Ship navigation collision avoidance control system based on multi-source data

Abstract

The invention relates to the technical field of ship navigation collision avoidance. The invention relates to a ship navigation collision avoidance control system based on multi-source data. The system comprises a collision avoidance range setting unit, a ship recording unit, a route deviation adjusting unit, a collision area analysis unit and a collision avoidance analysis unit. The collision avoidance range setting unit is used for setting a collision avoidance monitoring range, performing image recognition and radar analysis on ships in the collision avoidance monitoring range, initiating route sharing connection to the ships according to the image recognition and the radar analysis, and dividing the ships into shared ships and non-shared ships according to a connection result; the method comprises the following steps: acquiring length and width parameters through a body shape analysis module, combining with a real-time route of a shared ship, a deviation driving route of a non-shared ship and a driving route of a driving ship, simulating a future trajectory, calculating an overlapping region, integrating data through a formula, excluding a transient approaching low-risk region, accurately positioning a high-risk collision region, and obtaining a high-risk collision result. And a visual space reference is provided for collision avoidance decision making.

Description

Technical Field

[0001] The present invention relates to the technical field of ship navigation collision avoidance, and more specifically, to a ship navigation collision avoidance control system based on multi-source data. Background Art

[0002] In the field of ship navigation safety, the existing technologies mainly achieve ship collision avoidance control through means such as radar detection, image recognition, and automatic identification systems. Its core purpose is to monitor information such as the positions and motion states of surrounding ships, analyze the collision risks, and generate collision avoidance strategies to ensure the safety of ships during navigation. However, when a ship is in a port, due to the narrow waterways and dense ships, the traditional system uses a fixed large-radius monitoring range, which is prone to introducing a large amount of invalid data, resulting in a sharp increase in data processing load, and it is unable to accurately identify high-risk targets, which may delay collision avoidance decisions due to information overload. Secondly, when in open waters, the fixed monitoring range is difficult to meet the long-distance early warning requirements. Especially for large ships such as cargo ships, due to their large inertia and long braking distance, if the monitoring range is insufficient, it may lead to a hasty collision warning time and unable to effectively avoid risks. In order to reduce this situation, a ship navigation collision avoidance control system based on multi-source data is proposed. Summary of the Invention

[0003] The purpose of the present invention is to provide a ship navigation collision avoidance control system based on multi-source data to solve the problems raised in the above background art.

[0004] To achieve the above purpose, a ship navigation collision avoidance control system based on multi-source data is provided, including a collision avoidance range setting unit, a ship recording unit, a route deviation adjustment unit, a collision area analysis unit, and a collision avoidance analysis unit. The collision avoidance range setting unit is used to set the collision avoidance monitoring range, perform image recognition and radar analysis on the ships within the collision avoidance monitoring range, initiate a route sharing connection to the ship according to the image recognition and radar analysis, and classify the ships into shared ships and non-shared ships according to the connection result. The ship recording unit is used to analyze the driving directions of non-shared ships, identify risk non-shared ships according to the driving directions combined with the positions of the driving ships, and record the historical driving data of each non-shared ship according to the image recognition and radar analysis. The route deviation adjustment unit is used to perform a similarity match of the driving routes according to the historical driving data combined with the real-time positions of the risk non-shared ships, and analyze the deviation range according to the historical driving data of the risk non-shared ships, and perform a deviation driving route analysis by combining the deviation range with the matched historical driving route. The collision area analysis unit is used to analyze the body size parameters of shared ships and non-shared ships, and then combine the body size parameters with the driving routes, deviation driving routes of the shared ships, and the driving routes of the driving ships to conduct collision area analysis; The collision avoidance analysis unit is used to obtain the driving parameters of the driving ship, conduct driving route collision avoidance analysis based on the driving parameters in combination with the collision area, and generate a new driving route for the driving ship according to the analysis results.

[0005] As a further improvement of this technical solution, the collision avoidance range setting unit establishes a data collection module in the ship control port. The data collection module uses the image device of the ship to collect image data in each direction of the ship, and then accesses the radar system of the ship to detect other ships using the radar system.

[0006] As a further improvement of this technical solution, the collision avoidance range setting unit includes a monitoring range setting module, a shared connection module, and a ship classification module; The monitoring range setting module is used to set the collision avoidance monitoring range with the ship as the center; The collision avoidance monitoring range can be dynamically adjusted according to the position of the ship. When approaching the port, the collision avoidance monitoring range is dynamically reduced, and when far from the port, the collision avoidance monitoring range is dynamically expanded; The shared connection module is used to perform image recognition and radar analysis on the ships within the collision avoidance monitoring range using the radar system and the image device, determine the number of ships entering the collision avoidance monitoring range according to the image recognition data and radar analysis data, and then initiate a route sharing connection to each ship according to the number of ships; The ship classification module is used to classify ships according to the shared connection results. When a route sharing connection is successfully established with a ship, the ship is a shared ship. On the contrary, when a route sharing connection cannot be established with a ship, the ship is a non-shared ship.

