Navigation radome structure corrosion prevention method and device and storage medium

Abstract

The invention discloses a marine antenna housing structure corrosion prevention method and device, and a storage medium. The method comprises the steps of querying a first time node for performing corrosion prevention maintenance on an antenna housing structure of a maritime antenna in a ship last time; track point data generated after the first time node is collected for the ship; each piece of track point data comprises coordinate data and date data; determining a risk detection network; inputting the coordinate data into the first backbone structure to extract target coordinate features; inputting the date data into a second backbone structure to extract target date features; inputting the target coordinate features and the target date features into a feature fusion module to be fused into target spatial-temporal features; inputting the target spatial-temporal characteristics into the head structure, and classifying to obtain a corrosion risk level; and according to the corrosion risk level, prompting the antenna housing structure to carry out anti-corrosion maintenance. According to the embodiment of the invention, the anti-corrosion overhaul accuracy of the radome structure of the maritime antenna in the ship is improved, and the sailing safety of the ship 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 preventing corrosion of a marine radome structure. Background Art

[0002] Maritime antennas are installed on ships, which are one of the components of ship communication and navigation systems, and realize communication, navigation and other functions in maritime activities. Some types of maritime antennas are equipped with radome structures to protect the maritime antennas. When ships sail on the ocean, environmental factors such as seawater and salt spray will cause certain corrosion to the radome structure.

[0003] At present, some parts of the radome structure are made of corrosion-resistant materials, such as glass fiber reinforced plastic (GFRP), polytetrafluoroethylene (PTFE), etc., or anti-corrosion coatings, such as fluorocarbon coatings, ceramic layers, etc., are added to the surface of the radome structure.

[0004] Whether it is corrosion-resistant materials or anti-corrosion coatings, their coverage is limited, and some parts of the radome structure have a higher risk of corrosion. Therefore, the radome structure should be inspected for corrosion protection to ensure the safety of the radome structure, thereby ensuring the safety of the maritime antenna.

[0005] However, small vessels such as yachts and fishing boats lack professionals, and their owners usually arrange for professionals to perform anti-corrosion inspections on the antenna cover structure at intervals. The length of time varies, resulting in low inspection accuracy. There is a risk of damage to the maritime antenna due to corrosion and damage to the antenna cover structure during navigation. Summary of the invention

[0006] In view of this, the present invention provides a method, device and storage medium for preventing corrosion of a marine radome structure, so as to improve the accuracy of anti-corrosion maintenance of the radome structure on a ship.

[0007] A first aspect of the present invention provides a method for preventing corrosion of a marine radome structure, comprising:

[0008] Query the first time node of the last anti-corrosion maintenance of the radome structure of the maritime antenna in the ship;

[0009] During the navigation of the ship, track point data generated after the first time node is collected for the ship; each track point data has coordinate data and date data;

[0010] Determine a risk detection network; the risk detection network has a first backbone structure, a second backbone structure, a feature fusion module and a head structure;

[0011] Inputting the coordinate data into the first backbone structure to extract target coordinate features;

[0012] Inputting the date data into the second backbone structure to extract target date features;

[0013] Inputting the target coordinate feature and the target date feature into the feature fusion module and fusing them into target spatiotemporal features;

[0014] Inputting the target spatiotemporal characteristics into the head structure for classification to obtain a corrosion risk level;

[0015] According to the risk level of corrosion, the antenna cover structure is prompted to perform anti-corrosion maintenance.

[0016] A second aspect of the present invention provides a marine radome structure corrosion prevention device, comprising:

[0017] A time node query module is used to query the first time node of the last anti-corrosion maintenance of the radome structure of the maritime antenna in the ship;

[0018] A track point data collection module, used for collecting track point data generated after the first time node for the ship during the navigation of the ship; each track point data has coordinate data and date data;

[0019] A risk detection network determination module, used to determine a risk detection network; the risk detection network has a first backbone structure, a second backbone structure, a feature fusion module and a head structure;

[0020] A target coordinate feature extraction module, used for inputting the coordinate data into the first backbone structure to extract target coordinate features;

[0021] A target date feature extraction module, used for inputting the date data into the second backbone structure to extract target date features;

[0022] A target spatiotemporal feature fusion module, used for inputting the target coordinate feature and the target date feature into the feature fusion module and fusing them into target spatiotemporal features;

[0023] A risk level classification module, used for inputting the target spatiotemporal characteristics into the head structure to classify and obtain the risk level of corrosion;

[0024] The anti-corrosion maintenance prompt module is used to prompt the antenna cover structure to perform anti-corrosion maintenance according to the risk level of corrosion.

