Intelligent ship collision avoidance method and device

Abstract

The invention provides an intelligent ship collision avoidance method and device, and belongs to the technical field of ship navigation control, and the method comprises the steps: building a water area environment model according to the real-time navigation data and water area environment data of a to-be-controlled ship, and determining a real-time collision avoidance demand according to the water area environment model; determining a predicted relative trajectory according to the real-time navigation data of the to-be-controlled ship and the surrounding ships and the water area environment model, and determining the motion conflict radius of each surrounding ship according to the predicted relative trajectory; performing collision avoidance strategy generation optimization according to the motion conflict radius and the real-time collision avoidance demand to obtain a ship collision avoidance strategy; and adjusting the navigation state of the to-be-controlled ship according to the ship collision avoidance strategy and a preset collision avoidance rule until no collision risk exists and a re-navigation condition is met. Static obstacle information and dynamic change information of surrounding ships can be comprehensively considered in the collision avoidance decision, the accuracy of the collision avoidance decision is improved, and the collision risk is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of ship navigation control, and particularly to an intelligent ship collision avoidance method and device. Background Art

[0002] During the navigation process, ship collision avoidance is an issue that cannot be ignored. Currently, a large number of collision avoidance algorithms have been successfully applied in open waters. However, in restricted waters, the autonomous collision avoidance decision-making of ships still has great limitations. Due to the existence of a large number of dynamic and static obstacles in restricted waters, the navigable water area is limited, and at the same time, the ship is subject to certain uncertainties in the external environment interference, which directly leads to complex traffic encounter situations and threatens the navigation safety of ships.

[0003] Currently, the ship collision avoidance method in open waters is mainly based on the "International Regulations for Preventing Collisions at Sea" (referred to as the "Collision Regulations"). To reduce collision accidents caused by head-on encounters or near head-on encounters between ships, the international community generally adopts the traffic separation scheme. However, compared with open waters, the restricted water environment has a more complex marine environment and dynamic incoming ship information. For example, in the multi-ship intersection scenario in the key waterway area, when a ship is navigating, in addition to taking collision avoidance strategies based on the positions of static obstacles and other ships, it also needs to consider whether the dynamic changes in the shipping lanes that occur when other ships adjust their shipping lanes or avoid collisions in restricted waters will conflict with the collision avoidance strategies of this ship. Traditional methods often make collision avoidance decisions based on static obstacle information and the position information of surrounding ships, without considering the conflict between the dynamic change information of surrounding ships and the collision avoidance strategy of this ship, and it is difficult to make the optimal collision avoidance decision, resulting in a collision risk.

[0004] Therefore, the prior art has the technical problems that the conflict between the dynamic change information of surrounding ships and the collision avoidance strategy of this ship is not considered, it is difficult to make the optimal collision avoidance decision, and there is a collision risk. Summary of the Invention

[0005] In view of this, it is necessary to provide an intelligent ship collision avoidance method and device for comprehensively considering the dynamic change information of surrounding ships in ship collision avoidance decision-making, improving the accuracy of collision avoidance decision-making, and reducing the collision risk.

[0006] In a first aspect, the present invention provides an intelligent ship collision avoidance method, including: Construct a water area environment model based on the real-time navigation data of the ship to be controlled and the water area environment data, and determine the real-time collision avoidance requirements according to the water area environment model; Determine the predicted relative trajectory according to the real-time navigation data of the ship to be controlled and surrounding ships and the water area environment model, and determine the movement conflict radius of each surrounding ship according to the predicted relative trajectory; Generate and optimize a collision avoidance strategy according to the movement conflict radius and the real-time collision avoidance requirements to obtain a ship collision avoidance strategy; Adjust the navigation state of the ship to be controlled according to the ship collision avoidance strategy and the preset collision avoidance rules until there is no collision risk and the conditions for resuming navigation are met.

[0007] In some possible implementation manners, the real-time navigation data includes navigation position information, and the water area environment data includes waterway information and marine environment information. A water area environment model is constructed according to the real-time navigation data and the water area environment data of the ship to be controlled, including: Perform artificial potential field method modeling according to the navigation position information and the waterway information to obtain a waterway centerline gravitational function and a boundary repulsive force function; Perform water flow field modeling according to the marine environment information to obtain an environmental water flow field; Construct a water area environment model according to the waterway centerline gravitational function, the boundary repulsive force function, and the environmental water flow field.