[0007] As a further improvement of this technical solution, the ship record unit includes a driving direction analysis module and a historical driving record module; The driving direction analysis module is used to determine the bow direction of the non-shared ship according to the image recognition data for the non-shared ship, then determine the moving path of the ship according to the radar analysis data, and then combine the bow direction with the moving path to determine the driving direction of the non-shared ship, and then combine the driving direction with the position of the driving ship to identify the risk non-shared ship; The historical driving record module is used to record the historical driving data of each non-shared ship according to the image recognition data and radar analysis data.

[0008] As a further improvement of this technical solution, the method for the driving direction analysis module to identify the risk non-shared ship is as follows: If the driving direction of a non - shared ship contains the position of the driving ship, it is identified as a risk non - shared ship; If the driving direction of a non - shared ship does not contain the position of the driving ship, continuous monitoring is maintained; If the position of a non - shared ship is in the driving direction of the driving ship, it is identified as a risk non - shared ship.

[0009] As a further improvement of this technical solution, the route deviation adjustment unit includes a route matching module and a deviation calculation module; The route matching module is used to perform driving route similarity matching according to historical driving data in combination with the real - time position and moving path of the risk non - shared ship, and match the historical driving route with the highest similarity to the real - time position and moving path in the historical driving data; The deviation calculation module is used to analyze the deviation range based on the historical driving data of the risk non - shared ship, set the deviation range according to the analysis result, then expand the route by combining the deviation range with the matched historical driving route, and set the expanded historical driving route as the deviation driving route.

[0010] As a further improvement of this technical solution, the collision area analysis unit includes a ship size analysis module and a collision area analysis module; The ship size analysis module is used to analyze the ship size parameters of shared ships and non - shared ships, and obtain the ship size parameters of shared ships and non - shared ships in the collision avoidance monitoring range; The collision area analysis module is used to perform collision area analysis by combining the ship size parameters with the driving routes of shared ships, the deviation driving routes of risk non - shared ships, and the driving routes of driving ships, and obtain the collision areas that will cause accidents according to the analysis results.

[0011] As a further improvement of this technical solution, the collision avoidance analysis unit includes a driving parameter extraction module and a driving route generation module; The driving parameter extraction module is used to obtain the driving parameters of the driving ship in real - time through the data collection module; The driving route generation module is used to perform driving route collision avoidance analysis by combining the driving parameters with the collision areas, generate a new driving route according to the analysis result, and then send the new driving route to the driving ship through the data collection module for driving control.

[0012] As a further improvement of this technical solution, the steps for the collision avoidance analysis unit to generate a new driving route for the driving ship are as follows: ; Among them, F is the fusion function, E is the driving parameter vector, C is the set of collision regions, Z is the collision risk value, Y is the predicted trajectory. For the ship motion that satisfies the synthesis of uniform acceleration linear motion and uniform angular acceleration turning, the future trajectory is calculated recursively with the time step; ; Among them, T is the time compensation, B is the total time step, Y b is the predicted trajectory point, c r is the r-th collision region, Overlap(Y b , c r ) is the overlap value between the trajectory point and the collision region. When Y b falls into c r which is the spatial range and the time is within the time window of c r , return 1, otherwise return 0. w(c r ) is the collision region weight. The larger the area and the more urgent the time window, the higher the weight; ; Among them, R* is the optimal collision avoidance route, R is the new driving route, w1, w2, w3 are weight coefficients, R collision is the collision risk function, R deviation is the route deviation function, R stability is the navigation stability function, argmin is to select the optimal route value; ; Among them, d min (R*, c r ) is the minimum spatial distance between the new route R* and the collision region c r , ε is a very small positive number; ; Among them, L is the total length of R*, p(s) is the point on the new route with arc length s, R plan (s) is the point on the original planned route corresponding to the arc length, d(p(s), R plan (s)) is the perpendicular distance from the point to the route; ; Among them, Δθ o is the change in the steering angle between adjacent time steps, max|Δθ o | is the maximum steering angle change, a o is the acceleration at each time step, max|a o | is the absolute value of the maximum acceleration, j o is the acceleration, ∑|j o | is the sum of the absolute values of the acceleration, k θ , k a , kj is the weight coefficient; ; where U is the control instruction vector, including rudder angle and throttle opening parameters, and G is the instruction generation function; ; where G TJ is the rudder angle calculation of G, L d is the total preview distance, e y is the lateral offset of the current position relative to the target trajectory, v 2 is the square of the speed, Δθ heading is the difference between the current heading angle and the target heading angle; ; where G YM is the throttle opening calculation of G, Δv is the difference between the target speed and the current speed, t0 is the cut-off time point for integral calculation, t is the time variable, Δt is the control period, and k1, k2, k3 are PID parameters, representing proportional control, integral control, and derivative control respectively.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. In the ship navigation collision avoidance control system based on multi-source data, the length and width parameters are obtained through the body type analysis module, combined with the real-time route of the shared ship, the deviation driving route of the non-shared ship, and the driving route of the driving ship, to simulate the future trajectory and calculate the overlapping area. The data is integrated through the formula, and the low-risk areas of short-term approach are excluded, accurately positioning the high-risk collision area, providing an intuitive spatial reference for collision avoidance decision-making.