[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 preventing corrosion of the marine radome structure 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, which, when executed by a processor, implements the method for preventing corrosion of the marine antenna cover structure as described in the first aspect above.

[0030] A fifth aspect of the present invention provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the method for preventing corrosion of the marine antenna cover structure as described in the first aspect above is implemented.

[0031] In this embodiment, the first time node of the last anti-corrosion maintenance of the radome structure of the maritime antenna in the ship is queried; during the navigation of the ship, the track point data generated after the first time node is collected for the ship; each track point data has coordinate data and date data; a risk detection network is determined; the risk detection network has a first backbone structure, a second backbone structure, a feature fusion module and a head structure; the coordinate data is input into the first backbone structure to extract the target coordinate feature; the date data is input into the second backbone structure to extract the target date feature; the target coordinate feature and the target date feature are input into the feature fusion module to be fused into the target spatiotemporal feature; the target spatiotemporal feature is input into the head structure to classify the corrosion risk level; the radome structure is prompted to perform anti-corrosion maintenance according to the corrosion risk level. This embodiment mines the spatiotemporal dependence of coordinates and dates to evaluate the corrosion process of the radome structure of the maritime antenna in the ship, so as to select the appropriate time to prompt the anti-corrosion maintenance, improve the accuracy of the anti-corrosion maintenance of the radome structure of the maritime antenna in the ship, reduce the risk of the maritime antenna being damaged due to corrosion and damage of the radome structure during the navigation of the ship, and ensure the safety of the ship's navigation.

[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 preventing corrosion of a marine radome structure provided in Embodiment 1 of the present invention.

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

[0036] Figure 3 It is a schematic diagram of a radome structure provided in Embodiment 1 of the present invention.

[0037] Figure 4 It is a schematic diagram of a risk detection network provided in Example 1 of the present invention.

[0038] Figure 5 It is a structural schematic diagram of a marine radome structure corrosion prevention device provided in Embodiment 2 of the present invention.

[0039] Figure 6 It is a structural schematic diagram of an electronic device provided in Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0040] 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.

[0041] 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.

[0042] Embodiment 1

[0043] See also Figure 1 , shows a flow chart of a method for preventing corrosion of a marine radome structure provided by a first embodiment of the present invention, the method can be executed by a device for preventing corrosion of a marine radome structure, the device for preventing corrosion of a marine radome structure can be implemented in the form of hardware and / or software, and the device for preventing corrosion of a marine radome structure can be configured in an electronic device. Figure 1 As shown, the method includes:

[0044] Step 101, query the first time node of the last anti-corrosion maintenance of the radome structure of the maritime antenna in the ship.

[0045] like Figure 2 and Figure 3 As shown, one or more maritime antennas 201 are installed on the ship 200 . The maritime antenna 201 is provided with a radome structure 202 . The manufacturer of the maritime antenna provides various maintenance services for the maritime antenna 201 on the ship 200 .

[0046] In the server of the manufacturer of the maritime antenna, a database is used to record log information of various inspections and repairs of the maritime antenna in each ship.

[0047] In this embodiment, the time when the corresponding ship last (ie, most recently) performed anti-corrosion maintenance on the radome structure of its maritime antenna can be queried in the log information of the database according to the ship ID, and recorded as the first time node.

[0048] Step 102: During the navigation of the ship, track point data generated after the first time node is collected from the ship.

[0049] A satellite positioning chip is configured in the maritime antenna of the ship, and a timer is configured in the satellite positioning chip. The satellite positioning chip can be used to record the coordinate data of the ship (such as longitude and latitude), and the timer can be used to record date data (such as converting the combination of month and date into a number), thereby generating track point data, that is, each track point data has coordinate data and date data.

[0050] The corrosion of the radome structure of the maritime antenna in the ship is mainly due to the long-term combined effects of factors such as seawater and salt spray, which are mainly seasonal changes in different sea areas (affecting the temperature of seawater, local climate, etc.), and are less affected by short-term weather changes (such as typhoons). The seasonal changes in different sea areas are strongly correlated with coordinate data and date data. Therefore, in this embodiment, the seasonal changes in different sea areas are modeled and characterized by coordinate data and date data, so as to realize the spatiotemporal expression of the seasonality of different sea areas.

[0051] Considering that the radome structure of the maritime antenna in the ship is in a better condition after the anti-corrosion inspection, the trajectory point data generated after the first time node can be intercepted from all the trajectory point data of the ship, and the corrosion risk of this section of trajectory point data can be analyzed independently.

[0052] Step 103: Determine the risk detection network.

[0053] In this embodiment, a risk detection network can be constructed and trained in advance based on deep learning, and the risk detection network can be used to detect corrosion risks.

