Method and device for detecting strength of marine antenna supporting structure and storage medium

Abstract

The invention discloses a method and device for detecting the strength of a marine antenna supporting structure and a storage medium, and the method comprises the steps: respectively reading an attitude signal collected by an inertial measurement unit, a vibration signal collected by a vibration sensor and a resistance signal collected by a resistance sensor when a ship sails on a sea area; inputting the attitude signal into a first encoder to extract target attitude features; inputting the vibration signal into a second encoder to extract a first target vibration feature; inputting the resistance signal into a third encoder to extract a target resistance feature; inputting the target attitude feature, the first target vibration feature and the target resistance feature into a decoder, and decoding the target attitude feature, the first target vibration feature and the target resistance feature into a target multi-modal feature; inputting the target multi-modal feature into the classification strength of the supporting structure of the maritime antenna in the head structure; and performing alarm operation on the supporting structure of the maritime antenna according to the strength. According to the embodiment, the risk that the maritime antenna is damaged due to the fact that the supporting structure of the maritime antenna is damaged in the ship navigation process is reduced, and the ship navigation safety is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of deep learning, and in particular to a method, device and storage medium for detecting the strength of a support structure of a navigation antenna. Background Art

[0002] Maritime antennas are installed on ships. They are one of the components of ship communication and navigation systems. They realize communication, navigation and other functions in maritime activities. When ships sail on the ocean, environmental factors such as seawater and salt spray cause certain corrosion to the various structures of maritime antennas.

[0003] At present, small ships such as yachts and fishing boats lack professionals. Their owners usually arrange professionals to inspect and maintain maritime antennas at regular intervals. The inspection intervals vary in length, resulting in low inspection accuracy. There is a risk of maritime antennas being damaged due to structural corrosion and damage during navigation. Summary of the invention

[0004] In view of this, the present invention provides a method, device and storage medium for detecting the strength of a marine antenna support structure, so as to improve the accuracy of maintenance of a radome structure on a ship.

[0005] A first aspect of the present invention provides a method for detecting the strength of a navigation antenna support structure, comprising:

[0006] When the ship is sailing on the sea, the inertial measurement unit, the vibration sensor and the resistance sensor are respectively started; the inertial measurement unit is deployed on the hull of the ship, and the vibration sensor and the resistance sensor are both deployed on the support structure of the maritime antenna of the ship;

[0007] Respectively reading the attitude signal collected by the inertial measurement unit, the vibration signal collected by the vibration sensor, and the resistance signal collected by the resistance sensor;

[0008] Determine a structural classification network; the structural classification network includes a first encoder, a second encoder, a third encoder, a decoder and a head structure;

[0009] Inputting the posture signal into the first encoder to extract the target posture feature;

[0010] Inputting the vibration signal into the second encoder to extract a first target vibration feature;

[0011] Inputting the resistance signal into the third encoder to extract the target resistance feature;

[0012] Inputting the target posture feature, the first target vibration feature and the target resistance feature into the decoder to be decoded into a target multimodal feature;

[0013] Inputting the target multimodal features into the head structure to classify the strength of the support structure of the maritime antenna;

[0014] An alarm operation is performed on the supporting structure of the maritime antenna according to the strength.

[0015] A second aspect of the present invention provides a device for detecting the strength of a support structure of a navigation antenna, comprising:

[0016] A sensor start module, used to start an inertial measurement unit, a vibration sensor and a resistance sensor respectively when the ship is sailing on the sea; the inertial measurement unit is deployed on the hull of the ship, and the vibration sensor and the resistance sensor are both deployed on the support structure of the maritime antenna of the ship;

[0017] A signal reading module, used to read the attitude signal collected by the inertial measurement unit, the vibration signal collected by the vibration sensor, and the resistance signal collected by the resistance sensor respectively;

[0018] A structure classification network determination module, used to determine a structure classification network; the structure classification network includes a first encoder, a second encoder, a third encoder, a decoder and a head structure;

[0019] A target posture feature extraction module, used for inputting the posture signal into the first encoder to extract the target posture feature;

[0020] A target vibration feature extraction module, used for inputting the vibration signal into the second encoder to extract a first target vibration feature;

[0021] A target resistance feature extraction module, used for inputting the resistance signal into the third encoder to extract the target resistance feature;

[0022] A target multimodal feature decoding module, used for inputting the target posture feature, the first target vibration feature and the target resistance feature into the decoder to decode them into a target multimodal feature;

[0023] a strength classification module, for inputting the target multimodal features into the head structure to classify the strength of the support structure of the maritime antenna;

[0024] An alarm operation execution module is used to execute an alarm operation on the supporting structure of the maritime antenna according to the strength.