[0008] In some possible implementation manners, determining real-time collision avoidance requirements according to the water area environment model includes: Determine collision constraints according to the water area environment model, and determine real-time collision avoidance requirements according to the collision constraints.

[0009] In some possible implementation manners, determining a predicted relative trajectory according to the real-time navigation data and the water area environment model of the ship to be controlled and surrounding ships includes: Perform kinematic model analysis on the real-time navigation data of the ship to be controlled and surrounding ships to obtain a relative motion trajectory; Perform trajectory correction on the relative motion trajectory according to the water area environment model to obtain a predicted relative trajectory; Wherein, the real-time navigation data includes static data and dynamic data, the static data includes ship size, ship steering performance, and ship power performance, and the dynamic data includes real-time positioning data.

[0010] In some possible implementation manners, the water area environment model includes a waterway sub-model and a water flow field sub-model. Performing trajectory correction on the relative motion trajectory according to the water area environment model to obtain a predicted relative trajectory includes: Perform hydrodynamic analysis on surrounding ships according to the water flow field sub-model to obtain flow field correction parameters, and perform trajectory correction on the relative motion trajectory according to the flow field correction parameters to obtain a preliminary corrected trajectory; Perform waterway analysis on surrounding ships according to the waterway sub-model to obtain waterway correction parameters, and perform trajectory correction on the preliminary corrected trajectory according to the waterway correction parameters to obtain a predicted relative trajectory.

[0011] In some possible implementation manners, determining the motion conflict radius of each surrounding ship according to the predicted relative trajectory includes: Perform kinematic analysis according to the predicted relative trajectory and the preset collision avoidance rules to obtain the motion conflict radius of each surrounding ship.

[0012] In some possible implementation manners, a ship collision avoidance strategy is generated and optimized according to a motion conflict radius and real-time collision avoidance requirements, including: Construct an initial collision avoidance strategy according to real-time collision avoidance requirements; Construct a reward function according to the motion conflict radius and real-time collision avoidance requirements; Optimize the initial collision avoidance strategy according to the initial collision avoidance strategy and the reward function by using the deep deterministic policy gradient algorithm to obtain the ship collision avoidance strategy; Among them, the reward function includes a collision avoidance reward function, a homeward navigation reward function, and a boundary safety reward function.

[0013] In some possible implementation manners, generating and optimizing a collision avoidance strategy according to a motion conflict radius and real-time collision avoidance requirements further includes: If there are at least two ships to be controlled, cooperatively optimize the initial collision avoidance strategies of each ship to be controlled based on a game theory model.

[0014] In some possible implementation manners, adjusting the navigation state of the ship to be controlled according to the ship collision avoidance strategy and preset collision avoidance rules until there is no collision risk and the resumption of navigation conditions are met, including: Perform Kalman filtering on the predicted relative trajectory to determine whether there is a collision risk; Perform PID control on the speed, course, and turning angle of the ship to be controlled according to the ship collision avoidance strategy and preset collision avoidance rules until there is no collision risk and the preset resumption of navigation conditions are met.

[0015] In a second aspect, the present invention provides an intelligent ship collision avoidance device, including: A collision avoidance modeling unit, configured to construct a water area environment model according to the real-time navigation data and water area environment data of the ship to be controlled, and determine real-time collision avoidance requirements according to the water area environment model; A motion conflict analysis unit, configured to determine a predicted relative trajectory according to the real-time navigation data of the ship to be controlled and surrounding ships and the water area environment model, and determine the motion conflict radius of each surrounding ship according to the predicted relative trajectory; A collision avoidance strategy optimization unit, configured to generate and optimize a collision avoidance strategy according to the motion conflict radius and real-time collision avoidance requirements to obtain a ship collision avoidance strategy; A ship collision avoidance control unit, configured to adjust the navigation state of the ship to be controlled according to the ship collision avoidance strategy and preset collision avoidance rules until there is no collision risk and the resumption of navigation conditions are met.

