Bimodal combined ship electronic chart system and method based on Beidou satellite

Abstract

The invention discloses a bimodal combined ship electronic chart system and method based on a Beidou satellite, and belongs to the technical field of ship electronic charts. The method comprises the following steps: collecting a ship multivariate time sequence and key equipment image information, and integrating the information in a terminal server; singular term elimination is carried out on the ship multivariate time sequence, and the singular term is provided for the Beidou satellite information transmission terminal module; extracting a key structure and transmitting an image to a ship operation and maintenance image semantic coding module; a PearSGAN model is established; processing the bimodal information by using a Beidou coding structure; and executing an improved Mercator projection transformation method based on an elliptic dynamic equation, and providing a processing result to a centralized control center high-definition display screen module. The ship multivariate time sequence and the image information are deeply combined to form bimodal combined input information, and a high application value and a general paradigm are provided for further development of navigation guarantee, comprehensive management and other technologies of ocean ships in China.

Description

Technical Field

[0001] The present invention relates to a Beidou satellite-based dual-mode combined ship electronic chart system and method, belonging to the technical field of ship electronic charts. Background Art

[0002] Ship electronic charts are an important basis for ensuring the safe navigation of ships. Tasks such as ship collision avoidance, tide-riding scheduling, and pilotage to dock require ship electronic charts as the implementation basis. Existing electronic charts integrate and display various types of ocean data, navigation routes, navigation decisions and other information. Through internal and external interface connections, they realize the comprehensive display and control of complex navigation data and the situational awareness of ocean-going ships. As technical personnel's demand for ship situation continues to increase, the electronic chart system needs to display more and more multivariate time series, and the image information of key parts of the ship also needs to be perceived and transmitted as an important situation. Combining multivariate time series with images to form bimodal data, and jointly displaying bimodal information on the ship electronic chart to form a bimodal joint ship electronic chart system, and designing a dual-mode joint ship electronic chart display method, are currently in need of further in-depth design and research.

[0003] BeiDou satellites, independently developed by my country, possess both positioning and short message communication capabilities. Using BeiDou's positioning function to display a ship's position in real time, and using its short message communication function to transmit critical multivariate time-series data and imagery required for ship situational awareness, they provide strong support for the design of my country's electronic nautical chart systems and methods.

[0004] Beidou satellite positioning capabilities have reached internationally recognized levels and require no further development or improvement. However, the narrow bandwidth of Beidou satellite communications has long limited the application prospects of short message communication capabilities in various industries, resulting in the following issues:

[0005] 1. The amount of multivariate time series data of ships is huge, and the channel capacity of thousands of bits per time cannot meet the transmission requirements;

[0006] 2. The lack of a key structure extraction module in ship operation and maintenance images results in non-essential information dominating the transmission in limited channels;

[0007] 3. The semantic information of ship operation and maintenance images has not been fully explored, and short message communication remains at the stage of precise bit transmission supported by Shannon's theorem;

[0008] 4. The dual-modal information of ships cannot be accurately positioned in the Beidou satellite electronic chart system after rendering.

[0009] In response to the above problems, the present invention aims to propose a Beidou satellite-based dual-mode joint ship electronic nautical chart system and method to achieve optimized display of existing Beidou satellite ship electronic nautical charts. Summary of the Invention

[0010] In order to solve the problems existing in the background technology, the present invention provides a Beidou satellite-based dual-mode combined ship electronic chart system and method.

[0011] To achieve the above objectives, the present invention adopts the following technical solutions: a Beidou satellite-based dual-mode joint ship electronic chart system, including a ship multivariate time series singularity detection module, which is used to integrate the ship multivariate time series, perform multivariate time series singularity detection, and input the detection results into the Beidou satellite information transmission terminal module;

[0012] The ship operation and maintenance image key structure extraction module is used to collect images of key ship equipment in real time, extract the key structures of the ship operation and maintenance images, and output the extraction results to the ship operation and maintenance image semantic coding module;

[0013] The ship operation and maintenance image semantic encoding module is used to receive image information provided by the ship operation and maintenance image key structure extraction module, execute PearSGAN to achieve semantic communication, and deploy the trained PearSGAN semantic communication model in the Beidou satellite information transmission terminal module;

[0014] Beidou satellite information transmission terminal module, used to receive dual-mode data and conduct ship-to-shore communications in accordance with Beidou short message communication frequency specifications;

[0015] The Beidou satellite electronic chart positioning module is used to transmit the dual-mode information of ships sent by the Beidou satellite information transmission terminal module, executes the improved Mercator projection transformation method based on the elliptical dynamic equation, and displays the processed information on the high-definition display module in the centralized control center;

[0016] The high-definition display screen module in the centralized control center is used to display the multivariate time series of the ship and image information of key parts of the ship, realizing a dual-modal joint ship electronic chart system based on Beidou satellites.