[0054] Considering that the radome structure of the maritime antenna in the ship itself has a certain anti-corrosion ability, the corrosion progress of the radome structure of the maritime antenna in the ship is slow. Therefore, when the ship starts the sea voyage for the mth time (m is a positive integer), a sample structure sensitive to corrosion is added to the surface of the radome structure of the maritime antenna in the ship, and interference with the maritime antenna is avoided, so as to simulate the corrosion degree of the radome structure of the maritime antenna in the ship using the sample structure. When the ship ends the sea voyage for the nth time (n is a positive integer), the corrosion level is added to the radome structure of the maritime antenna in the ship according to the corrosion degree of the sample structure.

[0055] In order to fully simulate the corrosion degree of the radome structure of the maritime antenna in the ship, when the distance and time of a voyage are relatively long, the data of one voyage can be used as a sample. At this time, m is the same as n, that is, there can be no other voyages between the mth voyage and the nth voyage. When the distance and time of a voyage are relatively short, the data of one voyage can be used as a sample. At this time, m is different from n, that is, there are one or more other voyages between the mth voyage and the nth voyage.

[0056] The trajectory point data between the mth voyage and the nth voyage are used as samples, the corrosion level is used as the label, and the cross entropy is selected as the loss function to conduct supervised training on the risk detection network.

[0057] In practical applications, such as Figure 4 As shown, the risk detection network has a first backbone structure Backbone1, a second backbone structure Backbone2, a feature fusion module (Feature Fusion Module, FFM) and a head structure Head.

[0058] Among them, the first backbone structure Backbone1 is used to extract features from coordinate data, the second backbone structure Backbone2 is used to extract features from date data, the feature fusion module FFM is used to fuse the features of coordinate data with the features of date data, and the head structure Head is used to perform multi-classification tasks based on the fused features and output the corrosion level.

[0059] Step 104: Input the coordinate data into the first backbone structure to extract target coordinate features.

[0060] In this embodiment, if Figure 4 As shown, the first backbone network Backbone1 is one of the main parts of the risk detection network. It is usually composed of multiple convolutional layers, pooling layers and other basic components with different depths and complexities. It extracts and abstracts the input coordinate data and converts the original coordinate data into a more representative feature representation, which is recorded as the target coordinate feature.

[0061] In one design, Figure 4 As shown, the first backbone structure Backbone1 includes a first long short-term memory network LSTM1, a first convolutional layer Conv1, a second convolutional layer Conv2, a third convolutional layer Conv3 and a first fully connected layer FC1.

[0062] Then, in this design, the coordinate data is input into the first long short-term memory network LSTM1 to extract the first coordinate sequence features. Since the sequence of coordinate data is long and of varying lengths, the first long short-term memory network LSTM1 can reduce the dimension of the coordinate data and align the length.

[0063] Reshape the first coordinate sequence feature into a two-dimensional first coordinate feature map.

[0064] The first coordinate feature map is input into the first convolutional layer Conv1 to extract the second coordinate feature map.

[0065] The hyperbolic tangent function tanh is used to activate the second coordinate feature map.

[0066] The first coordinate feature map is input into the second convolutional layer Conv2 to extract the third coordinate feature map.

[0067] The sigmoid function is used to activate the third coordinate feature map.

[0068] The activated second coordinate feature map is added to the activated third coordinate feature map to obtain a fourth coordinate feature map.

[0069] Reshape the fourth coordinate feature map into a one-dimensional second coordinate sequence feature.

[0070] The second coordinate sequence feature is input into the first fully connected layer FC1 and mapped into the target coordinate feature.

[0071] Furthermore, the tanh function helps to guide the capture of positive and negative symmetric relationships and bidirectional changes in features, and the sigmoid function can provide a suitable output form, making the risk detection network more in line with task requirements. Since they have different functional characteristics and data transformation methods, their combined use can increase the complexity and diversity of the risk detection network, allowing the risk detection network to better generalize to new data and reduce overfitting of samples.

[0072] Step 105: Input date data into the second backbone structure to extract target date features.

[0073] In this embodiment, if Figure 4 As shown in the figure, the second backbone network Backbone2 is one of the main parts of the risk detection network. It is usually composed of multiple convolutional layers, pooling layers and other basic components with different depths and complexities. It extracts and abstracts the input date data and converts the original date data into a more representative feature representation, which is recorded as the target coordinate feature.

[0074] In one design, Figure 4 As shown, the second backbone structure Backbone2 includes a second long short-term memory network LSTM2, a first temporal convolutional network TCN1, a second temporal convolutional network TCN2 and a second fully connected layer FC2.