[0014] 2. In the ship navigation collision avoidance control system based on multi-source data, the collision avoidance monitoring range is dynamically adjusted according to the relative position of the ship and the port. When approaching the port, the monitoring radius is reduced to reduce the interference of invalid data caused by dense ships in the port and reduce the data processing pressure. When far from the port, the monitoring radius is expanded, and combined with factors such as ship type and sea area environment, ships coming from a long distance are pre-warned in advance to ensure sufficient warning distance in open waters, realizing the accurate matching of the monitoring range and the navigation safety requirements.

[0015] 3. In the ship navigation collision avoidance control system based on multi-source data, the collision risk value is calculated and the trajectory is predicted by combining the driving parameters and the collision area fusion function, and then the collision avoidance route is optimized based on the weight coefficient, balancing the collision risk, route deviation degree, and navigation stability. Finally, the control instructions that can be directly used in the autopilot system are generated, realizing real-time collision avoidance adjustment, reducing route deviation while avoiding collisions, and ensuring navigation efficiency. Description of the Drawings

[0016] Figure 1 This is the overall structural schematic diagram of the present invention.

[0017] The meanings of each label in the figure are as follows: 10, Collision avoidance range setting unit; 20, Ship record unit; 30, Route deviation adjustment unit; 40, Collision area analysis unit; 50, Collision avoidance analysis unit. Specific implementation manners

[0018] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0019] Please refer to Figure 1 As shown, the purpose of this embodiment is to provide a ship navigation collision avoidance control system based on multi-source data, including a collision avoidance range setting unit 10, a ship record unit 20, a route deviation adjustment unit 30, a collision area analysis unit 40, and a collision avoidance analysis unit 50; The collision avoidance range setting unit 10 is used to set the collision avoidance monitoring range, perform image recognition and radar analysis on the ships within the collision avoidance monitoring range, initiate a route sharing connection to the ship according to the image recognition and radar analysis, and classify the ships into shared ships and non-shared ships according to the connection result; The collision avoidance range setting unit 10 establishes a data collection module in the ship control port, and uses the data collection module to collect image data in each direction of the ship through the ship's image device, and then accesses the ship's radar system to detect other ships using the radar system.

[0020] Call the image device configured on the ship (such as a surround camera, a specific direction monitoring lens, etc.), and obtain the image data in each direction (bow, stern, port side, starboard side, etc.) of the ship at the set acquisition frequency, and record the basic information such as the time stamp corresponding to the image and the shooting azimuth; Access the ship radar system through the data collection module, and use the radar active detection mechanism to obtain the radar echo data such as the distance, azimuth, and motion trend (calculated through multiple frames of data) of other ships and obstacles within the collision avoidance monitoring range, and synchronously record the detection time, radar working parameters and other information.