[0017] Furthermore, the ship operation and maintenance image semantic coding module includes

[0018] Original image input and generative image output modules, used for image input and output;

[0019] Pearson correlation calculation module, used to calculate the correlation between image pixels;

[0020] Generator module, used to generate images;

[0021] The discriminator module is used to provide the discrimination results between the original image and the generated image.

[0022] The present invention provides a display method for a Beidou satellite-based dual-mode combined ship electronic chart system, the method comprising the following steps:

[0023] S1: Collect multivariate time series of ships and key equipment image information;

[0024] S2: Build a local area network to integrate dual-modal information consisting of ship multivariate time series and key equipment images into the shipboard computing terminal server;

[0025] S3: In the ship multivariate time series singularity detection module, the first-order difference singular term of the ship multivariate time series is eliminated, and the processed ship multivariate time series is provided to the Beidou satellite information transmission terminal module;

[0026] S3 said singular item elimination comprises the following steps:

[0027] S301: Singular item judgment:

[0028] S30101: Develop a set of error limits based on the maximum, minimum, and median values ​​of a multivariate time series ,in: Indicates the maximum change between adjacent points;

[0029] S30102: Definition The true value at the moment is , The predicted value at time ,like , it means that the point is a singular item;

[0030] S30102 Predicted value at time The first-order difference equation is used for calculation, and the calculation formula is as follows:

[0031] (1)

[0032] In formula (1):

[0033] express The true value of the moment;

[0034] express The true value of the moment.

[0035] S302: Singular term substitution:

[0036] Determine whether:

[0037] (2)

[0038] In formula (2):

[0039] is the maximum error limit;

[0040] ;

[0041] If formula (2) is satisfied, the historical data does not need to be eliminated; otherwise, , then reselect historical data to replace;

[0042] When the number of replacements exceeds five, reselect the starting point and repeat S301 and S302;

[0043] S303: Reselect singular items.

[0044] The S303 includes the following steps:

[0045] S30301: Find three consecutive data points:

[0046] (3)

[0047] S30302: Perform reverse deduction to obtain the replacement The true value of the moment of Predicted value at time :

[0048] (4).

[0049] S4: The ship operation and maintenance image key structure extraction module takes the key equipment image information as input, executes the total score variation joint key structure extraction method, extracts the key structure, and transmits the processed image to the ship operation and maintenance image semantic encoding module;

[0050] The extraction of the key structure described in S4 includes the following steps:

[0051] S401: With the goal of minimizing the extraction deviation, a regularization term is added to constrain overfitting, and the target equation is as follows:

[0052] (5)

[0053] In formula (5):

[0054] Represents the two-dimensional pixel index of the image;

[0055] represents the regularization term;

[0056] Indicates the key structure extraction results;

[0057] Represents image input;

[0058] S402: Define the x-direction partial derivative of the image in the real coordinate system and the y-direction partial derivative , then:

[0059] (6)

[0060] S403: Propose a key structure extraction method combining total score and variation:

[0061] (7)

[0062] In formula (7):

[0063] Both represent linear adjustment coefficients, which are used to dynamically adjust the weights of the x- and y-directions of the real coordinate system;

[0064] (8)

[0065] represents the total change in x direction, Represents the total change in the y direction, which is used to find the difference between adjacent pixel blocks;

[0066] Indicates the change in x direction, Indicates the change in y direction, which is used to describe the difference within each pixel block;

[0067] in:

[0068] Represents the Gaussian filter kernel function;

[0069] Represents the partial derivative of the key structure extraction result in the x direction;

[0070] Represents the partial derivative of the key structure extraction result in the y direction;

[0071] Represents a regular term to prevent overfitting;

[0072] (9)

[0073] in:

[0074] is the horizontal coordinate of pixel point p;

[0075] is the horizontal coordinate of pixel q;

[0076] is the vertical coordinate of pixel point p;

[0077] is the vertical coordinate of pixel q;

[0078] and is the center coordinate of the pixel block corresponding to the pixel point p;

[0079] and It is the center coordinate of the pixel block corresponding to pixel point q.