[0016] The beneficial effects of adopting the above embodiments are as follows: The intelligent ship collision avoidance method provided by the present invention determines the predicted relative trajectory between ships through real-time navigation data and the water area environment model. It can consider in real time during the collision avoidance planning process the channel changes of surrounding ships affected by restricted waters and the collision avoidance conflicts of the ship to be controlled, and generate a motion conflict radius based on the predicted relative trajectory. It can also consider in real time during the collision avoidance planning process the collision avoidance actions of surrounding ships in restricted waters and the collision avoidance conflicts of the ship to be controlled, so as to effectively combine dynamic change information during the intelligent ship collision avoidance process. Furthermore, the present invention determines the real-time collision avoidance requirements through the water area environment model, can obtain the static obstacle information of the ship, and optimizes the generation of collision avoidance strategies by combining the real-time collision avoidance requirements and the motion conflict radius. It can comprehensively consider the static obstacle information and the dynamic change information of surrounding ships in the collision avoidance decision-making, improve the accuracy of the collision avoidance decision-making, and reduce the collision risk. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those skilled in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0018] Figure 1 It is a schematic flowchart of an embodiment of the intelligent ship collision avoidance method provided by the present invention; Figure 2 It is a schematic flowchart of constructing the water area environment model in the embodiment of the present invention; Figure 3 It is a schematic flowchart of determining the predicted relative trajectory in the embodiment of the present invention; Figure 4 It is a schematic flowchart of trajectory correction in the embodiment of the present invention; Figure 5 It is a schematic flowchart of optimizing the generation of collision avoidance strategies in the embodiment of the present invention; Figure 6 It is a schematic flowchart of ship collision avoidance control in the embodiment of the present invention; Figure 7 It is a schematic structural diagram of an embodiment of the intelligent ship collision avoidance device provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0020] It should be understood that the schematic drawings are not drawn to scale. The flowcharts used in the present invention illustrate operations implemented according to some embodiments of the present invention. It should be understood that the operations in the flowchart may not be implemented in sequence, and steps without logical context relationships may be reversed or implemented simultaneously. In addition, those skilled in the art can add one or more other operations to the flowchart or remove one or more operations from the flowchart under the guidance of the content of the present invention. Some of the block diagrams shown in the drawings are functional entities, and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor systems and / or microcontroller systems.

[0021] The descriptions such as "first" and "second" involved in the embodiments of the present invention are only for descriptive purposes, and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Therefore, the technical features defined with "first" and "second" may explicitly or implicitly include at least one such feature.

[0022] Referring to "embodiments" herein means that the specific features, structures, or characteristics described in connection with the embodiments can be included in at least one embodiment of the present invention. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0023] The present invention provides an intelligent ship collision avoidance method and device, which will be described separately below.

[0024] Figure 1 is a schematic flowchart of an embodiment of the intelligent ship collision avoidance method provided by the present invention, as Figure 1 shown, the intelligent ship collision avoidance method includes: S101. Construct a water area environment model according to the real-time navigation data and water area environment data of the ship to be controlled, and determine the real-time collision avoidance requirement according to the water area environment model; Among them, the real-time navigation data includes data such as the current position information, speed, and course of the ship to be controlled, and the water area environment data includes information such as channel planning, obstacles, and ocean currents. The water area environment model constructed according to the real-time navigation data and water area environment data integrates the environmental information, channel information, and corresponding constraint conditions required for collision avoidance in the current water area, and determines the real-time collision avoidance requirement of the ship to be controlled in the current water area based on this.

[0025] S102. Determine the predicted relative trajectory based on the real-time navigation data of the ship to be controlled and the surrounding ships and the water area environment model, and determine the motion conflict radius of each surrounding ship according to the predicted relative trajectory; Among them, in order to ensure the effectiveness of the collision avoidance strategy, the embodiment also considers the dynamic transformation information of the surrounding ships. The embodiment can deduce the trajectory through the real-time navigation data of the ship to be controlled and the surrounding ships, and correct it in combination with the water area environment model, so as to obtain the predicted relative trajectory of the ships in the future time period. According to the obtained predicted relative trajectory, it can be judged whether there is a collision risk, and the motion conflict radius of each surrounding ship with a collision risk relative to the ship to be controlled can be determined.