[0021] The collision avoidance range setting unit 10 includes a monitoring range setting module, a shared connection module, and a ship classification module; The monitoring range setting module is used to set the collision avoidance monitoring range with the ship as the center; The collision avoidance monitoring range can be dynamically adjusted according to the position of the ship. When approaching the port, the collision avoidance monitoring range is dynamically reduced, and when moving away from the port, the collision avoidance monitoring range is dynamically expanded. The specific steps are as follows: Initialization of the reference monitoring range: Taking the position of the ship itself as the geometric center, set the initial collision avoidance monitoring range according to the general safety requirements of ship navigation and the ship type (such as different safety distance requirements for cargo ships and passenger ships). Then, obtain the geographical location information of the ship in real time, and identify the relative position relationship between the ship and the port through the ship positioning system (such as GPS, Beidou). Execution of the dynamic adjustment rule: According to the characteristics of the port waters (such as narrow channels and dense ships), shrink the monitoring range from the initial radius D0 towards the position of the central ship according to the preset adjustment algorithm (or empirical rule). The adjusted monitoring radius D1 meets the safety monitoring requirements of the port waters, and at the same time avoids introducing too many invalid interferences due to too large a range (such as dense port operation ships, large-scale monitoring will increase the data processing pressure). Expansion when moving away from the port: When the ship moves away from the port and enters the open sea, in order to give early warnings of approaching ships, obstacles, etc. from a long distance, according to the principle of safety redundancy, expand the monitoring range from D0 outwards to obtain a new monitoring radius D2, ensuring sufficient warning distance and reaction time to ensure the navigation safety of the ship in the open waters. The formula is as follows: ; Among them, D(d) is the collision avoidance monitoring range, D0 is the radius of the initial collision avoidance monitoring range, k is an adjustment coefficient, which is related to the ship type, sea area environment, and safety standard, and determines the influence degree of the distance change on the adjustment of the monitoring range. When the cargo ship has a large volume and large inertia, k takes a larger value to ensure a farther monitoring distance. sign is the sign function, which returns 1 when d - d0 > 0 and returns -1 when d - d0 < 0, controlling the adjustment direction of the monitoring range. d is the real-time distance between the ship and the port, the actual spatial distance obtained through the positioning system, and d0 is the port distance threshold, which is the critical distance triggering the adjustment of the monitoring range. When the ship is at a distance d0 from the port, the monitoring range is the initial radius D0.

[0022] The shared connection module is used to perform image recognition and radar analysis on the ships within the collision avoidance monitoring range by using the radar system and the image device, determine the number of ships entering the collision avoidance monitoring range according to the image recognition data and radar analysis data, and then initiate a route sharing connection to each ship according to the number of ships; Multi-source data acquisition and ship perception: First, use the radar system equipped on the ship to emit electromagnetic waves, receive the echoes reflected by the targets within the collision avoidance monitoring range, and obtain radar data such as the distance, azimuth, and speed of the ships; at the same time, use the image device (such as a camera) to take pictures of the monitoring range to obtain image data. Through the preliminary processing of the radar data and image data, identify the ship objects existing within the monitoring range; Ship quantity statistics: Summarize the ship information obtained through image recognition and radar analysis, count the specific number of ships entering the collision avoidance monitoring range, and clarify how many ships need to carry out subsequent interactions; Route sharing connection initiation: According to the counted number of ships, send a route sharing connection request to each identified ship one by one. This request includes the identity identifier of its own ship and basic navigation information (such as the current route, speed, etc.) so that the other ship can judge whether to accept the connection; The ship classification module is used to classify ships according to the sharing connection result. When a route sharing connection is successfully established with a ship, then this ship is a shared ship. On the contrary, when a route sharing connection cannot be established with a ship, then this ship is a non-shared ship; Wait and receive the feedback from the other ship on the route sharing connection request. If a ship's feedback is to agree to the connection and data sharing and interaction of the navigation route, etc. can be realized, then classify it as a shared ship; if the feedback is to reject the connection or the connection establishment fails (such as communication failures, etc.) and route sharing cannot be carried out, then classify this ship as a non-shared ship.

[0023] The ship recording unit 20 is used to analyze the driving direction of non-shared ships, identify risk non-shared ships according to the driving direction combined with the position of the ship being driven, and record the historical driving data of each non-shared ship according to image recognition and radar analysis; The ship recording unit 20 includes a driving direction analysis module and a historical driving record module; The driving direction analysis module is used to determine the bow direction of non-shared ships according to the image recognition data for non-shared ships, then determine the moving path of the ships according to the radar analysis data, then combine the bow direction with the moving path to judge the driving direction of non-shared ships, and then combine the driving direction with the position of the ship being driven to identify risk non-shared ships; With the help of image recognition technology, analyze the image data of non-shared ships, and determine the azimuth pointed by the bow of each non-shared ship by identifying the appearance features (such as the bow shape, logo, etc.) in the ship image to clarify its orientation information; Use the radar system to continuously collect the radar echo data of non-shared ships, and based on the information such as the distance and azimuth of the ships in the radar data, through data processing and trajectory calculation, determine the moving path of the ships within a period of time to understand its navigation trajectory trend; Combine the identified bow direction with the determined moving path, take the bow direction as a reference, and combine the trend of the moving path to comprehensively judge the overall driving direction of non-shared ships.