[0080] S5: The ship operation and maintenance image semantic encoding module receives the key structural information of the ship operation and maintenance image, establishes the PearSGAN model, considers the Pearson correlation between image information, and improves the image generation effect of the model;

[0081] The S5 comprises the following steps:

[0082] S501: Setting a Pearson correlation calculation module to calculate the correlation between adjacent pixels and cutting the image into different proportions as input to the discriminator module;

[0083] S502: Establish the generator module and discriminator module models:

[0084] (10)

[0085] (11)

[0086] In formula (10):

[0087] PearSG represents the generator module;

[0088] PearSD represents the discriminator module;

[0089] represents the Pearson correlation function;

[0090] Conv2d represents the convolution block;

[0091] It is 8 residual blocks connected in series;

[0092] It is 3 sampling blocks connected in series;

[0093] S503: A random signal is input to the generator module, and a generated image is output; the discriminator module obtains the original image and the generated image for identification, and feeds the identification result back to the generator module. The generator module and the discriminator module are repeatedly trained until a given error or a maximum number of training times is reached, and then the training is stopped.

[0094] S6: The BeiDou satellite information transmission terminal module simultaneously receives the multivariate time series extracted by the joint key structure of total-score variation and the image information processed by the PearSGAN model. It uses the coding structure of the BeiDou 4.0 protocol to process the dual-modal information and realize the parallel transmission of the two modes.

[0095] S7: The Beidou satellite electronic chart positioning module processes the ship dual-modal information provided by the Beidou satellite information transmission terminal module, executes the improved Mercator projection transformation method based on the ellipse dynamic equation, and provides the processing results to the high-definition display module of the centralized control center.

[0096] The S7 comprises the following steps:

[0097] S701: Define the longitude and latitude of the ship Time is , the Beidou short message communication delay time is , then the center area of ​​the ship's movement after the communication delay for:

[0098] (12)

[0099] In formula (12):

[0100] Indicates the ship's moving speed;

[0101] is the current time;

[0102] is the ship's moving yaw angle;

[0103] S702: Define the dynamic equation of the ellipse:

[0104] (13)

[0105] In formula (13):

[0106] Represents the x-direction ellipse parameter;

[0107] Indicates the y-direction ellipse parameter;

[0108] express The latitude and longitude of the ship at the time;

[0109] Then we get Moment ellipse dynamic equation as follows:

[0110] (14)

[0111] In formula (14):

[0112] is the maximum ellipse range coefficient;

[0113] S703: Design an improved Mercator projection transformation formula based on the ellipse dynamic equation as follows:

[0114] (15)

[0115] (16)

[0116] In formulas (15) and (16):

[0117] represents the geographic parameter coefficient;

[0118] Indicates the coordinate latitude that needs to be transformed;

[0119] Indicates the radius of the Earth's ellipsoid in the equatorial direction;

[0120] represents the meridian curvature radius;

[0121] represents the radius of curvature of the Maoyou circle;

[0122] Indicates the offset angle of a given longitude relative to the central meridian;

[0123] Similarly, by calculating the longitude, the geographic coordinates can be converted to the actual location on the display screen.

[0124] Compared with the prior art, the present invention has the following beneficial effects:

[0125] 1. The present invention designs a ship multivariate time series singularity detection module that is more compatible with ship electronic charts. Compared with the prior art that only processes one singularity, the ship multivariate time series singularity detection module of the present invention realizes the detection and elimination of multiple singular items through three sub-steps: singular item judgment, replacement, and reselection. The reselection sub-step can recalculate and determine the starting point of the abnormal item, effectively avoiding sensor data offset or abnormal fluctuation caused by noise and electromagnetic interference at the ship operation site, and realizing effective processing of the ship's multivariate time series.

[0126] 2. The present invention designs a key structure extraction module for ship operation and maintenance images to extract the key structures of the images, that is, it provides a key structure extraction method that combines total-score-variation and effectively removes the background part of the ship operation and maintenance scene. It solves the problem that the existing ship electronic nautical charts do not have image display capabilities, the image data volume far exceeds the multivariate time series, and the data difference seriously limits the application of Beidou satellite short message communication function in ship electronic nautical charts.

[0127] 3. The present invention designs a semantic encoding module for ship operation and maintenance images. The core of the present invention lies in the design of a Pearson correlation generative adversarial network model. The Pearson correlation is integrated into the GAN model, and the relationship between each pixel in the image is considered. The Pearson correlation between pixels is used as the guiding input for model training, which helps the generative GAN model to obtain real images more accurately and solve the problem of Beidou satellite image transmission. This solves the problem that the existing Beidou short message communication is not suitable for the design scenario of the dual-modal information transmission of the present invention because the Beidou satellite short message communication is a precise bit transmission based on Shannon theory. Its information transmission is strictly limited by the length of the short message string.