[0026] It should be noted that the motion conflict radius refers to the radius of the motion area required to execute the collision avoidance action based on the speed of the surrounding ship in the future time period when a collision may occur. Considering the motion conflict radius in the optimization of the collision avoidance strategy can avoid the collision risk when the ships execute the collision avoidance action.

[0027] S103. Generate and optimize the collision avoidance strategy according to the motion conflict radius and the real-time collision avoidance requirements to obtain the ship collision avoidance strategy; Among them, the initial ship collision avoidance strategy in the embodiment can be directly determined according to the real-time collision avoidance requirements, and then the collision avoidance strategy is optimized by the method of reinforcement learning according to the motion conflict radius and the real-time collision avoidance requirements to obtain the ship collision avoidance strategy. Among them, the reinforcement learning can be the deep deterministic policy gradient method, or other algorithms can be used according to the application requirements.

[0028] S104. Adjust the navigation state of the ship to be controlled according to the ship collision avoidance strategy and the preset collision avoidance rules until there is no collision risk and the resumption of navigation conditions are met.

[0029] Among them, the ship collision avoidance strategy can include but is not limited to various strategies such as decelerating, changing the navigation, or changing the course. The preset collision avoidance rules include the rules that ships need to abide by for collision avoidance under the International Regulations for Preventing Collisions at Sea. Ships need to abide by the preset collision avoidance rules during the process of executing the collision avoidance strategy. When the ship to be controlled executes the collision avoidance operation, it needs to judge that there is no collision risk and meet the resumption of navigation conditions stipulated by the International Regulations for Preventing Collisions at Sea before it can exit the collision avoidance process and resume normal navigation.

[0030] Compared with the prior art, the intelligent ship collision avoidance method provided by the present invention can determine the predicted relative trajectories between ships through real-time navigation data and a water area environment model. It can consider in real time during the collision avoidance planning process the changes in the waterway affected by the restricted water area of surrounding ships and the collision avoidance conflicts of the ship to be controlled, and generate a motion conflict radius based on the predicted relative trajectories. It can also consider in real time during the collision avoidance planning process the collision avoidance actions of surrounding ships in the restricted water area and the collision avoidance conflicts of the ship to be controlled, thus effectively combining dynamic change information during the intelligent ship collision avoidance process. Furthermore, the present invention can determine the real-time collision avoidance requirements through the water area environment model, obtain the static obstacle information of the ship, and optimize the generation of collision avoidance strategies by combining the real-time collision avoidance requirements and the motion conflict radius. It can comprehensively consider the static obstacle information and the dynamic change information of surrounding ships in the collision avoidance decision-making, improve the accuracy of the collision avoidance decision-making, and reduce the collision risk.

[0031] In some embodiments of the present invention, the real-time navigation data includes navigation position information, and the water area environment data includes waterway information and marine environment information. Figure 2 It is a schematic flow chart of constructing a water area environment model according to an embodiment of the present invention. As Figure 2 shown, constructing a water area environment model based on the real-time navigation data and water area environment data of the ship to be controlled includes: S201. Perform artificial potential field method modeling according to the navigation position information and waterway information to obtain a waterway centerline gravitational function and a boundary repulsive function; S202. Perform water flow field modeling according to the marine environment information to obtain an environmental water flow field; S203. Construct a water area environment model according to the waterway centerline gravitational function, the boundary repulsive function, and the environmental water flow field.

[0032] Specifically, in the embodiment, the artificial potential field method is used to model the waterway in the water area. Among them, the waterway centerline gravitational function can be expressed as:

[0033] Among them, represents the proportional gain function of the waterway centerline gravitational potential field, is the waterway width, is the distance between the ship itself and the left boundary line of the selected waterway.

[0034] The boundary repulsive function can be expressed as:

[0035] Among them, represents the proportional gain coefficient of the boundary potential field.

[0036] For the water flow field, in the embodiments, calculations are performed and presented in the form of a mathematical model based on the ocean current information in the ocean environment information. For example, the NS equation, Euler equation, etc.

[0037] Then, based on the constructed channel potential field and water flow field, the embodiments can establish a water area environment model for the navigation area of the ship to be controlled.

[0038] In some embodiments of the present invention, determining the real-time collision avoidance requirement according to the water area environment model includes: Determining the collision constraint according to the water area environment model, and determining the real-time collision avoidance requirement according to the collision constraint.