[0024] The method for the driving direction analysis module to identify risk non-shared ships is as follows: If the driving direction of a non - shared ship contains the position of the ship being driven, it is identified as a risk non - shared ship; If the driving direction of a non - shared ship does not contain the position of the ship being driven, continue to monitor; If the position of a non - shared ship is in the driving direction of the ship being driven, it is identified as a risk non - shared ship.

[0025] The historical driving record module is used to record the historical driving data of each non - shared ship based on image recognition data and radar analysis data. The specific steps are as follows: Data acquisition planning: Determine the dimensions of the historical driving data to be recorded. Based on the capabilities of image recognition and radar analysis, clarify the basic and key data content to be collected, including the position, driving direction, speed, and navigation time interval of the ship, etc., to set the acquisition scope for subsequent recording work; Multi - source data association: Associate the visual information of non - shared ships obtained by image recognition (such as the position and attitude of the ship in the picture at different times) with the data such as distance, azimuth, and motion parameters obtained by radar analysis. Through identifiers such as timestamps, match the image data and radar data of the same ship at the same moment to ensure the consistency and accuracy of the data sources; Sort the associated non - shared ship data in chronological order. Starting from the initial moment when the ship enters the monitoring range, sort out the driving data at each time node in turn, construct a sequence arranged along the time axis to clearly present the changes in the ship's driving process, and store the sorted non - shared ship driving data arranged in time series into a dedicated database or data file, creating an independent historical driving data file for each non - shared ship for subsequent links such as route analysis and risk prediction.

[0026] The route deviation adjustment unit 30 is used to perform driving route similarity matching based on historical driving data combined with the real - time position of the risk non - shared ship, and at the same time analyze the deviation range based on the historical driving data of the risk non - shared ship, and perform deviation driving route analysis by combining the deviation range with the matched historical driving route; The route deviation adjustment unit 30 includes a route matching module and a deviation calculation module; The route matching module is used to perform driving route similarity matching based on historical driving data combined with the real - time position and moving path of the risk non - shared ship, and match the historical driving route with the highest similarity to the real - time position and moving path in the historical driving data; Collect the historical driving data of the risk non - shared ship, including information such as its position, heading, and speed at different past times. At the same time, obtain the current real - time position and the moving path data of the ship, and integrate the historical and real - time data to provide a comprehensive data basis for subsequent analysis; Compare the real-time position and moving path of the risk non-sharing ship with each route in the historical voyage data. Through matching algorithms and metrics, evaluate the similarity between the real-time path and each historical route, and find the historical voyage route with the highest similarity as the reference route. The deviation calculation module is used to analyze the deviation range based on the historical voyage data of the risk non-sharing ship, set the deviation range according to the analysis results, then combine the deviation range with the matched historical voyage route for route expansion, and set the expanded historical voyage route as the deviation voyage route.

[0027] For the historical voyage route with the highest similarity found, combined with the historical voyage data of the risk non-sharing ship, analyze the fluctuations of parameters such as position, heading, and speed during its past voyages, count the degree and range of deviation of these parameters from the historical route, and determine a reasonable deviation range value to represent the possible deviation degree of the ship's voyage. Based on the matched historical voyage route, expand on both sides and in front of the route according to the determined deviation range. Take the expanded route as the deviation voyage route. This route takes into account various possible deviation situations during the ship's voyage and more comprehensively reflects the possible future voyage trajectory of the ship. The formula is as follows: ; where Sim(h i , R) is the similarity, h i is the i-th historical route, Q is the real-time path, α, β, χ are weight coefficients, psim is the position similarity, based on the minimum distance from a point to the path, vsim is the speed similarity, comparing the difference between the historical and real-time speeds, dsim is the direction similarity, calculating the cosine similarity of the heading angle, and selecting the most similar historical route h match According to h match generate the deviation route for each point (x, y, t) on it as follows: ; where M(x, y, t) is the set of deviation points corresponding to the point (x, y, t), Δx and Δy are the deviation amounts representing the spatial coordinates, that is, the offset values relative to the reference position (x, y), x, y are the spatial position coordinates of the risk non-sharing ship at a certain moment respectively, t is the time parameter, μ(0, ∑) is a two-dimensional normal distribution with a mean of 0 and a covariance matrix of Σ; ; where M route is the complete deviation voyage route after integrating all deviation points.

[0028] The collision area analysis unit 40 is used to analyze the body size parameters of shared vessels and non-shared vessels, and then combine the body size parameters with the driving routes of the shared vessels, the deviation driving routes, and the driving routes of the piloted vessels to conduct collision area analysis; The collision area analysis unit 40 includes a body size analysis module and a collision area analysis module; The body size analysis module is used to analyze the body size parameters of shared vessels and non-shared vessels, and obtain the body size parameters of shared vessels and non-shared vessels within the collision avoidance monitoring range; For example, determine the length and width by analyzing the ship's contour through images, and use data such as radar echo intensity to assist in calculating parameters such as draft depth and height, and establish a body size parameter file containing information such as length, width, and height for each ship.