[0128] 4. The present invention takes into account the mobility of ships during navigation, designs an elliptical dynamic equation, and forms an improved Mercator projection transformation method based on the elliptical dynamic equation to determine the display of the ship's latitude and longitude information. The ship's key information can be displayed by clicking the mouse within the area divided by the elliptical dynamic equation, effectively ensuring that the ship's key information is displayed within the correct screen area, ensuring the effective and accurate implementation of the technology.

[0129] 5. This invention deeply combines multivariate time series of ships with image information to form dual-modal input information. This provides high application value and a universal paradigm for the further development of navigation support and integrated management technologies for my country's ocean-going ships. BRIEF DESCRIPTION OF THE DRAWINGS

[0130] Figure 1 is a structural block diagram of the nautical chart system of the present invention;

[0131] Figure 2 This is the structural diagram of the semantic coding module for ship operation and maintenance images;

[0132] Figure 3 This is an example of the display effect of the key structure extraction module of ship operation and maintenance images;

[0133] Figure 4 This is an example diagram of the display results of multivariate time series in the ship's electronic chart system;

[0134] Figure 5 This is an example of the display results of ship operation and maintenance images in the ship electronic chart system. DETAILED DESCRIPTION

[0135] The technical solutions 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 invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0136] A dual-mode combined ship electronic chart system based on Beidou satellite, including

[0137] The ship multivariate time series singularity detection module is used to integrate the ship multivariate time series and provide key information such as the ship's navigation speed, navigation direction, and sway attitude. The ship is equipped with various sensors on site, which are connected to the ship database server in real time for storage. The present invention does not specify the specific type and number of sensors. The module then uses the ship multivariate time series as input, performs multivariate time series singularity detection based on the first-order difference singular term elimination method, and inputs the detection results into the Beidou satellite information transmission terminal module.

[0138] The ship operation and maintenance image key structure extraction module is used to collect real-time video surveillance images of key ship equipment. The hardware typically uses a video camera and manually captures key video screenshots as input images. These images include, but are not limited to, key equipment such as ship gas turbines, transmission mechanisms, and onboard motors. A combined total-score-variance key structure extraction method is then used to extract the key structures of the ship operation and maintenance images. The extracted images no longer contain background noise, but only the main structure of the equipment in the image. Finally, the extraction results are output to the ship operation and maintenance image semantic encoding module.

[0139] The ship operation and maintenance image semantic encoding module is used to receive image information provided by the ship operation and maintenance image key structure extraction module, execute PearSGAN (Pearson correlation generative adversarial network model) to achieve semantic communication, and deploy the trained PearSGAN semantic communication model in the Beidou satellite information transmission terminal module;

[0140] The BeiDou satellite information transmission terminal module is used to receive dual-modal data (data from the ship multivariate time series singularity detection module and data from the ship operation and maintenance image semantic coding module) and conduct ship-to-shore communication in accordance with the BeiDou short message communication frequency specification. This module usually requires a 24V main power supply, a BeiDou satellite antenna, a GPU generative model training server, and a SIM BeiDou communication card. The above hardware components are combined to ensure the stable operation of the BeiDou satellite information transmission terminal module.

[0141] The Beidou satellite electronic chart positioning module is used to transmit dual-mode ship information from the Beidou satellite information transmission terminal module. It executes the improved Mercator projection transformation method based on the ellipse dynamic equation to convert Cartesian coordinates and geographic coordinates, finds the actual position corresponding to the display of key ship information on the screen, and displays the processed information on the high-definition display module of the centralized control center.

[0142] The high-definition display screen module of the centralized control center is used to display the multivariate time series of the ship and the image information of the key parts of the ship, and to display key data in the ship's electronic nautical chart. In order to display it more clearly, the present invention has designed a ship operation and maintenance image display interface. When the user clicks on the area within the ship on the screen, the display will jump to the image information of the ship's operation and maintenance status, realizing a dual-mode joint ship electronic nautical chart system based on Beidou satellite.

[0143] Furthermore, the ship operation and maintenance image semantic coding module includes

[0144] Original image input and generative image output modules, used for image input and output;

[0145] Pearson correlation calculation module, used to calculate the correlation between image pixels;

[0146] Generator module (PearSG), used to generate images;

[0147] The discriminator module (PearSD) is used to provide the discrimination results between the original image and the generated image.