[0075] In this design, the date data is input into the second long short-term memory network LSTM2 to extract the first date sequence features. Since the sequence of date data is long and of varying lengths, the second long short-term memory network LSTM2 can preliminarily extract features from the coordinate data.

[0076] The first date sequence features are input into the first temporal convolutional network TCN1 to extract the second date sequence features.

[0077] The second date sequence features are input into the second temporal convolutional network TCN2 to extract the third date sequence features.

[0078] The first date sequence feature and the third date sequence feature are concatenated to form the fourth date sequence feature.

[0079] The fourth date sequence feature is input into the second fully connected layer FC2 and mapped into the target date feature.

[0080] Furthermore, the first temporal convolutional network TCN1 and the second temporal convolutional network TCN2 both belong to temporal convolutional networks (TCN). TCN can improve the multi-scale information extraction capability by stacking convolutional layers and increasing the receptive field of the convolution kernel, and effectively capture the local dependencies in the coordinate data.

[0081] Step 106: Input the target coordinate feature and the target date feature into a feature fusion module and fuse them into a target spatiotemporal feature.

[0082] In this embodiment, if Figure 4 As shown in the figure, the target coordinate features and the target date features are input into the feature fusion module FFM. The feature fusion module FFM is responsible for integrating the target coordinate features and the target date features at different levels, and fusing the two into the target spatiotemporal features, thereby improving the expression ability of the risk detection network and helping the risk detection network to better adapt to the classification task of corrosion grades.

[0083] In one design, Figure 4 As shown, the feature fusion module FFM includes a self-attention layer Self-Attention, a first attention module AM1 and a second attention module AM2, wherein the first attention module AM1 and the second attention module AM2 are modules that encapsulate the attention mechanism.

[0084] In this design, the first candidate spatiotemporal feature Feature is randomly generated.

[0085] The first candidate spatiotemporal feature Feature is input into the self-attention layer, and the second candidate spatiotemporal feature is generated under the self-attention mechanism.

[0086] The second candidate spatiotemporal feature, target coordinate feature and target date feature are input into the first attention module AM1. Under the attention mechanism, the three are fused into the third candidate spatiotemporal feature according to the attention of the target coordinate feature and the target date feature relative to the second candidate spatiotemporal feature.

[0087] Furthermore, the first attention module AM1 includes a first attention layer Attention1 and a second attention layer Attention2.

[0088] At this time, the second candidate spatiotemporal features and the target coordinate features are input into the first attention layer Attention1, with the second candidate spatiotemporal features as the Query (query matrix) in the attention mechanism, and the target coordinate features as the Key (key matrix) and Value (value matrix) in the attention mechanism, and the two are fused into the first basic spatiotemporal features under the attention mechanism.

[0089] The first basic spatiotemporal features and the target date features are input into the second attention layer Attention2, with the first basic spatiotemporal features as the Query (query matrix) in the attention mechanism, and the target date features as the Key (key matrix) and Value (value matrix) in the attention mechanism, and the two are fused into the third candidate spatiotemporal features under the attention mechanism.

[0090] The third candidate spatiotemporal feature, target coordinate feature and target date feature are input into the second attention module AM2. Under the attention mechanism, the three are fused into the target spatiotemporal feature according to the attention of the target coordinate feature and the target date feature relative to the third candidate spatiotemporal feature.

[0091] Furthermore, the second attention module includes a third attention layer Attention3 and a fourth attention layer Attention4.

[0092] At this time, the third candidate spatiotemporal features and the target coordinate features are input into the third attention layer Attention3, with the third candidate spatiotemporal features as the Query (query matrix) in the attention mechanism, and the target coordinate features as the Key (key matrix) and Value (value matrix) in the attention mechanism, and the two are fused into the second basic spatiotemporal features under the attention mechanism.

[0093] The second basic spatiotemporal features and the target date features are input into the fourth attention layer Attention4, with the second basic spatiotemporal features as the Query (query matrix) in the attention mechanism, and the target date features as the Key (key matrix) and Value (value matrix) in the attention mechanism, and the two are fused into the target spatiotemporal features under the attention mechanism.

[0094] There are complex spatial dependencies and dynamic changes in temporal patterns between different types of ships in different sailing routes. In this design, the first backbone structure Backbone1 and the second backbone structure Backbone2 can effectively extract local spatial dependencies and temporal dependencies, and use the attention mechanism to capture more implicit long-term spatiotemporal dependencies, thereby effectively improving the expression of features.

[0095] Step 107: Input the target spatiotemporal features into the head structure to classify and obtain the risk level of corrosion.

[0096] In this embodiment, if Figure 4 As shown, the target spatiotemporal features are input into the head structure Head, and the head structure Head performs a multi-classification operation and outputs the corrosion risk level Level.