[0025] A third aspect of the present invention provides an electronic device, the electronic device comprising:

[0026] at least one processor; and

[0027] a memory communicatively connected to the at least one processor; wherein,

[0028] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for detecting the strength of the support structure of the navigation antenna as described in the first aspect above.

[0029] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the method for detecting the strength of a support structure of a navigation antenna as described in the first aspect above.

[0030] A fifth aspect of the present invention provides a computer program product, the computer program product comprising a computer program, and when the computer program is executed by a processor, the method for detecting the strength of a support structure of a navigation antenna as described in the first aspect above is implemented.

[0031] In this embodiment, when the ship is sailing on the sea, the inertial measurement unit, the vibration sensor and the resistance sensor are started respectively; the inertial measurement unit is deployed on the hull of the ship, and the vibration sensor and the resistance sensor are deployed on the support structure of the maritime antenna of the ship; the attitude signal collected by the inertial measurement unit, the vibration signal collected by the vibration sensor and the resistance signal collected by the resistance sensor are read respectively; the structural classification network is determined; the structural classification network includes a first encoder, a second encoder, a third encoder, a decoder and a head structure; the attitude signal is input into the first encoder to extract the target attitude feature; the vibration signal is input into the second encoder to extract the first target vibration feature; the resistance signal is input into the third encoder to extract the target resistance feature; the target attitude feature, the first target vibration feature and the target resistance feature are input into the decoder to be decoded into the target multimodal feature; the target multimodal feature is input into the head structure to classify the strength of the support structure of the maritime antenna; and an alarm operation is performed on the support structure of the maritime antenna according to the strength. This embodiment explores the characteristics of the support structure of the maritime antenna in terms of vibration, resistance, etc. to evaluate the strength of the support structure of the maritime antenna, thereby performing an alarm operation, realizing self-inspection of the support structure of the maritime antenna, improving the accuracy of inspection and maintenance of the support structure of the maritime antenna in the ship, reducing the risk of damage to the maritime antenna due to damage to the support structure of the maritime antenna during the navigation of the ship, and ensuring the safety of the navigation of the ship.

[0032] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 It is a flow chart of a method for detecting the strength of a navigation antenna support structure provided in Embodiment 1 of the present invention.

[0035] Figure 2 It is a schematic diagram of a ship and a maritime antenna provided in Embodiment 1 of the present invention.

[0036] Figure 3 It is a schematic diagram of a structural classification network provided in Example 1 of the present invention.

[0037] Figure 4 It is a structural schematic diagram of a device for detecting the strength of a support structure of a navigation antenna provided in a second embodiment of the present invention.

[0038] Figure 5 It is a structural schematic diagram of an electronic device provided in Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0040] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can cover sequential implementations other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0041] Embodiment 1

[0042] See also Figure 1, shows a flow chart of a method for detecting the strength of a support structure of a navigation antenna provided by a first embodiment of the present invention. The method can be executed by a device for detecting the strength of a support structure of a navigation antenna. The device for detecting the strength of a support structure of a navigation antenna can be implemented in the form of hardware and / or software. The device for detecting the strength of a support structure of a navigation antenna can be configured in an electronic device. Figure 1 As shown, the method includes:

[0043] Step 101: When the ship is sailing on the sea, the inertial measurement unit, the vibration sensor and the resistance sensor are started respectively.

[0044] like Figure 2 As shown, one or more maritime antennas 201 are installed on the ship 200 , and the manufacturer of the maritime antennas provides various maintenance services for the maritime antennas 201 on the ship 200 .

[0045] In this embodiment, when the ship is sailing on the sea, the inertial measurement unit (IMU), the vibration sensor and the resistance sensor can be started respectively.

[0046] Among them, the inertial measurement unit is deployed on the hull of the ship, and the vibration sensor and the resistance sensor are deployed on the supporting structure of the ship's maritime antenna.

[0047] Step 102 : respectively reading the attitude signal collected by the inertial measurement unit, the vibration signal collected by the vibration sensor, and the resistance signal collected by the resistance sensor.

[0048] In practical applications, the inertial measurement unit collects the attitude signal of the ship hull at a preset first frequency, the vibration sensor collects the vibration signal of the supporting structure of the maritime antenna at a preset second frequency, and the resistance sensor collects the resistance signal of the supporting structure of the maritime antenna at a preset third frequency.

[0049] At this time, the attitude signal collected by the inertial measurement unit, the vibration signal collected by the vibration sensor, and the resistance signal collected by the resistance sensor can be continuously read respectively.

[0050] Step 103: Determine the structural classification network.

[0051] In this embodiment, a structure classification network can be constructed and trained in advance based on deep learning, where Figure 3 As shown, the structural classification network includes a first encoder Encoder1, a second encoder Encoder2, a third encoder Encoder3, a decoder Decoder and a head structure Head.