[0148] The present invention provides a display method for a Beidou satellite-based dual-mode combined ship electronic chart system, the method comprising the following steps:

[0149] S1: The ship's display and control equipment collects the ship's multivariate time series (sensor information such as navigation speed, direction, and attitude) in real time for transmission, and uses video monitoring equipment to collect image information of key equipment such as the ship's on-site gas turbine, transmission mechanism, and motor;

[0150] S2: Through the Modbus and TCP / IP fusion communication protocols, a local area network based on the C / S framework is built in the on-site router, and the dual-modal information consisting of the ship's multi-dimensional time series and key equipment images is integrated into the shipboard computing terminal server;

[0151] S3: In the ship multivariate time series singularity detection module, the first-order difference singular term of the ship multivariate time series is eliminated, and the processed ship multivariate time series is provided to the Beidou satellite information transmission terminal module;

[0152] S3 said singular item elimination comprises the following steps:

[0153] S301: Singular item judgment:

[0154] S30101: Develop a set of error limits based on the maximum, minimum, and median values ​​of a multivariate time series ,in: Indicates the maximum change between adjacent points;

[0155] S30102: Definition The true value at the moment is , The predicted value at time ,like , it means that the point is a singular item;

[0156] S30102 Predicted value at time The first-order difference equation is used for calculation, and the calculation formula is as follows:

[0157] (1)

[0158] In formula (1):

[0159] express The true value of the moment;

[0160] express The true value of the moment.

[0161] S302: Singular term substitution:

[0162] Existing methods usually use a one-by-one replacement method to update singular terms. However, when the system detects multiple singular terms continuously, the values ​​after one-by-one replacement will still have large deviations. Therefore, when there are continuous singular terms in the multivariate time series of ships, it is necessary to observe and judge whether the following conditions are met:

[0163] (2)

[0164] In formula (2):

[0165] is the maximum error limit;

[0166] ;

[0167] If formula (2) is satisfied, the historical data does not need to be eliminated; otherwise, , then reselect historical data to replace;

[0168] When the number of substitutions exceeds five, it indicates that there are too many continuity singular terms, and S301 and S302 are repeated after reselecting the starting point;

[0169] S303: Singular item reselection:

[0170] Singular term reselection aims to solve the problem that the first term of the ship multivariate time series is singular by finding three consecutive data points that meet a specific relationship and performing reverse deduction and replacement.

[0171] The S303 includes the following steps:

[0172] S30301: Find three consecutive data points:

[0173] (3)

[0174] S30302: After finding three consecutive data points of the relationship in formula (3), reverse deduction is performed to obtain the replacement The true value of the moment of Predicted value at time :

[0175] (4).

[0176] S4: The ship operation and maintenance image key structure extraction module takes the image information of key equipment such as ship gas turbines, transmission structures, and motors as input, executes the key structure extraction method of total differential variation combined, extracts key structures, combines the total differential variation and differential variation information, reduces the background area data in the equipment site image, and transmits the processed image to the ship operation and maintenance image semantic encoding module;

[0177] The extraction of the key structure described in S4 includes the following steps:

[0178] S401: With the goal of minimizing the extraction deviation, a regularization term is added to constrain overfitting, and the target equation is as follows:

[0179] (5)

[0180] In formula (5):

[0181] Represents the two-dimensional pixel index of the image;

[0182] represents the regularization term;

[0183] Indicates the key structure extraction results;

[0184] Represents image input;

[0185] S402: Define the x-direction partial derivative of the image in the real coordinate system and the y-direction partial derivative , then:

[0186] (6)

[0187] S403: S401 and S402 are the existing image key structure extraction processes, but key structures and textures often overlap, which leads to unclear division between the two. Therefore, a key structure extraction method based on total score variation is proposed:

[0188] (7)

[0189] In formula (7):

[0190] Both represent linear adjustment coefficients, which are used to dynamically adjust the weights of the x- and y-directions of the real coordinate system;

[0191] (8)

[0192] represents the total change in x direction, Represents the total change in the y direction, which is used to find the difference between adjacent pixel blocks;

[0193] Indicates the change in x direction, Indicates the change in y direction, which is used to describe the difference within each pixel block;

[0194] in:

[0195] Represents the Gaussian filter kernel function, specifying that each pixel block has pixels;

[0196] Represents the partial derivative of the key structure extraction result in the x direction;

[0197] Represents the partial derivative of the key structure extraction result in the y direction;

[0198] Represents a regular term to prevent overfitting, which can usually be set to random noise here;

[0199] (9)

[0200] in:

[0201] is the horizontal coordinate of pixel point p;

[0202] is the horizontal coordinate of pixel q;

[0203] is the vertical coordinate of pixel point p;

[0204] is the vertical coordinate of pixel q;

[0205] and is the center coordinate of the pixel block corresponding to the pixel point p;

[0206] and It is the center coordinate of the pixel block corresponding to pixel point q.

