediction probability confidence distribution of each underwater acoustic signal sample and the collapsed model distribution as the goal, construct the KL divergence loss of each underwater acoustic signal sample; obtain the target KL divergence loss function based on the sum of the KL divergence losses of multiple underwater acoustic signal samples; Construct an evidence collapse regularization term based on the target KL divergence loss function and the warm start parameter; Use the evidence collapse regularization term to correct the underwater acoustic target recognition loss function, and re-obtain multiple underwater acoustic signal samples of the same category from datasets in different sea areas, and use the corrected underwater acoustic target recognition loss function to train the convolutional neural network module and the basic probability assignment module.

2. The underwater acoustic target recognition method according to claim 1, wherein The construction process of the underwater acoustic target recognition loss function includes: Construct the loss function of each underwater acoustic signal sample based on the predicted category and the true category of each underwater acoustic signal sample; the loss function is the mean square error loss function, the cross-entropy loss function or the likelihood value loss function; Taking the optimization of the expectation of the loss function of each underwater acoustic signal sample with respect to its prediction probability confidence distribution as the goal, construct the recognition loss function of each underwater acoustic signal sample; Obtain the underwater acoustic target recognition loss function based on the sum of the recognition loss functions of multiple underwater acoustic signal samples.

3. The underwater acoustic target recognition method according to claim 2, characterized in that, The prediction probability confidence distribution of each underwater acoustic signal sample is expressed as: ， Among them, represents the predicted probability confidence distribution of the th underwater acoustic signal sample; represents the predicted probability vector of the th underwater acoustic signal sample belonging to each category; represents the Dirichlet distribution parameter vector of the th underwater acoustic signal sample; represents the Dirichlet distribution normalization constant of the th underwater acoustic signal sample; represents the predicted probability that the th underwater acoustic signal sample belongs to the th category; represents the number of categories; represents the Dirichlet prediction parameter that the th underwater acoustic signal sample belongs to the th category; is the set composed of the predicted probabilities that the th underwater acoustic signal sample belongs to each category; When the loss function is the mean square error loss function, the recognition loss function of each underwater acoustic signal sample is: ， Among them, represents the recognition loss function of the th underwater acoustic signal sample; represents the mean square error loss function of the th underwater acoustic signal sample; represents the true probability vector of the th underwater acoustic signal sample belonging to various categories; represents the true probability that the th underwater acoustic signal sample belongs to the th category; represents that the random variable follows a Dirichlet distribution with a parameter vector of ; represents a random variable; represents a Dirichlet distribution with a parameter vector of ; represents the expectation of the predicted probability that the th underwater acoustic signal sample belongs to the th category under the condition that the random variable follows a Dirichlet distribution with a parameter vector of ; represents the expectation of the square of the predicted probability that the th underwater acoustic signal sample belongs to the th category under the condition that the random variable follows a Dirichlet distribution with a parameter vector of ; When the loss function is the cross-entropy loss function, the underwater acoustic target recognition loss function is: ， Among them, represents the cross-entropy loss function of the -th underwater acoustic signal sample; When the loss function is the likelihood value loss function, the underwater acoustic target recognition loss function is: ， Among them, represents the likelihood value loss function of the nth underwater acoustic signal sample.

4. The underwater acoustic target recognition method according to claim 1, wherein, The corrected underwater acoustic target recognition loss function is: ， Among them, represents the corrected loss function for underwater acoustic target recognition; represents the loss function for underwater acoustic target recognition; represents the number of underwater acoustic signal samples; represents the warm start parameter; represents the target KL divergence loss function; represents the KL divergence loss of the th underwater acoustic signal sample; represents the Dirichlet distribution parameter vector obtained by removing the Dirichlet prediction parameters of the categories with non-zero prediction probabilities from the th underwater acoustic signal sample; represents the prediction probability confidence distribution when the Dirichlet distribution parameter vector of the th underwater acoustic signal sample is represents the all-ones Dirichlet distribution.