[0029] The collision area analysis module is used to combine the body size parameters with the driving routes of the shared vessels, the deviation driving routes of risk non-shared vessels, and the driving routes of the piloted vessels to conduct collision area analysis, and obtain the collision areas that may cause accidents according to the analysis results. The specific steps are as follows: Integration of driving route data: Summarize the real-time driving routes of shared vessels, the deviation driving routes of risk non-shared vessels (historical data deviation ranges have been considered), and the current driving routes of piloted vessels, ensure that each route data contains key information such as position, direction, and speed, and form a multi-ship driving route data set; Collision area modeling analysis: Combine the body size parameters of each ship with the driving routes, and based on the ship's motion laws (such as turning radius, braking distance) and geometric relationships, simulate the motion trajectories of different ships within a certain period in the future. By analyzing the overlapping parts and proximity of the trajectories, calculate the coordinate range of the possible collision area and mark it as the potential collision area; Collision risk confirmation: Conduct risk assessment on the preliminarily determined potential collision areas, combine dynamic data such as ship speed and course change trends, exclude areas that are briefly close but actually have no collision risk, and finally determine the high-risk accident collision areas to provide a basis for subsequent collision avoidance decisions.

[0030] The collision avoidance analysis unit 50 is used to obtain the driving parameters of the piloted vessel, conduct driving route collision avoidance analysis according to the driving parameters combined with the collision area, and generate a new driving route for the piloted vessel according to the analysis results.

[0031] The collision avoidance analysis unit 50 includes a driving parameter extraction module and a driving route generation module; The driving parameter extraction module is used to obtain the driving parameters of the piloted vessel in real time through the data collection module; The data collection module through the ship control port reads various dynamic parameters of the ship in real time, including but not limited to the current speed, heading angle, acceleration, steering rate, etc. At the same time, it obtains the static parameters of the ship itself (such as the ship length, ship width, draft), providing basic data support for collision avoidance analysis.

[0032] The travel route generation module is used to perform travel route collision avoidance analysis by combining driving parameters with the collision area. According to the analysis results, a new travel route is generated, and then the new travel route is sent to the ship in motion through the data collection module for travel control (the generated new travel route is transmitted to the ship's autopilot system through the data collection module. After the system receives the route data, it automatically parses the sequence of coordinate points and converts them into specific rudder control instructions, throttle instructions, etc., to realize the real-time adjustment and control of the ship's travel route).

[0033] Associate the driving parameters (such as speed and heading) collected in real time with the coordinates of the potential collision area output by the collision area analysis unit in the spatial and temporal dimensions. For example, according to the current speed and heading, predict the travel trajectory of the ship in motion in the next period of time, and superimpose it on the spatial range of the collision area to determine whether there is a risk of trajectory overlap. The formula is as follows: ; Among them, F is the fusion function, which outputs the collision risk value (such as a probability between 0 and 1) and the future predicted trajectory under the current parameters. E is the driving parameter vector, C is the set of collision areas, Z is the collision risk value, Y is the predicted trajectory. For the ship's motion that satisfies the synthesis of uniformly accelerated linear motion and uniformly angular accelerated steering, the future trajectory is calculated recursively with a time step; ; Among them, T is the time compensation, B is the total time step, Y b is the predicted trajectory point, c r is the r-th collision area, Overlap(Y b , c r ) is the overlap value between the trajectory point and the collision area. When Y b falls within the spatial range of c r and the time is within the time window of c r , 1 is returned, otherwise 0 is returned. w(c r ) is the weight of the collision area. The larger the area and the more urgent the time window, the higher the weight; ; Among them, R* is the optimal collision avoidance route, R is the new travel route, w1, w2, w3 are weight coefficients, corresponding to the priorities of collision risk, route deviation degree, and navigation stability respectively. R collision is the collision risk function, which is positively correlated with the risk value output by F(E,C). Rdeviation is the route deviation function, which calculates the distance deviation between the new route and the original planned route, R stability is the navigation stability function, which is positively correlated with the steering angle and acceleration fluctuation, and argmin is used to select the optimal route value; ; where d min (R*, c r ) is the minimum spatial distance between the new route R* and the collision area c r , and ε is a very small positive number; ; where L is the total length of R*, p(s) is the point on the new route with arc length s, R plan (s) is the corresponding point on the original planned route with the same arc length, and d(p(s), R plan (s)) is the perpendicular distance from the point to the route; ; where Δθ o is the change in the steering angle between adjacent time steps, max|Δθ o | is the maximum change in the steering angle, a o is the acceleration at each time step, max|a o | is the maximum absolute value of the acceleration, j o is the acceleration, ∑|j o | is the sum of the absolute values of the acceleration, k θ , k a , k j are the weight coefficients; ; where U is the control instruction vector, which includes parameters such as the rudder angle and throttle opening, and G is the instruction generation function; ; where G TJ is the rudder angle calculation of G, L d is the total preview distance, e y is the lateral offset of the current position relative to the target trajectory, v 2 is the square of the speed, and Δθ heading is the difference between the current heading angle and the target heading angle; ; where G YM is the throttle opening calculation of G, Δv is the difference between the target speed and the current speed, t0 is the cut-off time point for integral calculation, t is the time variable, Δt is the control period, and k1, k2, k3 are the PID parameters, representing proportional control, integral control, and derivative control respectively.