[0207] S5: The ship operation and maintenance image semantic encoding module receives the key structural information of the ship operation and maintenance image and establishes the PearSGAN model. Based on the existing GAN model, it considers the Pearson correlation between image information and uses the correlation between pixels to improve the image generation effect of the model.

[0208] The S5 comprises the following steps:

[0209] S501: Setting a Pearson correlation calculation module to calculate the correlation between adjacent pixels and cutting the image into different proportions as input to the discriminator module; specifically, cutting the image into 1 / 2 and 1 / 4, and using the entire image, 1 / 2 image, and 1 / 4 image as input to the discriminator module;

[0210] S502: Establish the generator module and discriminator module models:

[0211] (10)

[0212] (11)

[0213] In formula (10):

[0214] PearSG represents the generator module;

[0215] PearSD represents the discriminator module;

[0216] represents the Pearson correlation function;

[0217] Conv2d represents the convolution block;

[0218] It is 8 residual blocks connected in series;

[0219] It is 3 sampling blocks connected in series;

[0220] Meanwhile, for the discriminator module, the number of channels of the five concatenated Conv2ds are 64, 128, 256, 512, and 1, respectively.

[0221] S503: A random signal is input to the generator module, and the generator module outputs a generated image to the generative image output module. The generative image output module outputs a complete generated image, a 1 / 2 generated image, and a 1 / 4 generated image respectively. The discriminator module obtains the original image and the generated image for identification, and feeds the identification result back to the generator module. The generator module and the discriminator module are repeatedly trained until a given error or a maximum number of training times is reached, and then the training is stopped.

[0222] S6: The BeiDou satellite information transmission terminal module simultaneously receives the multivariate time series extracted from the key structure of the total-score-variogram (TSV) joint method and the image information processed by the PearSGAN model. It uses the BeiDou 4.0 protocol's coding structure to encode and perform parity checking on the dual-modal information. For multi-card devices, data parallel transmission mode can be used to achieve parallel transmission of the dual modes.

[0223] S7: The Beidou satellite electronic chart positioning module processes the ship dual-modal information provided by the Beidou satellite information transmission terminal module, executes the improved Mercator projection transformation method based on the ellipse dynamic equation, and provides the processing results to the high-definition display module of the centralized control center.

[0224] The S7 comprises the following steps:

[0225] S701: Define the longitude and latitude of the ship Time is , the Beidou short message communication delay time is , then the center area of ​​the ship's movement after the communication delay for:

[0226] (12)

[0227] In formula (12):

[0228] Indicates the ship's moving speed;

[0229] is the current time;

[0230] is the ship's moving yaw angle;

[0231] S702: Define the dynamic equation of the ellipse:

[0232] (13)

[0233] In formula (13):

[0234] Represents the x-direction ellipse parameter;

[0235] Indicates the y-direction ellipse parameter;

[0236] express The latitude and longitude of the ship at the time;

[0237] Then we get Moment ellipse dynamic equation as follows:

[0238] (14)

[0239] In formula (14):

[0240] is the maximum ellipse range coefficient, meaning Within the range, users can click the screen to jump to the ship operation and maintenance image interface;

[0241] S703: Design an improved Mercator projection transformation formula based on the ellipse dynamic equation as follows:

[0242] (15)

[0243] (16)

[0244] In formulas (15) and (16):

[0245] represents the geographic parameter coefficient;

[0246] Indicates the coordinate latitude that needs to be transformed;

[0247] Indicates the radius of the Earth's ellipsoid in the equatorial direction;

[0248] represents the meridian curvature radius;

[0249] represents the radius of curvature of the Maoyou circle;

[0250] Indicates the offset angle of a given longitude relative to the central meridian;

[0251] Similarly, longitude calculation is performed, and after the above transformation, the geographic coordinates can be converted into the actual position on the display screen.

[0252] Aiming at the demand of ship electronic nautical chart system, the present invention combines two modalities, multivariate time series and ship image information, and provides a Beidou satellite-based dual-modal joint ship electronic nautical chart system and method.