[0097] In one design, Figure 4As shown, the head structure Head includes a third fully connected layer FC3, which maps the target spatiotemporal features in the third fully connected layer FC3 into corrosion risk features, activates the corrosion risk features using activation functions such as sigmoid, obtains the probabilities under various risk levels, and outputs the risk level Level with the highest probability.

[0098] Step 108: Perform anti-corrosion maintenance on the radome structure according to the corrosion risk level.

[0099] In this embodiment, a statistical analysis can be performed on the corrosion risk level. When a higher corrosion risk accumulates, the antenna cover structure is prompted to perform anti-corrosion maintenance. At this time, a prompt message is sent to the account of the relevant user (such as the owner or lessee of the ship, etc.), prompting the user to pay attention to the degree of corrosion of the antenna cover structure of the maritime antenna, contact the manufacturer of the maritime antenna in time, and arrange for professionals to perform anti-corrosion maintenance on the antenna cover structure.

[0100] In addition, relevant users (such as ship owners or lessees) can also use the application provided by the manufacturer of the maritime antenna to actively detect the degree of corrosion of the radome structure of the maritime antenna, contact the manufacturer of the maritime antenna as needed, and arrange for professionals to perform anti-corrosion inspections on the radome structure.

[0101] In one embodiment of the present invention, step 108 may include the following steps:

[0102] Step 1081, querying the ship for the second time node for the next anti-corrosion maintenance of the radome structure of the maritime antenna.

[0103] In this embodiment, the time when the corresponding ship will next perform anti-corrosion maintenance on the radome structure of its maritime antenna can be queried in the scheduling information of the database according to the ship ID, and recorded as the second time node.

[0104] Step 1082: Calculate the duration between the current time and the second time node.

[0105] In this embodiment, the current time is read in the operating system or the network, and the second time node is subtracted from the current time to obtain the duration of the interval between the two.

[0106] Step 1083: If the corrosion risk level is higher than or equal to the preset target level, a prompt frequency is generated based on the duration, risk level and duration.

[0107] As the ship continues to sail, the corrosion process continues to deepen, and the corrosion risk level increases. If the corrosion risk level is greater than or equal to the preset target level, it means that the degree of corrosion is high. At this time, the prompt frequency can be generated based on the duration, risk level and duration. Among them, the prompt frequency is negatively correlated with the duration, and the prompt frequency is positively correlated with the risk level.

[0108] In a specific implementation, the first benchmark frequency and the second benchmark frequency configured for the duration can be queried; wherein the first benchmark frequency is the highest prompt frequency under the duration, and the second benchmark frequency is the lowest prompt frequency under the duration. Therefore, the first benchmark frequency is higher than the second benchmark frequency, and both the first benchmark frequency and the second benchmark frequency are negatively correlated with the duration.

[0109] The difference between the first reference frequency and the second reference frequency is calculated to obtain a frequency deviation value.

[0110] A linear or nonlinear function is used to map the risk level to a sliding scale, wherein the sliding scale is positively correlated with the risk level.

[0111] Exemplarily, the risk level is added to a preset first adjustment coefficient to obtain a first adjustment level, a ratio between the first adjustment level and a preset second adjustment coefficient is calculated to obtain a second adjustment level, and the second adjustment level is squared to obtain a floating ratio.

[0112] The product of the frequency deviation value and the floating ratio is calculated as the floating deviation value.

[0113] The sum of the second reference frequency and the floating deviation value is rounded to an integer to obtain the prompt frequency.

[0114] Then, the prompt frequency can be expressed as:

[0115] Pbest=Round(Pmin+sqrt((Risk+a) / b)×(Pmax-Pmin));

[0116] Among them, Pbest is the prompt frequency, Pmax is the first benchmark frequency, Pmin is the second benchmark frequency, Risk is the risk level of corrosion, a is the first adjustment coefficient, b is the second adjustment coefficient, Round is the integer function, and sqrt is the square root function.

[0117] Step 1084, perform anti-corrosion maintenance on the antenna cover structure according to the prompt frequency.

[0118] In this embodiment, the anti-corrosion maintenance of the antenna cover structure can be prompted by instant messaging, email, text messages, etc. according to the reminder frequency.