[0039] Specifically, based on the constructed water area environment model, the embodiments can analyze the collision constraint in the current situation by combining the situation analysis technology, and determine the real-time collision avoidance requirement of the ship to be controlled in the subsequent navigation in combination with the real-time navigation situation of the ship.

[0040] It should be noted that during the navigation of the ship, it is necessary to obtain and update the real-time navigation data of the ship and the water area environment data around it in real time, and update the environment model regularly, so as to adjust the collision avoidance requirement in real time.

[0041] In some embodiments of the present invention, Figure 3 is a schematic flow chart of determining the predicted relative trajectory according to an embodiment of the present invention. As Figure 3 shown, determining the predicted relative trajectory according to the real-time navigation data of the ship to be controlled and the surrounding ships and the water area environment model includes: S301. Performing kinematic model analysis on the real-time navigation data of the ship to be controlled and the surrounding ships to obtain the relative motion trajectory; S302. Performing trajectory correction on the relative motion trajectory according to the water area environment model to obtain the predicted relative trajectory; Among them, the real-time navigation data includes static data and dynamic data. The static data includes ship dimensions, ship steering performance, and ship power performance, and the dynamic data includes real-time positioning data.

[0042] Specifically, the embodiments query the static data of the ships through the AIS system, ship registration information or database of each ship, such as ship dimensions (ship length and width), ship steering performance, and ship power performance (maximum acceleration and maximum speed), and collect dynamic data (ship position change signal, dashboard monitoring signal) through the AIS system, radar system, optical sensor, and autopilot system.

[0043] The embodiments calculate the speeds and headings of each ship in real time based on dynamic data, and considering the accelerations and steering performances of each ship, determine the relative positions, speeds and headings of each ship, and calculate the dynamic trajectories of each ship through a kinematic model to predict the relative motion trajectories of each ship within a preset time period. At the same time, considering that the water area environment will not only affect the navigation of the ship to be controlled, but also affect the navigation of other ships. Therefore, the embodiments also perform trajectory correction on the relative motion trajectory according to the water area environment model to obtain the final predicted relative trajectory.

[0044] In some embodiments of the present invention, the water area environment model includes a waterway sub-model and a water flow field sub-model. Figure 4 It is a schematic flow chart of the trajectory correction of the embodiments of the present invention. As Figure 4 shown, performing trajectory correction on the relative motion trajectory according to the water area environment model to obtain the predicted relative trajectory includes: S401. Perform hydrodynamic analysis on surrounding ships according to the water flow field sub-model to obtain flow field correction parameters, and perform trajectory correction on the relative motion trajectory according to the flow field correction parameters to obtain a preliminary corrected trajectory; S402. Perform waterway analysis on surrounding ships according to the waterway sub-model to obtain waterway correction parameters, and perform trajectory correction on the preliminary corrected trajectory according to the waterway correction parameters to obtain the predicted relative trajectory.

[0045] Preferably, the process of performing trajectory correction according to the water area environment model can be divided into two parts. The first part performs hydrodynamic analysis through the water flow field sub-model to obtain flow field correction parameters, that is, analyzes the influence of factors such as ocean currents in the current water area on different ships, and corrects the trajectory accordingly. The second part is to perform waterway analysis through the waterway sub-model to obtain waterway correction parameters, that is, in restricted waters, to ensure safety, each ship will try to ensure that it sails within a predetermined waterway, and the waterway factor is also one of the key factors for trajectory correction.

[0046] It should be understood that the above is only part of the trajectory correction methods, and it does not limit that the present application can only adopt the above methods. According to actual application scenarios, there can be other trajectory correction methods. For example, in a scenario with a high sea wind level, it is also necessary to analyze the influence of wind on different ships.

[0047] In some embodiments of the present invention, determining the motion conflict radius of each surrounding ship according to the predicted relative trajectory includes: Performing kinematic analysis on each surrounding ship according to the predicted relative trajectory and preset collision avoidance rules to obtain the motion conflict radius of each surrounding ship.

[0048] Specifically, after determining the predicted relative trajectory, the embodiments use a kinematic model in combination with preset collision avoidance rules to calculate the motion conflict radius of each surrounding ship in real time:

[0049] Among them, represents the motion conflict radius, represents the relative speed, refers to the preset time period, represents the turning radius, and the preset collision avoidance rules are the collision avoidance rules set according to the International Regulations for Preventing Collisions at Sea.