[0052] The first encoder Encoder1 is used to extract features from the attitude signal, the second encoder Encoder2 is used to extract features from the vibration signal, the third encoder Encoder3 is used to extract features from the resistance signal, the decoder Decoder is used to decode the features of the attitude signal, the vibration signal and the resistance signal into new features, and the head structure Head is used to perform multi-classification tasks based on the decoded features and output the strength of the supporting structure of the maritime antenna.

[0053] When training the structure classification network, the attitude signals collected by the historical inertial measurement unit, the vibration signals collected by the vibration sensor, and the resistance signals collected by the resistance sensor are selected as samples. The technicians annotate the strength of the supporting structure of the corresponding maritime antenna as the label Label, select cross entropy as the loss function, and perform supervised training on the structure classification network.

[0054] Step 104: Input the posture signal into the first encoder to extract the target posture feature.

[0055] In this embodiment, if Figure 3 As shown, a sequence composed of the posture signals Pose of the ship's hull is input into the first encoder Encoder1 for encoding, and target posture features are extracted from the sequence composed of the posture signals of the ship's hull.

[0056] In one design, Figure 3 As shown, the first encoder Encoder1 includes a first convolutional layer Conv1, a second convolutional layer Conv2, a third convolutional layer Conv3 and a first fully connected layer FC1.

[0057] In this design, the transpose of the posture signal Pose is multiplied by the posture signal to obtain the posture matrix. Then, the posture matrix can be expressed as P T *P, where P is the attitude signal and T is the transpose.

[0058] The posture matrix is ​​input into the first fully connected layer FC1 and mapped into the first candidate posture feature to achieve dimensionality reduction and alignment.

[0059] The first candidate posture feature is input into the first convolutional layer Conv1 to perform a convolution operation to obtain the second candidate posture feature.

[0060] The second candidate posture feature is input into the second convolutional layer Conv2 to perform a convolution operation to obtain a third candidate posture feature.

[0061] The first candidate posture feature and the third candidate posture feature are concatenated to obtain a fourth candidate posture feature.

[0062] The fourth candidate posture feature is input into the third convolutional layer Conv3 to perform a convolution operation to obtain the target posture feature.

[0063] Step 105: Input the vibration signal into a second encoder to extract the first target vibration feature.

[0064] In this embodiment, if Figure 3 As shown, a sequence composed of vibration signals Vibration of the support structure of the maritime antenna is input into the second encoder Encoder2 for encoding, and a first target vibration feature is extracted from the sequence composed of vibration signals Vibration of the support structure of the maritime antenna.

[0065] In practical applications, maritime antennas generate vibration signals as ships drift on the sea.

[0066] When the supporting structure of the maritime antenna is damaged, the stiffness and mass distribution of the supporting structure of the maritime antenna change, causing the natural frequency of the vibration signal to change. Some originally suppressed vibration modes may be excited, resulting in an increase in the amplitude at a specific frequency.

[0067] When the support structure of the maritime antenna is corroded, the material properties of the support structure of the maritime antenna are degraded, the effective cross-sectional area is reduced, the overall stiffness of the structure is reduced, and the corrosion is often localized, which will lead to uneven distribution of structural mass and may cause some additional low-frequency vibration components. The vibration response under the same excitation increases and the amplitude increases. Especially when the corrosion develops to a certain extent and the structure becomes loose locally, the increase in amplitude will be more obvious.

[0068] Therefore, there is a significant difference between the vibration signal generated when the supporting structure of the maritime antenna is intact and the vibration signal generated when the supporting structure of the maritime antenna is damaged or corroded. The vibration signal of the supporting structure of the maritime antenna can be used to detect the strength of the supporting structure of the maritime antenna.

[0069] In one design, Figure 3 As shown, the second encoder Encoder2 includes a fourth convolutional layer Conv4, a fifth convolutional layer Conv5, a sixth convolutional layer Conv6 and a second fully connected layer FC2.

[0070] In this design, DWT (Discrete Wavelet Transform) and other algorithms are used to perform time-frequency conversion on the vibration signal to obtain a vibration time-frequency graph. Converting the vibration signal into a vibration time-frequency graph helps to explore the characteristics of the vibration signal in terms of frequency, amplitude, etc.

[0071] The vibration time-frequency graph is input into the second fully connected layer FC2 and mapped as the first candidate vibration feature to achieve dimensionality reduction and alignment.

[0072] The first candidate vibration feature is input into the fourth convolutional layer Conv4 to perform a convolution operation to obtain a second candidate vibration feature.

[0073] The second candidate vibration feature is input into the fifth convolutional layer Conv5 to perform a convolution operation to obtain a third candidate vibration feature.

[0074] The first candidate vibration feature and the third candidate vibration feature are concatenated to obtain a fourth candidate vibration feature.

[0075] The fourth candidate vibration feature is input into the sixth convolutional layer Conv6 to perform a convolution operation to obtain the target vibration feature.