[0119] In this embodiment, the first time node of the last anti-corrosion maintenance of the radome structure of the maritime antenna in the ship is queried; during the navigation of the ship, the track point data generated after the first time node is collected for the ship; each track point data has coordinate data and date data; a risk detection network is determined; the risk detection network has a first backbone structure, a second backbone structure, a feature fusion module and a head structure; the coordinate data is input into the first backbone structure to extract the target coordinate feature; the date data is input into the second backbone structure to extract the target date feature; the target coordinate feature and the target date feature are input into the feature fusion module to be fused into the target spatiotemporal feature; the target spatiotemporal feature is input into the head structure to classify the corrosion risk level; the radome structure is prompted to perform anti-corrosion maintenance according to the corrosion risk level. This embodiment mines the spatiotemporal dependence of coordinates and dates to evaluate the corrosion process of the radome structure of the maritime antenna in the ship, so as to select the appropriate time to prompt the anti-corrosion maintenance, improve the accuracy of the anti-corrosion maintenance of the radome structure of the maritime antenna in the ship, reduce the risk of the maritime antenna being damaged due to corrosion and damage of the radome structure during the navigation of the ship, and ensure the safety of the ship's navigation.

[0120] Embodiment 2

[0121] See also Figure 5 , shows a schematic diagram of the structure of a marine antenna cover structure corrosion prevention device provided by the second embodiment of the present invention. Figure 5 As shown, the device comprises:

[0122] A time node query module 501 is used to query the first time node of the last anti-corrosion maintenance of the radome structure of the maritime antenna in the ship;

[0123] The track point data collection module 502 is used to collect track point data generated after the first time node for the ship during the navigation of the ship; each track point data has coordinate data and date data;

[0124] The risk detection network determination module 503 is used to determine the risk detection network; the risk detection network has a first backbone structure, a second backbone structure, a feature fusion module and a head structure;

[0125] A target coordinate feature extraction module 504, used for inputting the coordinate data into the first backbone structure to extract target coordinate features;

[0126] A target date feature extraction module 505, for inputting the date data into the second backbone structure to extract target date features;

[0127] A target spatiotemporal feature fusion module 506 is used to input the target coordinate feature and the target date feature into the feature fusion module and fuse them into a target spatiotemporal feature;

[0128] A risk level classification module 507, used for inputting the target spatiotemporal features into the head structure to classify and obtain the risk level of corrosion;

[0129] The anti-corrosion maintenance prompt module 508 is used to prompt the radome structure to perform anti-corrosion maintenance according to the risk level of corrosion.

[0130] In one embodiment of the present invention, the first backbone structure includes a first long short-term memory network, a first convolutional layer, a second convolutional layer, a third convolutional layer and a first fully connected layer, and the target coordinate feature extraction module 504 is further used to:

[0131] Inputting the coordinate data into the first long short-term memory network to extract first coordinate sequence features;

[0132] Converting the first coordinate sequence feature into a two-dimensional first coordinate feature graph;

[0133] Inputting the first coordinate feature map into the first convolutional layer to extract a second coordinate feature map;

[0134] Activating the second coordinate feature map using a hyperbolic tangent function;

[0135] Inputting the first coordinate feature map into the second convolutional layer to extract a third coordinate feature map;

[0136] Activating the third coordinate feature map using a S-type function;

[0137] Adding the activated second coordinate feature map to the activated third coordinate feature map to obtain a fourth coordinate feature map;

[0138] Converting the fourth coordinate feature map into a one-dimensional second coordinate sequence feature;

[0139] The second coordinate sequence feature is input into the first fully connected layer and mapped into the target coordinate feature.

[0140] In one embodiment of the present invention, the second backbone structure includes a second long short-term memory network, a first temporal convolutional network, a second temporal convolutional network and a second fully connected layer, and the target date feature extraction module 505 is further used to:

[0141] Inputting the date data into the second long short-term memory network to extract the first date sequence features;

[0142] Input the first date sequence feature into the first temporal convolutional network to extract the second date sequence feature;

[0143] Input the second date series feature into the second temporal convolutional network to extract the third date series feature;

[0144] concatenating the first date series feature and the third date series feature into a fourth date series feature;

[0145] The fourth date series feature is input into the second fully connected layer and mapped into a target date feature.

[0146] In one embodiment of the present invention, the feature fusion module includes a self-attention layer, a first attention module and a second attention module, and the target spatiotemporal feature fusion module 506 is further used to:

[0147] Randomly generate the first candidate spatiotemporal feature;

[0148] Inputting the first candidate spatiotemporal feature into the self-attention layer, and generating a second candidate spatiotemporal feature under the self-attention mechanism;

[0149] Inputting the second candidate spatiotemporal feature, the target coordinate feature, and the target date feature into the first attention module, and fusing them into a third candidate spatiotemporal feature under the attention mechanism;

[0150] The third candidate spatiotemporal feature, the target coordinate feature and the target date feature are input into the second attention module and fused into the target spatiotemporal feature under the attention mechanism.