[0034] The basic principles, main features and advantages of the present invention have been shown and described above. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and the descriptions in the specification are only preferred examples of the present invention and are not used to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of the present invention claimed is defined by the appended claims and their equivalents.

Claims

1. A ship navigation collision avoidance control system based on multi-source data, characterized in that: It includes a collision avoidance range setting unit (10), a ship recording unit (20), a route deviation adjustment unit (30), a collision area analysis unit (40), and a collision avoidance analysis unit (50); The collision avoidance range setting unit (10) is used to set the collision avoidance monitoring range, perform image recognition and radar analysis on ships within the collision avoidance monitoring range, initiate a route sharing connection to the ship according to the image recognition and radar analysis, and classify the ships into shared ships and non-shared ships according to the connection result; The ship recording unit (20) is used to analyze the traveling direction of non-shared ships, identify risk non-shared ships according to the traveling direction combined with the position of the driving ship, and record the historical traveling data of each non-shared ship according to the image recognition and radar analysis; The route deviation adjustment unit (30) is used to perform a similarity match of the traveling route according to the historical traveling data combined with the real-time position of the risk non-shared ship, analyze the deviation range according to the historical traveling data of the risk non-shared ship, and perform a deviation traveling route analysis by combining the deviation range with the matched historical traveling route; The collision area analysis unit (40) is used to analyze the body size parameters of shared ships and non-shared ships, and then perform a collision area analysis by combining the body size parameters with the traveling routes of shared ships, deviation traveling routes, and the traveling route of the driving ship; The collision avoidance analysis unit (50) is used to obtain the driving parameters of the driving ship, perform a traveling route collision avoidance analysis according to the driving parameters combined with the collision area, and generate a new traveling route for the driving ship according to the analysis result.

2. The ship navigation collision avoidance control system based on multi-source data according to claim 1 is characterized in that: The collision avoidance range setting unit (10) collects image data in each direction of the ship by using the image device of the ship through establishing a data collection module in the ship control port, and then detects other ships by accessing the radar system of the ship and using the radar system.

3. The ship navigation collision avoidance control system based on multi-source data according to claim 1, characterized in that: The collision avoidance range setting unit (10) includes a monitoring range setting module, a shared connection module, and a ship classification module; The monitoring range setting module is used to set the collision avoidance monitoring range with the ship as the center; The collision avoidance monitoring range can be dynamically adjusted according to the position of the ship. When approaching the port, the collision avoidance monitoring range is dynamically reduced, and when far from the port, the collision avoidance monitoring range is dynamically expanded; The shared connection module is used to perform image recognition and radar analysis on the ships within the collision avoidance monitoring range by using the radar system and the image device, determine the number of ships entering the collision avoidance monitoring range according to the image recognition data and radar analysis data, and then initiate a route sharing connection to each ship according to the number of ships; The ship classification module is used to classify ships according to the shared connection result. When a route sharing connection is successfully established with a ship, the ship is a shared ship. On the contrary, when a route sharing connection cannot be established with a ship, the ship is a non-shared ship.

4. The ship navigation collision avoidance control system based on multi-source data according to claim 1, characterized in that: The ship recording unit (20) includes a traveling direction analysis module and a historical traveling record module; The driving direction analysis module is used to determine the bow direction of a non - shared ship based on image recognition data, then determine the moving path of the ship according to radar analysis data, and then combine the bow direction with the moving path to determine the driving direction of the non - shared ship. Then, it combines the driving direction with the position of the driving ship to identify the risk non - shared ship; The historical driving record module is used to record the historical driving data of each non - shared ship according to image recognition data and radar analysis data.