[0076] Step 106: Input the resistance signal into a third encoder to extract target resistance characteristics.

[0077] In this embodiment, if Figure 3 As shown, the sequence composed of the resistance signal Resistance of the support structure of the maritime antenna is input into the third encoder Encoder3 for encoding, and the target resistance feature is extracted from the sequence composed of the resistance signal Resistance of the support structure of the maritime antenna.

[0078] When the support structure of the maritime antenna is damaged, the current path becomes longer or disconnected, and the resistance increases. The damage of the support structure of the maritime antenna may cause the contact point to loosen. Under the influence of vibration or other external factors, the contact resistance will change, resulting in large fluctuations in the resistance signal.

[0079] When the supporting structure of a maritime antenna corrodes, an oxide film or other corrosion products will form on the surface. The conductivity of these products is usually worse than that of the metal itself, which is equivalent to adding a resistance element in the current path, increasing the overall resistance.

[0080] Therefore, there is a significant difference between the resistance signal generated when the supporting structure of the maritime antenna is intact and the resistance signal generated when the supporting structure of the maritime antenna is damaged or corroded. The resistance signal of the supporting structure of the maritime antenna can be used to detect the strength of the supporting structure of the maritime antenna.

[0081] In one design, Figure 3 As shown, the third encoder Encoder3 includes a seventh convolutional layer Conv7, an eighth convolutional layer Conv8, a ninth convolutional layer Conv9 and a third fully connected layer FC3.

[0082] In this design, the transpose of the resistance signal is multiplied by the resistance signal to obtain the resistance matrix. Then, the resistance matrix can be expressed as R T*R, where R is the resistance signal and T is the transpose. Converting the resistance signal into a resistance matrix helps to mine the characteristics of the resistance signal when it fluctuates.

[0083] The resistance matrix is ​​input into the third fully connected layer FC3 and mapped into the first candidate resistance feature to achieve dimensionality reduction and alignment.

[0084] The first candidate resistance feature is input into the seventh convolutional layer Conv7 to perform a convolution operation to obtain a second candidate resistance feature.

[0085] The second candidate resistance feature is input into the eighth convolutional layer Conv8 to perform a convolution operation to obtain a third candidate resistance feature.

[0086] The first candidate resistance feature and the third candidate resistance feature are concatenated to form a fourth candidate resistance feature.

[0087] The fourth candidate posture feature is input into the ninth convolutional layer Conv9 to perform a convolution operation to obtain a target resistance feature.

[0088] Step 107: Input the target posture feature, the first target vibration feature, and the target resistance feature into a decoder and decode them into a target multimodal feature.

[0089] like Figure 3 As shown, the target posture feature, the first target vibration feature and the target resistance feature are input into a decoder and decoded into a target multimodal feature.

[0090] In one embodiment of the present invention, Figure 3 As shown, the decoder includes a modulator and a fusion module (FM). Then, in this embodiment, step 107 may include the following steps:

[0091] Step 1071: In the adjuster, adjust the first target vibration feature according to the target posture feature to obtain the second target vibration feature.

[0092] In practical applications, such as Figure 3 As shown, there is a nonlinear relationship between the ship's drift and the vibration of the supporting structure of the maritime antenna. Therefore, the target attitude feature and the first target vibration feature are input into the modulator, and the first target vibration feature is modulated according to the target attitude feature to filter out the interference of the ship's drift and obtain a relatively pure second target vibration feature.

[0093] In one design, Figure 3As shown, the modulator Modulator includes a tenth convolutional layer Conv10. Then, in the present design, in the tenth convolutional layer Conv10, a convolution operation is performed on the first target vibration feature with the target posture feature as a convolution kernel to obtain a second target vibration feature.

[0094] Step 1072: In the fusion module, the second target vibration feature and the target resistance feature are fused into a target multimodal feature.

[0095] In this embodiment, if Figure 3 As shown, the second target vibration feature and the target resistance feature are input into the fusion module FM, and the second target vibration feature and the target resistance feature are interacted and fused into a target multimodal feature.

[0096] In one design, Figure 3 As shown, the fusion module FM includes a first convolution module ConvModule1, a second convolution module ConvModule2, a neural network Transformer, a first atrous spatial pyramid pooling module ASPP1 and a second atrous spatial pyramid pooling module ASPP2.

[0097] Among them, the first convolution module ConvModule1 and the second convolution module ConvModule2 each include a convolutional layer (Convolutional Layer) and a pooling layer Pooling in sequence. The convolutional layer provides a convolution operation, and the pooling layer provides a pooling operation, such as a maximum pooling operation or an average pooling operation.

[0098] In this design, the second target vibration feature is input into the first convolution module ConvModule1 to perform convolution operation and pooling operation in sequence to extract the first vibration modal feature.