[0151] In one embodiment of the present invention, the first attention module includes a first attention layer and a second attention layer, and the second attention module includes a third attention layer and a fourth attention layer;

[0152] The target spatiotemporal feature fusion module 506 is also used for:

[0153] Inputting the second candidate spatiotemporal feature and the target coordinate feature into the first attention layer, and fusing them into the first basic spatiotemporal feature under the attention mechanism;

[0154] Input the first basic spatiotemporal feature and the target date feature into the second attention layer, and fuse them into the third candidate spatiotemporal feature under the attention mechanism;

[0155] The target spatiotemporal feature fusion module 506 is also used for:

[0156] Inputting the third candidate spatiotemporal feature and the target coordinate feature into the third attention layer, and fusing them into the second basic spatiotemporal feature under the attention mechanism;

[0157] The second basic spatiotemporal features and the target date features are input into the fourth attention layer and fused into the target spatiotemporal features under the attention mechanism.

[0158] In one embodiment of the present invention, the anti-corrosion maintenance prompt module 508 is also used to:

[0159] querying the ship for a second time point for the next anti-corrosion maintenance of the radome structure of the maritime antenna;

[0160] Calculate the duration between the current time and the second time node;

[0161] If the corrosion risk level is greater than or equal to the preset target level, a prompt frequency is generated according to the duration, the risk level and the duration;

[0162] According to the reminder frequency, the antenna cover structure is prompted to carry out anti-corrosion maintenance.

[0163] In one embodiment of the present invention, the anti-corrosion maintenance prompt module 508 is also used to:

[0164] querying a first reference frequency and a second reference frequency configured for the duration; the first reference frequency is higher than the second reference frequency, and both the first reference frequency and the second reference frequency are negatively correlated with the duration;

[0165] Calculating the difference between the first reference frequency and the second reference frequency to obtain a frequency deviation value;

[0166] Mapping the risk level to a floating ratio; the floating ratio is positively correlated with the risk level;

[0167] Calculate the product of the frequency deviation value and the floating ratio as the floating deviation value;

[0168] The sum of the second reference frequency and the floating deviation value is rounded to obtain the prompt frequency.

[0169] In one embodiment of the present invention, the anti-corrosion maintenance prompt module 508 is also used to:

[0170] Adding the risk level to a preset first adjustment coefficient to obtain a first adjustment level;

[0171] Calculating a ratio between the first adjustment level and a preset second adjustment coefficient to obtain a second adjustment level;

[0172] The second adjustment level is squared to obtain a floating ratio.

[0173] The marine radome structure corrosion prevention device provided in the embodiment of the present invention can execute the marine radome structure corrosion prevention method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the marine radome structure corrosion prevention method.

[0174] Embodiment 3

[0175] See also Figure 6 , 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.

[0176] like Figure 6 As 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.

[0177] 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.

[0178] 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 preventing corrosion of a marine radome structure.

[0179] In some embodiments, the method for preventing corrosion of a marine radome structure 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 preventing corrosion of a marine radome structure described above may be executed. Alternatively, in other embodiments, the processor 11 may be configured to execute the method for preventing corrosion of a marine radome structure in any other appropriate manner (e.g., by means of firmware).

[0180] 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.

[0181] 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.

[0182] 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.

[0183] 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).

[0184] 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.

[0185] 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.

[0186] Embodiment 4

[0187] 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 preventing corrosion of a marine radome structure as provided in any embodiment of the present invention is implemented.

[0188] 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).

[0189] 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.

[0190] 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 preventing corrosion of a marine radome structure, characterized in that: include: Query the first time node of the last anti-corrosion maintenance of the radome structure of the maritime antenna in the ship; During the navigation of the ship, track point data generated after the first time node is collected for the ship; each track point data has coordinate data and date data; Determine a risk detection network; the risk detection network has a first backbone structure, a second backbone structure, a feature fusion module and a head structure; Inputting the coordinate data into the first backbone structure to extract target coordinate features; Inputting the date data into the second backbone structure to extract target date features; Inputting the target coordinate feature and the target date feature into the feature fusion module and fusing them into target spatiotemporal features; Inputting the target spatiotemporal characteristics into the head structure for classification to obtain a corrosion risk level; According to the risk level of corrosion, the antenna cover structure is prompted to perform anti-corrosion maintenance.