5. The underwater acoustic target recognition method according to claim 1, characterized in that, The comprehensive basic probability assignment of multiple underwater acoustic signal samples belonging to various categories is expressed as: ， Among them, represents the combined basic probability assignment that multiple underwater acoustic signal samples belong to the th category; represents the predicted probability that the first underwater acoustic signal sample belongs to the th category; represents the predicted probability that the second underwater acoustic signal sample belongs to the th category; represents the predicted probability that the Nth underwater acoustic signal sample belongs to the th category; , represents the number of categories; The comprehensive probability distribution uncertainty is expressed as: ， Among them, represents the comprehensive probability distribution uncertainty; represents the normalization coefficient; represents the prediction probability confidence distribution uncertainty of the first underwater acoustic signal sample; represents the prediction probability confidence distribution uncertainty of the second underwater acoustic signal sample; represents the prediction probability confidence distribution uncertainty of the Nth underwater acoustic signal sample.

6. The underwater acoustic target recognition method according to claim 1, characterized in that Inputting each underwater acoustic signal sample into the convolutional neural network module includes: Perform frame segmentation, windowing and pre-emphasis processing on each underwater acoustic signal sample to obtain the target underwater acoustic signal sample; Perform short-time Fourier transform on each target underwater acoustic signal sample to obtain the spectrogram of the target underwater acoustic signal sample; Input the spectrogram of each target underwater acoustic signal sample into the convolutional neural network module.

7. The underwater acoustic target recognition method according to claim 1, wherein The steps for underwater acoustic recognition using the underwater acoustic target recognition model include: Obtain the underwater acoustic signal to be recognized, input the underwater acoustic signal to be recognized into the convolutional neural network module in the underwater acoustic recognition model, and output the prediction probabilities of the underwater acoustic signal to be recognized belonging to various categories; Input the prediction probabilities of the underwater acoustic signal to be recognized belonging to various categories into the basic probability assignment module in the underwater acoustic recognition model, perform parametric operation on the prediction probabilities of the underwater acoustic signal to be recognized belonging to various categories using the Dirichlet distribution, and output the uncertainty of the prediction probability confidence distribution of the underwater acoustic signal to be recognized; Based on the prediction probabilities of the underwater acoustic signal to be recognized belonging to various categories and the uncertainty of the prediction probability confidence distribution, determine whether there is a target in the underwater acoustic signal to be recognized.

8. An underwater acoustic target recognition device, characterized in that, Include: A data acquisition module for obtaining multiple underwater acoustic signal samples of the same category from datasets in different sea areas; A category probability prediction module for inputting each underwater acoustic signal sample into the convolutional neural network module and outputting the prediction probabilities of each underwater acoustic signal sample belonging to various categories; An uncertainty acquisition module for inputting the prediction probabilities of each underwater acoustic signal sample belonging to various categories into the basic probability assignment module, performing parametric operation on the prediction probabilities of each underwater acoustic signal sample belonging to various categories using the Dirichlet distribution, and outputting the uncertainty of the prediction probability confidence distribution of each underwater acoustic signal sample; An evidence fusion module for using the Dempster-Shafer evidence combination rule to fuse the prediction probabilities of multiple underwater acoustic signal samples belonging to various categories and the uncertainty of the prediction probability confidence distribution respectively, obtaining the comprehensive basic probability assignment and the comprehensive probability distribution uncertainty of multiple underwater acoustic signal samples belonging to various categories, thereby outputting the prediction categories of multiple underwater acoustic signal samples, and calculating the value of the underwater acoustic target recognition loss function; A model training and acquisition module for re-obtaining multiple underwater acoustic signal samples of the same category from datasets in different sea areas, training the convolutional neural network module and the basic probability assignment module until the value of the underwater acoustic target recognition loss function is minimized, and obtaining the trained underwater acoustic target recognition model; where, after constructing the underwater acoustic target loss function, it further includes: Taking the minimization of the difference between the prediction probability confidence distribution of each underwater acoustic signal sample and the collapsed model distribution as the goal, constructing the KL divergence loss of each underwater acoustic signal sample; obtaining the target KL divergence loss function based on the sum of the KL divergence losses of multiple underwater acoustic signal samples; Constructing an evidence collapse regularization term based on the target KL divergence loss function and the warm start parameter; Using the evidence collapse regularization term to correct the underwater acoustic target recognition loss function, and re-obtaining multiple underwater acoustic signal samples of the same category from datasets in different sea areas, and training the convolutional neural network module and the basic probability assignment module using the corrected underwater acoustic target recognition loss function.

9. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the steps of the underwater acoustic target recognition method according to any one of claims 1 to 7 are implemented.

Images