[0099] The first vibration mode feature is input into the neural network Transformer to extract the second vibration mode feature.

[0100] The target resistance feature is input into the second convolution module ConvModule2 to extract the first resistance modal feature.

[0101] The first resistance modal feature is input into the neural network Transformer to extract the second resistance modal feature.

[0102] The first vibration modal feature and the second resistance modal feature are concatenated to form a first candidate multi-modal feature.

[0103] The first resistance modal feature and the second vibration modal feature are concatenated to form a second candidate multi-modal feature.

[0104] The first candidate multimodal feature is input into the first atrous spatial pyramid pooling module ASPP1 to extract the third candidate multimodal feature.

[0105] The second candidate multimodal feature is input into the second atrous spatial pyramid pooling module ASPP2 to extract the fourth candidate multimodal feature.

[0106] The third candidate multimodal feature and the fourth candidate multimodal feature are concatenated into the target multimodal feature.

[0107] In practical applications, vibration and resistance have high-density fluctuations. In this scenario, this embodiment relies on the attention mechanism of Transformer and the multi-scale receptive field of the Atrous Spatial Pyramid Pooling (ASPP) module to capture global context features, effectively improving the quality of features fused under multi-modality.

[0108] Step 108: Input the target multimodal features into the head structure to classify the strength of the support structure of the maritime antenna.

[0109] In this embodiment, if Figure 3 As shown, the target multimodal features are input into the head structure Head, the head structure Head performs a multi-classification task, and outputs the strength of the supporting structure of the maritime antenna.

[0110] In one design, Figure 3 As shown, the head structure Head includes a fourth fully connected layer FC4, which inputs the target multimodal features into the fourth fully connected layer FC4 and maps them into structural strength features. The structural strength features are activated using activation functions such as sigmoid to obtain the probabilities under various intensities, and the intensity with the highest probability is output.

[0111] The structural classification network of this embodiment has a simple structure, low computational complexity, and low resource consumption. It is suitable for edge computing equipment deployed in ships. During the navigation process of the ship, it can still detect the strength of the supporting structure of the maritime antenna in real time in a weak network environment.

[0112] Step 109: Perform an alarm operation on the supporting structure of the maritime antenna according to the strength.

[0113] In this embodiment, the strength of the supporting structure of the maritime antenna can be statistically analyzed. When the supporting structure of the maritime antenna is at risk of damage, corrosion, etc., an alarm operation is performed on the supporting structure of the maritime antenna. At this time, a prompt message is sent to the account of relevant users (such as the owner or lessee of the ship, etc.), prompting them to pay attention to the supporting structure of the maritime antenna, contact the manufacturer of the maritime antenna in time, and arrange for professionals to inspect and repair the supporting structure of the maritime antenna.

[0114] In addition, relevant users (such as ship owners or lessees, etc.) can also use the application provided by the manufacturer of the maritime antenna to actively detect the strength of the supporting structure of the maritime antenna, contact the manufacturer of the maritime antenna as needed, and arrange for professionals to inspect and repair the supporting structure of the maritime antenna.

[0115] In a statistical analysis method, the strength of the supporting structure of the maritime antenna can be continuously detected, and multiple intensities can be smoothed to obtain an intensity sequence.

[0116] A sliding window is added to the intensity sequence. Each time the window is slid, the intensity in the window is averaged to obtain the average intensity.

[0117] If the average intensity is less than or equal to the preset intensity threshold, it means that the supporting structure of the maritime antenna is at risk of damage, corrosion, etc., and an alarm operation is performed on the supporting structure of the maritime antenna using instant messaging, email, text message, etc.

[0118] In this embodiment, when the ship is sailing on the sea, the inertial measurement unit, the vibration sensor and the resistance sensor are started respectively; the inertial measurement unit is deployed on the hull of the ship, and the vibration sensor and the resistance sensor are deployed on the support structure of the maritime antenna of the ship; the attitude signal collected by the inertial measurement unit, the vibration signal collected by the vibration sensor and the resistance signal collected by the resistance sensor are read respectively; the structural classification network is determined; the structural classification network includes a first encoder, a second encoder, a third encoder, a decoder and a head structure; the attitude signal is input into the first encoder to extract the target attitude feature; the vibration signal is input into the second encoder to extract the first target vibration feature; the resistance signal is input into the third encoder to extract the target resistance feature; the target attitude feature, the first target vibration feature and the target resistance feature are input into the decoder to be decoded into the target multimodal feature; the target multimodal feature is input into the head structure to classify the strength of the support structure of the maritime antenna; and an alarm operation is performed on the support structure of the maritime antenna according to the strength. This embodiment explores the characteristics of the support structure of the maritime antenna in terms of vibration, resistance, etc. to evaluate the strength of the support structure of the maritime antenna, thereby performing an alarm operation, realizing self-inspection of the support structure of the maritime antenna, improving the accuracy of inspection and maintenance of the support structure of the maritime antenna in the ship, reducing the risk of damage to the maritime antenna due to damage to the support structure of the maritime antenna during the navigation of the ship, and ensuring the safety of the navigation of the ship.