[0050] In some embodiments of the present invention, Figure 5 is a schematic flow chart for optimizing the generation of the collision avoidance strategy of the embodiment of the present invention, as Figure 5 shown, the collision avoidance strategy of the ship is obtained by optimizing the generation of the collision avoidance strategy according to the motion conflict radius and the real-time collision avoidance requirements, including: S501. Construct an initial collision avoidance strategy according to the real-time collision avoidance requirements; S502. Construct a reward function according to the motion conflict radius and the real-time collision avoidance requirements; S503. Optimize the deep deterministic policy gradient algorithm according to the initial collision avoidance strategy and the reward function to obtain the collision avoidance strategy of the ship; Among them, the reward function includes a collision avoidance reward function, a homeward reward function, and a boundary safety reward function.

[0051] Specifically, in the optimization of the collision avoidance strategy generation, the embodiment optimizes the initial collision avoidance strategy by using the DDPG (Deep Deterministic Policy Gradient) algorithm, and introduces an enhanced truncation function to optimize the convergence speed of the DDPG algorithm. The optimized algorithm formula is expressed as:

[0052] Among them, represents the state-action value under the current collision avoidance strategy, represents the current collision avoidance strategy, represents the target collision avoidance strategy, represents the discount factor, represents the immediate reward, is the policy function when using the collision avoidance strategy at that time.

[0053] Among them, the immediate reward is calculated according to the reward function, and the reward function includes a collision avoidance reward function, a homeward reward function, and a boundary safety reward function. The collision avoidance reward function determines the function value according to the collision avoidance result, the homeward reward function determines the function value according to the ship's navigation efficiency, and the boundary safety reward function can be dynamically adjusted based on the actual size of the ship itself.

[0054] In some embodiments of the present invention, the ship collision avoidance strategy is generated and optimized according to the motion conflict radius and real-time collision avoidance requirements, and further includes: When there are at least two ships to be controlled, the initial collision avoidance strategies of each ship to be controlled are collaboratively optimized based on the game theory model.

[0055] Preferably, when multiple ships to be controlled all require collision avoidance strategy planning, the embodiment introduces a collaborative collision avoidance strategy, uses the V2X communication technology between ships to exchange navigation states and target information in real time, and performs collaborative optimization based on the game theory model.

[0056] In some embodiments of the present invention, Figure 6 is a schematic flowchart of the ship collision avoidance control according to the embodiment of the present invention. As Figure 6 shown, the navigation state of the ship to be controlled is adjusted according to the ship collision avoidance strategy and the preset collision avoidance rules until there is no collision risk and the resumption condition is met, including: S601. Perform Kalman filtering on the predicted relative trajectory to determine whether there is a collision risk; Among them, in judging the collision risk, the embodiment introduces a Kalman filtering algorithm based on time series to track the predicted relative trajectory and predict the possible collision time. The formula is expressed as:

[0057]

[0058] Among them, represents the predicted state, represents the predicted covariance, represents the process noise, and represent the dynamic matrix, represents the control input, represents the time step.

[0059] S602. Perform PID control on the ship speed, heading and turning angle of the ship to be controlled according to the ship collision avoidance strategy and the preset collision avoidance rules until there is no collision risk and the preset resumption condition is met.

[0060] Among them, when there is a collision risk, the embodiment adjusts the ship speed, heading and turning angle of the ship through the steering gear and the recommendation system according to the ship collision avoidance strategy and the preset collision avoidance rules. In this process, the embodiment optimizes the adjustment process through the PID control algorithm:

[0061] Among them, represents the control input, is the error, , and respectively represent the proportional-integral and derivative gains.

[0062] In addition, the embodiment further includes that whenever the ship's speed, heading or rotational angular velocity changes, the real-time state of the ship is automatically updated, and the relative position and speed with surrounding ships are recalculated; calculate the effect of the current collision avoidance strategy and return a reward value. If the reward value shows that the collision avoidance effect is not ideal, then return and optimize the collision avoidance strategy again.