5. The ship navigation collision avoidance control system based on multi-source data according to claim 4, characterized in that: The method for the driving direction analysis module to identify the risk non - shared ship is as follows: If the driving direction of the non - shared ship contains the position of the driving ship, it is identified as a risk non - shared ship; If the driving direction of the non - shared ship does not contain the position of the driving ship, continue to monitor; If the position of the non - shared ship is in the driving direction of the driving ship, it is identified as a risk non - shared ship.

6. The ship navigation collision avoidance control system based on multi-source data according to claim 1, characterized in that: The route deviation adjustment unit (30) includes a route matching module and a deviation calculation module; The route matching module is used to perform driving route similarity matching according to historical driving data, combined with the real - time position and moving path of the risk non - shared ship, and match the historical driving route with the highest similarity to the real - time position and moving path in the historical driving data; The deviation calculation module is used to analyze the deviation range according to the historical driving data of the risk non - shared ship, set the deviation range according to the analysis result, and then expand the route by combining the deviation range with the matched historical driving route, and set the expanded historical driving route as the deviation driving route.

7. A ship navigation collision avoidance control system based on multi-source data according to claim 1, characterized in that: The collision area analysis unit (40) includes a ship size analysis module and a collision area analysis module; The ship size analysis module is used to analyze the ship size parameters of shared ships and non - shared ships, and obtain the ship size parameters of shared ships and non - shared ships in the collision avoidance monitoring range; The collision area analysis module is used to perform collision area analysis by combining the ship size parameters with the driving routes of shared ships, the deviation driving routes of risk non - shared ships, and the driving routes of driving ships, and obtain the collision areas that will cause accidents according to the analysis result.

8. A ship navigation collision avoidance control system based on multi-source data according to claim 1, characterized in that: The collision avoidance analysis unit (50) includes a driving parameter extraction module and a driving route generation module; The driving parameter extraction module is used to obtain the driving parameters of the driving ship in real - time through the data collection module; The driving route generation module is used to perform driving route collision avoidance analysis by combining the driving parameters with the collision area, generate a new driving route according to the analysis result, and then send the new driving route to the driving ship through the data collection module for driving control.

9. A ship navigation collision avoidance control system based on multi-source data according to claim 1, characterized in that: The steps for the collision avoidance analysis unit (50) to generate a new driving route for the driving ship are as follows: ； Among them, F is the fusion function, E is the driving parameter vector, C is the set of collision areas, Z is the collision risk value, Y is the predicted trajectory. For the ship motion satisfying the synthesis of uniformly accelerated linear motion and uniformly angular - accelerated turning, the future trajectory is calculated recursively with a time step; ； Where T is the time compensation, B is the total time step, and Y b is the predicted trajectory point, c r is the r-th collision area, Overlap(Y b , c r ) is the overlap value between the trajectory point and the collision area. When Y b falls into c r which is the spatial range and the time is within the time window of c r , return 1; otherwise, return 0. w(c r ) is the collision area weight. The larger the area and the more urgent the time window, the higher the weight; ； Among them, R* is the optimal collision avoidance route, R is the new driving route, w1, w2, and w3 are weight coefficients, R collision is the collision risk function, R deviation is the route deviation function, R stability is the navigation stability function, and argmin is to select the optimal route value; ； Among them, d min (R*, c r ) is the new route R* and the collision area c r The minimum spatial distance, ε is a very small positive number; ； where L is the total length of R*, p(s) is the point on the new route with arc length s, and R plan (s) is the point on the original planned route corresponding to the arc length, and d(p(s), R plan (s)) is the perpendicular distance from the point to the route; ； where, Δθ o is the change in steering angle between adjacent time steps, max|Δθ o | is the maximum change in steering angle, a o is the acceleration at each time step, max|a o | is the absolute value of the maximum acceleration, j o is the acceleration, ∑|j o | is the sum of the absolute values of the acceleration, k θ , k a , k j is the weight coefficient; ; Among them, U is the control instruction vector, including rudder angle and throttle opening parameters, and G is the instruction generation function; ； Among them, G TJ is the rudder angle calculation of G, L d is the total preview distance, e y is the lateral offset of the current position relative to the target trajectory, v 2 is the square of the speed, Δθ heading is the difference between the current heading angle and the target heading angle; ； Among them, G YM is the throttle opening calculation of G, Δv is the difference between the target speed and the current speed, t0 is the cut-off time point for integral calculation, t is the time variable, Δt is the control period, and k1, k2, and k3 are PID parameters, representing proportional control, integral control, and derivative control respectively.

Images