[0119] Embodiment 2

[0120] See also Figure 4 , shows a schematic diagram of the structure of a device for detecting the strength of a support structure of a navigation antenna provided by Embodiment 2 of the present invention. Figure 4 As shown, the device comprises:

[0121] The sensor start module 401 is used to start the inertial measurement unit, the vibration sensor and the resistance sensor respectively when the ship is sailing on the sea; the inertial measurement unit is deployed on the hull of the ship, and the vibration sensor and the resistance sensor are both deployed on the support structure of the maritime antenna of the ship;

[0122] A signal reading module 402 is used to read the attitude signal collected by the inertial measurement unit, the vibration signal collected by the vibration sensor, and the resistance signal collected by the resistance sensor respectively;

[0123] A structure classification network determination module 403 is used to determine a structure classification network; the structure classification network includes a first encoder, a second encoder, a third encoder, a decoder and a head structure;

[0124] A target posture feature extraction module 404, used for inputting the posture signal into the first encoder to extract the target posture feature;

[0125] A target vibration feature extraction module 405, configured to input the vibration signal into the second encoder to extract a first target vibration feature;

[0126] A target resistance feature extraction module 406, configured to input the resistance signal into the third encoder to extract a target resistance feature;

[0127] A target multimodal feature decoding module 407, configured to input the target posture feature, the first target vibration feature and the target resistance feature into the decoder and decode them into a target multimodal feature;

[0128] A strength classification module 408, for inputting the target multimodal features into the head structure to classify the strength of the support structure of the maritime antenna;

[0129] The warning operation execution module 409 is used to execute a warning operation on the supporting structure of the maritime antenna according to the strength.

[0130] In one embodiment of the present invention, the first encoder includes a first convolutional layer, a second convolutional layer, a third convolutional layer and a first fully connected layer, and the target posture feature extraction module 404 is further used for:

[0131] Multiplying the transpose of the attitude signal by the attitude signal to obtain an attitude matrix;

[0132] Inputting the posture matrix into the first fully connected layer and mapping it into a first candidate posture feature;

[0133] Inputting the first candidate posture feature into the first convolution layer to perform a convolution operation to obtain a second candidate posture feature;

[0134] Inputting the second candidate posture feature into the second convolutional layer to perform a convolution operation to obtain a third candidate posture feature;

[0135] splicing the first candidate posture feature and the third candidate posture feature into a fourth candidate posture feature;

[0136] The fourth candidate posture feature is input into the third convolution layer to perform a convolution operation to obtain a target posture feature.

[0137] In one embodiment of the present invention, the second encoder includes a fourth convolutional layer, a fifth convolutional layer, a sixth convolutional layer and a second fully connected layer, and the target vibration feature extraction module 405 is further used to:

[0138] Performing time-frequency conversion on the vibration signal to obtain a vibration time-frequency graph;

[0139] Inputting the vibration time-frequency graph into the second fully connected layer and mapping it into a first candidate vibration feature;

[0140] Inputting the first candidate vibration feature into the fourth convolutional layer to perform a convolution operation to obtain a second candidate vibration feature;

[0141] Inputting the second candidate vibration feature into the fifth convolutional layer to perform a convolution operation to obtain a third candidate vibration feature;

[0142] splicing the first candidate vibration feature and the third candidate vibration feature into a fourth candidate vibration feature;

[0143] The fourth candidate vibration feature is input into the sixth convolutional layer to perform a convolution operation to obtain a target vibration feature.

[0144] In one embodiment of the present invention, the third encoder includes a seventh convolutional layer, an eighth convolutional layer, a ninth convolutional layer and a third fully connected layer, and the target resistance feature extraction module 406 is further used to:

[0145] multiplying the resistance signal by the transpose of the resistance signal to obtain a resistance matrix;

[0146] Inputting the resistance matrix into the third fully connected layer and mapping it into a first candidate resistance feature;

[0147] Inputting the first candidate resistance feature into the seventh convolutional layer to perform a convolution operation to obtain a second candidate resistance feature;

[0148] Inputting the second candidate resistance feature into the eighth convolutional layer to perform a convolution operation to obtain a third candidate resistance feature;

[0149] splicing the first candidate resistance feature and the third candidate resistance feature into a fourth candidate resistance feature;

[0150] The fourth candidate posture feature is input into the ninth convolution layer to perform a convolution operation to obtain a target resistance feature.