[0063] After collision avoidance control, when the ship is judged to have no collision risk and meets the preset resumption of navigation conditions set according to the International Regulations for Preventing Collisions at Sea, a fuzzy control algorithm is used to adjust the deviation between the current navigation state of the ship and the original heading, and the normal navigation state is restored.

[0064] Otherwise, restart the collision avoidance program until there is no collision risk and the preset resumption of navigation conditions are met.

[0065] In summary, the intelligent ship collision avoidance method provided by the present invention can determine the predicted relative trajectory between ships through real-time navigation data and the water area environment model, can consider in real time during the collision avoidance planning process the channel changes of surrounding ships affected by restricted waters and the collision avoidance conflicts of the ship to be controlled, and can generate a movement conflict radius according to the predicted relative trajectory, and can consider in real time during the collision avoidance planning process the collision avoidance actions of surrounding ships in restricted waters and the collision avoidance conflicts of the ship to be controlled, so as to effectively combine dynamic change information in the intelligent ship collision avoidance process; Furthermore, the present invention can determine the real-time collision avoidance requirements through the water area environment model, can obtain the static obstacle information of the ship, and can generate and optimize the collision avoidance strategy by combining the real-time collision avoidance requirements and the movement conflict radius, and can comprehensively consider the static obstacle information and the dynamic change information of surrounding ships in the collision avoidance decision-making, improve the accuracy of the collision avoidance decision-making, and reduce the collision risk.

[0066] In order to better implement the intelligent ship collision avoidance method in the embodiments of the present invention, correspondingly, based on the intelligent ship collision avoidance method, the embodiments of the present invention further provide an intelligent ship collision avoidance device, as Figure 7 shown, the intelligent ship collision avoidance device 700 includes: A collision avoidance modeling unit 701, configured to construct a water area environment model according to the real-time navigation data and water area environment data of the ship to be controlled, and determine the real-time collision avoidance requirements according to the water area environment model; A movement conflict analysis unit 702, configured to determine the predicted relative trajectory according to the real-time navigation data of the ship to be controlled and surrounding ships and the water area environment model, and determine the movement conflict radius of each surrounding ship according to the predicted relative trajectory; A collision avoidance strategy optimization unit 703, configured to generate and optimize the collision avoidance strategy according to the movement conflict radius and the real-time collision avoidance requirements to obtain the ship collision avoidance strategy; The ship collision avoidance control unit 704 is configured to adjust the navigation state of the ship to be controlled according to the ship collision avoidance strategy and the preset collision avoidance rules until there is no collision risk and the conditions for resuming navigation are met.

[0067] The intelligent ship collision avoidance device 700 provided by the above embodiment can implement the technical solutions described in the above embodiment of the intelligent ship collision avoidance method. The specific implementation principles of the above modules or units can be referred to the corresponding content in the above embodiment of the intelligent ship collision avoidance method, which will not be elaborated here.

[0068] The intelligent ship collision avoidance method and device provided by the present invention have been introduced in detail above. Specific examples are used in this article to elaborate on the principles and implementation manners of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. An intelligent ship collision avoidance method, characterized in that, Including: Construct a water area environment model based on the real-time navigation data and water area environment data of the ship to be controlled, and determine the real-time collision avoidance requirements according to the water area environment model; Determine the predicted relative trajectory according to the real-time navigation data of the ship to be controlled and surrounding ships and the water area environment model, and determine the movement conflict radius of each surrounding ship according to the predicted relative trajectory; Generate and optimize a collision avoidance strategy according to the movement conflict radius and the real-time collision avoidance requirements to obtain a ship collision avoidance strategy; Adjust the navigation state of the ship to be controlled according to the ship collision avoidance strategy and preset collision avoidance rules until there is no collision risk and the resumption of navigation conditions are met.

2. The intelligent ship collision avoidance method according to claim 1, wherein, The real-time navigation data includes navigation position information, and the water area environment data includes channel information and marine environment information. The construction of the water area environment model according to the real-time navigation data and water area environment data of the ship to be controlled includes: Perform artificial potential field method modeling according to the navigation position information and the channel information to obtain a channel center line gravitational function and a boundary repulsive function; Perform water flow field modeling according to the marine environment information to obtain an environmental water flow field; Construct a water area environment model according to the channel center line gravitational function, the boundary repulsive function and the environmental water flow field.