2. The method according to claim 1, characterized in that: The first backbone structure includes a first long short-term memory network, a first convolutional layer, a second convolutional layer, a third convolutional layer and a first fully connected layer, and the step of inputting the coordinate data into the first backbone structure to extract the target coordinate features includes: Inputting the coordinate data into the first long short-term memory network to extract first coordinate sequence features; Converting the first coordinate sequence feature into a two-dimensional first coordinate feature graph; Inputting the first coordinate feature map into the first convolutional layer to extract a second coordinate feature map; Activating the second coordinate feature map using a hyperbolic tangent function; Inputting the first coordinate feature map into the second convolutional layer to extract a third coordinate feature map; Activating the third coordinate feature map using a S-type function; Adding the activated second coordinate feature map to the activated third coordinate feature map to obtain a fourth coordinate feature map; Converting the fourth coordinate feature map into a one-dimensional second coordinate sequence feature; The second coordinate sequence feature is input into the first fully connected layer and mapped into the target coordinate feature.

3. The method according to claim 2, characterized in that The second backbone structure includes a second long short-term memory network, a first time convolutional network, a second time convolutional network and a second fully connected layer, and the step of inputting the date data into the second backbone structure to extract the target date feature includes: Inputting the date data into the second long short-term memory network to extract the first date sequence features; Input the first date sequence feature into the first temporal convolutional network to extract the second date sequence feature; Input the second date series feature into the second temporal convolutional network to extract the third date series feature; concatenating the first date series feature and the third date series feature into a fourth date series feature; The fourth date series feature is input into the second fully connected layer and mapped into a target date feature.

4. The method according to claim 3, characterized in that The feature fusion module includes a self-attention layer, a first attention module and a second attention module. The step of inputting the target coordinate feature and the target date feature into the feature fusion module and fusing them into the target spatiotemporal feature includes: Randomly generate the first candidate spatiotemporal feature; Inputting the first candidate spatiotemporal feature into the self-attention layer, and generating a second candidate spatiotemporal feature under the self-attention mechanism; Inputting the second candidate spatiotemporal feature, the target coordinate feature, and the target date feature into the first attention module, and fusing them into a third candidate spatiotemporal feature under the attention mechanism; The third candidate spatiotemporal feature, the target coordinate feature and the target date feature are input into the second attention module and fused into the target spatiotemporal feature under the attention mechanism.

5. The method according to claim 4, characterized in that The first attention module includes a first attention layer and a second attention layer, and the second attention module includes a third attention layer and a fourth attention layer; The step of inputting the second candidate spatiotemporal feature, the target coordinate feature, and the target date feature into the first attention module, and fusing them into a third candidate spatiotemporal feature under the attention mechanism, includes: Inputting the second candidate spatiotemporal feature and the target coordinate feature into the first attention layer, and fusing them into the first basic spatiotemporal feature under the attention mechanism; Input the first basic spatiotemporal feature and the target date feature into the second attention layer, and fuse them into the third candidate spatiotemporal feature under the attention mechanism; The step of inputting the third candidate spatiotemporal feature, the target coordinate feature, and the target date feature into the second attention module and fusing them into the target spatiotemporal feature under the attention mechanism includes: Inputting the third candidate spatiotemporal feature and the target coordinate feature into the third attention layer, and fusing them into the second basic spatiotemporal feature under the attention mechanism; The second basic spatiotemporal features and the target date features are input into the fourth attention layer and fused into the target spatiotemporal features under the attention mechanism.

6. The method according to any one of claims 1 to 5, characterized in that The step of performing anti-corrosion maintenance on the radome structure according to the corrosion risk level includes: querying the ship for a second time point for the next anti-corrosion maintenance of the radome structure of the maritime antenna; Calculate the duration between the current time and the second time node; If the corrosion risk level is greater than or equal to the preset target level, a prompt frequency is generated according to the duration, the risk level and the duration; According to the reminder frequency, the antenna cover structure is prompted to carry out anti-corrosion maintenance.

7. The method according to claim 6, characterized in that The generating of the prompt frequency according to the duration, the risk level and the duration includes: querying a first reference frequency and a second reference frequency configured for the duration; the first reference frequency is higher than the second reference frequency, and both the first reference frequency and the second reference frequency are negatively correlated with the duration; Calculating the difference between the first reference frequency and the second reference frequency to obtain a frequency deviation value; Mapping the risk level to a floating ratio; the floating ratio is positively correlated with the risk level; Calculate the product of the frequency deviation value and the floating ratio as the floating deviation value; The sum of the second reference frequency and the floating deviation value is rounded to an integer to obtain the prompt frequency.

8. The method according to claim 7, characterized in that Mapping the risk level to a floating ratio includes: Adding the risk level to a preset first adjustment coefficient to obtain a first adjustment level; Calculating a ratio between the first adjustment level and a preset second adjustment coefficient to obtain a second adjustment level; The second adjustment level is squared to obtain a floating ratio.

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 execute the method for preventing corrosion of a marine radome 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 preventing corrosion of a marine radome structure according to any one of claims 1 to 8 is implemented.

Images