[0151] In one embodiment of the present invention, the decoder includes a moderator and a fusion module, and the target multimodal feature decoding module 407 includes:

[0152] a signal conditioning module, used for, in the conditioner, conditioning the first target vibration characteristic according to the target posture characteristic to obtain a second target vibration characteristic;

[0153] A signal fusion module is used to fuse the second target vibration feature and the target resistance feature into a target multimodal feature in the fusion module.

[0154] In one embodiment of the present invention, the moderator comprises a tenth convolutional layer; and the signal conditioning module is further configured to:

[0155] In the tenth convolutional layer, a convolution operation is performed on the first target vibration feature using the target posture feature as a convolution kernel to obtain a second target vibration feature.

[0156] In one embodiment of the present invention, the fusion module includes a first convolution module, a second convolution module, a neural network Transformer, a first atrous spatial pyramid pooling module, and a second atrous spatial pyramid pooling module; the first convolution module and the second convolution module each include a convolution layer and a pooling layer in sequence;

[0157] The signal fusion module is also used for:

[0158] Inputting the second target vibration feature into the first convolution module to extract the first vibration modal feature;

[0159] Inputting the first vibration modal feature into the neural network Transformer to extract the second vibration modal feature;

[0160] Inputting the target resistance feature into the second convolution module to extract the first resistance modal feature;

[0161] Inputting the first resistance modal feature into the neural network Transformer to extract the second resistance modal feature;

[0162] splicing the first vibration modal feature and the second resistance modal feature into a first candidate multi-modal feature;

[0163] splicing the first resistance modal feature and the second vibration modal feature into a second candidate multi-modal feature;

[0164] Inputting the first candidate multimodal feature into the first atrous spatial pyramid pooling module to extract a third candidate multimodal feature;

[0165] Inputting the second candidate multimodal feature into the second atrous spatial pyramid pooling module to extract a fourth candidate multimodal feature;

[0166] The third candidate multimodal feature and the fourth candidate multimodal feature are concatenated into a target multimodal feature.

[0167] In one embodiment of the present invention, the alarm operation execution module 409 is further used to:

[0168] smoothing a plurality of the intensities to obtain an intensity sequence;

[0169] adding a slidable window in the intensity sequence;

[0170] Calculating an average value of the intensities in the window to obtain an average intensity;

[0171] If the average intensity is less than or equal to a preset intensity threshold, an alarm operation is performed on the supporting structure of the maritime antenna.

[0172] The device for detecting the strength of a marine antenna support structure provided by an embodiment of the present invention can execute the method for detecting the strength of a marine antenna support structure provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method for detecting the strength of a marine antenna support structure.

[0173] Embodiment 3

[0174] See also Figure 5 , shows a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, blade servers, mainframe computers, and other suitable computers. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0175] like Figure 5As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0176] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0177] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a method for detecting the strength of a support structure of a marine antenna.

[0178] In some embodiments, the method for detecting the strength of the support structure of the marine antenna may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the method for detecting the strength of the support structure of the marine antenna described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the method for detecting the strength of the support structure of the marine antenna by any other appropriate means (e.g., by means of firmware).

[0179] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0180] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0181] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0182] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0183] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0184] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0185] Embodiment 4

[0186] An embodiment of the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the method for detecting the strength of a navigation antenna support structure provided by any embodiment of the present invention is implemented.

[0187] In the process of implementation, the computer program product can be written in one or more programming languages ​​or a combination thereof to perform the computer program code of the present invention, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0188] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0189] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting the strength of a marine antenna support structure, characterized in that: include: When the ship is sailing on the sea, the inertial measurement unit, the vibration sensor and the resistance sensor are activated respectively; The inertial measurement unit is deployed on the hull of the ship, and the vibration sensor and the resistance sensor are both deployed on a supporting structure of a maritime antenna of the ship; Respectively reading the attitude signal collected by the inertial measurement unit, the vibration signal collected by the vibration sensor, and the resistance signal collected by the resistance sensor; Determine a structural classification network; the structural classification network includes a first encoder, a second encoder, a third encoder, a decoder and a head structure; Inputting the posture signal into the first encoder to extract the target posture feature; Inputting the vibration signal into the second encoder to extract a first target vibration feature; Inputting the resistance signal into the third encoder to extract the target resistance feature; Inputting the target posture feature, the first target vibration feature and the target resistance feature into the decoder to be decoded into a target multimodal feature; Inputting the target multimodal features into the head structure to classify the strength of the support structure of the maritime antenna; An alarm operation is performed on the supporting structure of the maritime antenna according to the strength.