3. The intelligent ship collision avoidance method according to claim 1, wherein, The determination of the real-time collision avoidance requirements according to the water area environment model includes: Determine the collision constraint according to the water area environment model, and determine the real-time collision avoidance requirements according to the collision constraint.

4. The intelligent ship collision avoidance method according to claim 1, wherein The determination of the predicted relative trajectory according to the real-time navigation data of the ship to be controlled and surrounding ships and the water area environment model includes: Perform kinematic model analysis on the real-time navigation data of the ship to be controlled and surrounding ships to obtain a relative movement trajectory; Perform trajectory correction on the relative movement trajectory according to the water area environment model to obtain a predicted relative trajectory; Among them, the real-time navigation data includes static data and dynamic data. The static data includes ship size, ship steering performance and ship power performance, and the dynamic data includes real-time positioning data.

5. The intelligent ship collision avoidance method according to claim 4, wherein, The water area environment model includes a channel sub-model and a water flow field sub-model. The performing of trajectory correction on the relative movement trajectory according to the water area environment model to obtain a predicted relative trajectory includes: Perform hydrodynamic analysis on the surrounding ships according to the water flow field sub-model to obtain a flow field correction parameter, and perform trajectory correction on the relative movement trajectory according to the flow field correction parameter to obtain a preliminary corrected trajectory; Perform channel analysis on the surrounding ships according to the channel sub-model to obtain a channel correction parameter, and perform trajectory correction on the preliminary corrected trajectory according to the channel correction parameter to obtain a predicted relative trajectory.

6. The intelligent ship collision avoidance method according to claim 1, characterized in that, The determination of the movement conflict radius of each surrounding ship according to the predicted relative trajectory includes: Perform kinematic analysis according to the predicted relative trajectory and preset collision avoidance rules to obtain the movement conflict radius of each surrounding ship.

7. The intelligent ship collision avoidance method according to claim 1, characterized in that The generation and optimization of a collision avoidance strategy according to the movement conflict radius and the real-time collision avoidance requirements to obtain a ship collision avoidance strategy includes: Construct an initial collision avoidance strategy according to the real-time collision avoidance requirements; Construct a reward function according to the movement conflict radius and the real-time collision avoidance requirements; The deep deterministic policy gradient algorithm is optimized according to the initial collision avoidance strategy and the reward function to obtain a ship collision avoidance strategy; Among them, the reward function includes a collision avoidance reward function, a homeward navigation reward function, and a boundary safety reward function.

8. The intelligent ship collision avoidance method according to claim 7, wherein The generation and optimization of the collision avoidance strategy according to the motion conflict radius and the real-time collision avoidance requirement to obtain a ship collision avoidance strategy further includes: If there are at least two ships to be controlled, the initial collision avoidance strategies of the ships to be controlled are collaboratively optimized based on the game theory model.

9. The intelligent ship collision avoidance method according to claim 1, wherein, Adjusting the navigation state of the ship to be controlled according to the ship collision avoidance strategy and the preset collision avoidance rules until there is no collision risk and the resumption of navigation conditions are met, including: Performing Kalman filtering on the predicted relative trajectory to determine whether there is a collision risk; Performing PID control on the ship speed, heading, and turning angle of the ship to be controlled according to the ship collision avoidance strategy and the preset collision avoidance rules until there is no collision risk and the preset resumption of navigation conditions are met.

10. An intelligent ship collision avoidance device, characterized in that, Including: A collision avoidance modeling unit, configured to construct a water area environment model according to the real-time navigation data of the ship to be controlled and the water area environment data, and determine the real-time collision avoidance requirement according to the water area environment model; A motion conflict analysis unit, configured to determine a predicted relative trajectory according to the real-time navigation data of the ship to be controlled and surrounding ships and the water area environment model, and determine the motion conflict radius of each surrounding ship according to the predicted relative trajectory; A collision avoidance strategy optimization unit, configured to generate and optimize a collision avoidance strategy according to the motion conflict radius and the real-time collision avoidance requirement to obtain a ship collision avoidance strategy; A ship collision avoidance control unit, configured to adjust the navigation state of the ship to be controlled according to the ship collision avoidance strategy and the preset collision avoidance rules until there is no collision risk and the resumption of navigation conditions are met.

Images