[0253] The present invention aims to utilize the Beidou satellite independently developed by my country. Under the constraints of the Beidou satellite short message communication string length and the maximum communication frequency, a ship multivariate time series singularity detection module, a ship operation and maintenance image key structure extraction module, a ship operation and maintenance image semantic coding module, and a Beidou satellite electronic nautical chart positioning module are designed and invented.

[0254] The ship multivariate time series singularity detection module designs a first-order difference singular term elimination method to achieve the task of eliminating outliers in the ship multivariate time series sensor data; the ship operation and maintenance image key structure extraction module aims to adapt to the narrow bandwidth channel constraints of the Beidou satellite, and designs a key structure extraction method combining total-score variation to achieve the task of extracting the key structure of the ship operation and maintenance image; on this basis, the ship operation and maintenance image semantic encoding module further designs the Pearson correlation generative adversarial network model PearSGAN, and deploys the trained PearSGAN model in advance in the communication servers between the ship and the shore, transforming the traditional precise bit communication into semantic communication based on the PearSGAN model; the Beidou satellite electronic nautical chart positioning module designs a Mercator projection transformation method based on the improved elliptical dynamic equation.

[0255] The above four key modules are combined with the ship's multi-dimensional time series and image information acquisition, Beidou satellite information transmission terminal, and high-definition display screen of the centralized control center to finally design and implement a Beidou satellite-based dual-modal joint ship electronic nautical chart system and method.

[0256] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim to which it relates.

[0257] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A BeiDou satellite-based dual-mode combined ship electronic chart system, characterized by: include The ship multivariate time series singularity detection module is used to integrate the ship multivariate time series, perform multivariate time series singularity detection, and input the detection results to the Beidou satellite information transmission terminal module; The ship operation and maintenance image key structure extraction module is used to collect images of key ship equipment in real time, extract the key structures of the ship operation and maintenance images, and output the extraction results to the ship operation and maintenance image semantic coding module; The ship operation and maintenance image semantic encoding module is used to receive image information provided by the ship operation and maintenance image key structure extraction module, execute PearSGAN to achieve semantic communication, and deploy the trained PearSGAN semantic communication model in the Beidou satellite information transmission terminal module; Beidou satellite information transmission terminal module, used to receive dual-mode data and conduct ship-to-shore communications in accordance with Beidou short message communication frequency specifications; The Beidou satellite electronic chart positioning module is used to transmit the dual-mode information of ships sent by the Beidou satellite information transmission terminal module, executes the improved Mercator projection transformation method based on the elliptical dynamic equation, and displays the processed information on the high-definition display module in the centralized control center; The high-definition display screen module in the centralized control center is used to display the multivariate time series of the ship and image information of key parts of the ship, realizing a dual-modal joint ship electronic chart system based on Beidou satellites.

2. The BeiDou satellite-based dual-mode combined ship electronic chart system according to claim 1, characterized in that: The ship operation and maintenance image semantic coding module includes Original image input and generative image output modules, used for image input and output; Pearson correlation calculation module, used to calculate the correlation between image pixels; Generator module, used to generate images; The discriminator module is used to provide the discrimination results between the original image and the generated image.

3. A display method for a BeiDou satellite-based dual-mode combined ship electronic chart system according to claim 1 or 2, characterized in that: The method comprises the following steps: S1: Collect multivariate time series of ships and key equipment image information; S2: Build a local area network to integrate dual-modal information consisting of ship multivariate time series and key equipment images into the shipboard computing terminal server; S3: In the ship multivariate time series singularity detection module, the first-order difference singular term of the ship multivariate time series is eliminated, and the processed ship multivariate time series is provided to the Beidou satellite information transmission terminal module; S4: The ship operation and maintenance image key structure extraction module takes the key equipment image information as input, executes the total score variation joint key structure extraction method, extracts the key structure, and transmits the processed image to the ship operation and maintenance image semantic encoding module; S5: The ship operation and maintenance image semantic encoding module receives the key structural information of the ship operation and maintenance image, establishes the PearSGAN model, considers the Pearson correlation between image information, and improves the image generation effect of the model; S6: The BeiDou satellite information transmission terminal module simultaneously receives the multivariate time series extracted by the joint key structure of total-score variation and the image information processed by the PearSGAN model. It uses the coding structure of the BeiDou 4.0 protocol to process the dual-modal information and realize the parallel transmission of the two modes. S7: The Beidou satellite electronic chart positioning module processes the ship dual-modal information provided by the Beidou satellite information transmission terminal module, executes the improved Mercator projection transformation method based on the ellipse dynamic equation, and provides the processing results to the high-definition display module of the centralized control center.