2. The method according to claim 1, characterized in that The first encoder includes a first convolution layer, a second convolution layer, a third convolution layer and a first fully connected layer, and the step of inputting the posture signal into the first encoder to extract the target posture feature includes: Multiplying the transpose of the attitude signal by the attitude signal to obtain an attitude matrix; Inputting the posture matrix into the first fully connected layer and mapping it into a first candidate posture feature; Inputting the first candidate posture feature into the first convolution layer to perform a convolution operation to obtain a second candidate posture feature; Inputting the second candidate posture feature into the second convolutional layer to perform a convolution operation to obtain a third candidate posture feature; splicing the first candidate posture feature and the third candidate posture feature into a fourth candidate posture feature; The fourth candidate posture feature is input into the third convolution layer to perform a convolution operation to obtain a target posture feature.

3. The method according to claim 2, characterized in that The second encoder includes a fourth convolution layer, a fifth convolution layer, a sixth convolution layer and a second fully connected layer, and the step of inputting the vibration signal into the second encoder to extract the first target vibration feature includes: Performing time-frequency conversion on the vibration signal to obtain a vibration time-frequency graph; Inputting the vibration time-frequency graph into the second fully connected layer and mapping it into a first candidate vibration feature; Inputting the first candidate vibration feature into the fourth convolutional layer to perform a convolution operation to obtain a second candidate vibration feature; Inputting the second candidate vibration feature into the fifth convolutional layer to perform a convolution operation to obtain a third candidate vibration feature; splicing the first candidate vibration feature and the third candidate vibration feature into a fourth candidate vibration feature; The fourth candidate vibration feature is input into the sixth convolutional layer to perform a convolution operation to obtain a target vibration feature.

4. The method according to claim 3, characterized in that The third encoder includes a seventh convolutional layer, an eighth convolutional layer, a ninth convolutional layer and a third fully connected layer, and the step of inputting the resistance signal into the third encoder to extract the target resistance feature includes: multiplying the resistance signal by the transpose of the resistance signal to obtain a resistance matrix; Inputting the resistance matrix into the third fully connected layer and mapping it into a first candidate resistance feature; Inputting the first candidate resistance feature into the seventh convolutional layer to perform a convolution operation to obtain a second candidate resistance feature; Inputting the second candidate resistance feature into the eighth convolutional layer to perform a convolution operation to obtain a third candidate resistance feature; splicing the first candidate resistance feature and the third candidate resistance feature into a fourth candidate resistance feature; The fourth candidate posture feature is input into the ninth convolution layer to perform a convolution operation to obtain a target resistance feature.

5. The method according to claim 4, characterized in that The decoder includes a moderator and a fusion module, and the target posture feature, the first target vibration feature and the target resistance feature are input into the decoder and decoded into a target multimodal feature, including: In the adjuster, the first target vibration feature is adjusted according to the target posture feature to obtain a second target vibration feature; In the fusion module, the second target vibration feature and the target resistance feature are fused into a target multimodal feature.

6. The method according to claim 5, characterized in that The moderator includes a tenth convolutional layer; in the moderator, the first target vibration feature is mediated according to the target posture feature to obtain a second target vibration feature, including: In the tenth convolutional layer, a convolution operation is performed on the first target vibration feature using the target posture feature as a convolution kernel to obtain a second target vibration feature.

7. The method according to claim 6, characterized in that The fusion module includes a first convolution module, a second convolution module, a neural network Transformer, a first atrous spatial pyramid pooling module, and a second atrous spatial pyramid pooling module; the first convolution module and the second convolution module each include a convolution layer and a pooling layer in sequence; In the fusion module, the second target vibration feature and the target resistance feature are fused into a target multimodal feature, including: Inputting the second target vibration feature into the first convolution module to extract the first vibration modal feature; Inputting the first vibration modal feature into the neural network Transformer to extract the second vibration modal feature; Inputting the target resistance feature into the second convolution module to extract the first resistance modal feature; Inputting the first resistance modal feature into the neural network Transformer to extract the second resistance modal feature; splicing the first vibration modal feature and the second resistance modal feature into a first candidate multi-modal feature; splicing the first resistance modal feature and the second vibration modal feature into a second candidate multi-modal feature; Inputting the first candidate multimodal feature into the first atrous spatial pyramid pooling module to extract a third candidate multimodal feature; Inputting the second candidate multimodal feature into the second atrous spatial pyramid pooling module to extract a fourth candidate multimodal feature; The third candidate multimodal feature and the fourth candidate multimodal feature are concatenated into a target multimodal feature.

8. The method according to any one of claims 1 to 7, characterized in that The performing an alarm operation on the supporting structure of the maritime antenna according to the strength includes: smoothing a plurality of the intensities to obtain an intensity sequence; adding a slidable window in the intensity sequence; Calculating an average value of the intensities in the window to obtain an average intensity; If the average intensity is less than or equal to a preset intensity threshold, an alarm operation is performed on the supporting structure of the maritime antenna.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the method for detecting the strength of a navigation antenna support structure according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for detecting the strength of a navigation antenna support structure according to any one of claims 1 to 8 is implemented.

Images