4. The method according to claim 3, wherein: S3 said singular item elimination comprises the following steps: S301: Singular item judgment: S30101: Develop a set of error limits based on the maximum, minimum, and median values ​​of a multivariate time series ,in: Indicates the maximum change between adjacent points; S30102: Definition The true value at the moment is , The predicted value at time ,like , it means that the point is a singular item; S302: Singular term substitution: Determine whether: （2） In formula (2): is the maximum error limit; ； If formula (2) is satisfied, the historical data does not need to be eliminated; otherwise, , then reselect historical data to replace; When the number of replacements exceeds five, reselect the starting point and repeat S301 and S302; S303: Reselect singular items.

5. The method according to claim 4, characterized in that: S30102 Predicted value at time The first-order difference equation is used for calculation, and the calculation formula is as follows: （1） In formula (1): express The true value of the moment; express The true value of the moment.

6. The method according to claim 4, characterized in that: The S303 includes the following steps: S30301: Find three consecutive data points: （3） S30302: Perform reverse deduction to obtain the replacement The true value of the moment of Predicted value at time : （4）。 7. The method according to claim 3, wherein: The extraction of the key structure described in S4 includes the following steps: S401: With the goal of minimizing the extraction deviation, a regularization term is added to constrain overfitting, and the target equation is as follows: （5） In formula (5): Represents the two-dimensional pixel index of the image; represents the regularization term; Indicates the key structure extraction results; Represents image input; S402: Define the x-direction partial derivative of the image in the real coordinate system and the y-direction partial derivative , then: （6） S403: Propose a key structure extraction method combining total score and variation: （7） In formula (7): Both represent linear adjustment coefficients, which are used to dynamically adjust the weights of the x- and y-directions of the real coordinate system; （8） represents the total change in x direction, Represents the total change in the y direction, which is used to find the difference between adjacent pixel blocks; Indicates the change in x direction, Indicates the change in y direction, which is used to describe the difference within each pixel block; in: Represents the Gaussian filter kernel function; Represents the partial derivative of the key structure extraction result in the x direction; Represents the partial derivative of the key structure extraction result in the y direction; Represents a regular term to prevent overfitting; （9） in: is the horizontal coordinate of pixel point p; is the horizontal coordinate of pixel q; is the vertical coordinate of pixel point p; is the vertical coordinate of pixel q; and is the center coordinate of the pixel block corresponding to the pixel point p; and It is the center coordinate of the pixel block corresponding to pixel point q.

8. The method according to claim 3, wherein: The S5 comprises the following steps: S501: Setting a Pearson correlation calculation module to calculate the correlation between adjacent pixels and cutting the image into different proportions as input to the discriminator module; S502: Establish the generator module and discriminator module models: （10） （11） In formula (10): PearSG represents the generator module; PearSD represents the discriminator module; represents the Pearson correlation function; Conv2d represents the convolution block; It is 8 residual blocks connected in series; It is 3 sampling blocks connected in series; S503: A random signal is input to the generator module, and a generated image is output; the discriminator module obtains the original image and the generated image for identification, and feeds the identification result back to the generator module. The generator module and the discriminator module are repeatedly trained until a given error or a maximum number of training times is reached, and then the training is stopped.

9. The method according to claim 3, wherein: The S7 comprises the following steps: S701: Define the longitude and latitude of the ship Time is , the Beidou short message communication delay time is , then the center area of ​​the ship's movement after the communication delay for: （12） In formula (12): Indicates the ship's moving speed; is the current time; is the ship's moving yaw angle; S702: Define the dynamic equation of the ellipse: （13） In formula (13): Represents the x-direction ellipse parameter; Indicates the y-direction ellipse parameter; express The latitude and longitude of the ship at the time; Then we get Moment ellipse dynamic equation as follows: （14） In formula (14): is the maximum ellipse range coefficient; S703: Design an improved Mercator projection transformation formula based on the ellipse dynamic equation as follows: （15） （16） In formulas (15) and (16): represents the geographic parameter coefficient; Indicates the coordinate latitude that needs to be transformed; Indicates the radius of the Earth's ellipsoid in the equatorial direction; represents the meridian curvature radius; represents the radius of curvature of the Maoyou circle; Indicates the offset angle of a given longitude relative to the central meridian; Similarly, by calculating the longitude, the geographic coordinates can be converted to the actual location on the display screen.

Images