Intelligent ship collision avoidance decision-making method and system

Abstract

The invention discloses a ship intelligent collision avoidance decision-making method and system. The method comprises the following steps: acquiring ship operation data, ship target encounter parameters and ship navigation historical data; establishing a risk assessment model based on an improved artificial potential field algorithm according to the ship operation data, the ship target encounter parameters and the ship navigation historical data; and calculating a potential field force vector based on the risk assessment model, and generating a short-term collision avoidance instruction and a long-term path planning strategy in real time according to the potential field force vector. According to the method and the system provided by the invention, the problems of unreasonable collision avoidance path and poor environmental adaptability of the ship in a complex navigation environment can be solved.

Description

Technical Field

[0001] This invention relates to the field of ship navigation technology, and in particular to a ship intelligent collision avoidance decision-making method and system. Background Technology

[0002] With the rapid development of the global shipping industry, the maritime traffic environment is becoming increasingly complex, and the frequency of encounters between ships has increased significantly. Especially in high-density navigation areas such as ports, narrow waterways, and channel intersections, ship collision avoidance safety has become a key focus of navigation safety management. Traditional ship collision avoidance systems largely rely on Automatic Identification Systems (AIS) and radar information, calculating the closest encounter distance (CPA) and closest encounter time (TCPA) to assist crew members in assessing risks. Furthermore, existing automatic collision avoidance algorithms applied to ship collision avoidance systems have certain limitations in practical applications, easily leading to unreasonable collision avoidance paths and poor environmental adaptability. Summary of the Invention

[0003] In view of this, the purpose of this invention is to provide a ship intelligent collision avoidance decision-making method and system, which can solve the problems of unreasonable collision avoidance paths and poor environmental adaptability of ships in complex navigation environments.

[0004] In a first aspect, embodiments of the present invention provide a ship intelligent collision avoidance decision-making method, including:

[0005] Acquire ship operation data, ship-target encounter parameters, and ship navigation history data;

[0006] A risk assessment model based on an improved artificial potential field algorithm is established based on ship operation data, ship-target encounter parameters, and ship navigation history data.

[0007] The potential field composite force vector is calculated based on the risk assessment model, and short-term collision avoidance commands and long-term path planning strategies are generated in real time based on the potential field composite force vector.

[0008] Furthermore, the ship operation data includes ship position data, ship environmental data, ship target encounter parameters including nearest encounter distance, nearest encounter time, and visibility, and ship navigation history data including the ship's historical safe track data, channel collision accident hotspot area data, and the ship's operation records.

[0009] Furthermore, after collecting ship operation data, the ship operation data is preprocessed.

[0010] Furthermore, the expression for the risk assessment model based on the improved artificial potential field algorithm is as follows:

[0011] U(p)=U tar (p)+U hot (p)+U hist (p)

[0012] In the formula, U(p) represents the total power field of the vessel. tar (p) represents the target repulsive force, U hot (p) represents the repulsive force in the hotspot area of ​​channel collision accidents, U hist (p) represents the gravity of the historical safe course, and p is the coordinate vector of the ship.

[0013] Furthermore, the target repulsive force U tar The expression for (p) is:

[0014]

[0015] In the formula, i represents each target vessel; W tar The target repulsion force weight is used to control the repulsion force intensity for collision avoidance by the target vessel; μ i σ is the linear normalization exponent; tar The radius of influence of the target repulsive force is used to control the range of action of the target repulsive force; c i Let be the coordinate vector of the nearest point corresponding to the i-th target vessel;

[0016] Repulsive force U in hotspot areas of channel collision accidents hot The expression for (p) is:

[0017]

[0018] In the formula, j represents each channel collision hotspot area; W hot The repulsive force weight for hotspot areas of channel collision accidents is used to control the repulsive force intensity in these areas. σ is the hazard coefficient of the repulsive force in the hotspot area of ​​a collision accident in the j-th waterway, used to represent the hazard level of the hotspot area of ​​a collision accident in the waterway; hot h represents the radius of influence of the repulsive force in the hotspot area of ​​a channel collision accident, used to control the range of action of the repulsive force in the hotspot area of ​​a channel collision accident; j This is the j-th channel collision hotspot area;

[0019] Historical safe flight path gravity U hist The expression for (p) is:

[0020]

[0021] In the formula, k represents each historical safe flight path; W hist η is the gravitational weight for the historical safe course, used to control the attraction strength for the vessel to return to its historical safe course; k d represents the weight of the k-th historical safe flight path, indicating its importance; k(p) represents the shortest distance between this vessel and the kth historical safe track; σ hist The attraction radius of a historical safe flight path is used to control the attraction bandwidth of that path.

[0022] Furthermore, the linear normalization exponent μ i The expression is:

[0023]

[0024] In the formula, CPA is the nearest encounter distance; TCPA is the nearest encounter time; λ is the hazard coefficient, used to reflect the degree of danger of the nearest encounter distance, the nearest encounter time, and visibility; α d α is a sensitivity coefficient representing the nearest encounter distance, used to control the effect of the nearest encounter distance on the attenuation of the danger level; t d is a sensitivity coefficient representing the nearest encounter distance, used to control the impact of the nearest encounter time on the attenuation of the degree of danger; safe Safety distance, a distance indicator used to measure the risk of ship collision; T ref This is the normalized time of the most recent meeting time.

[0025] Furthermore, the expression for the risk factor λ is:

[0026] λ = 0.4λ CPA +0.4λ TCPA +0.2λ V

[0027] In the formula, λ CPA λ is the nearest encounter distance score; TCPA The score for the most recent encounter time; λ V Visibility score;

[0028] The nearest encounter will be the distance score λ. CPA The most recent time score λ TCPA Visibility score λ V The values ​​are all set through risk status, which includes low risk, medium risk and high risk.

[0029] Secondly, embodiments of the present invention provide a ship intelligent collision avoidance decision-making system, comprising:

[0030] Data acquisition module: used to acquire ship operation data, ship-target encounter parameters, and ship navigation history data.

[0031] Risk assessment module: used to build a risk assessment model based on an improved artificial potential field algorithm, based on acquired ship operation data, ship-target encounter parameters, and ship navigation history data.

[0032] Decision output module: used to calculate the potential field composite force vector based on the risk assessment model, and generate short-term collision avoidance instructions and long-term path planning strategies in real time based on the potential field composite force vector.

[0033] The embodiments of the present invention bring the following beneficial effects: The intelligent collision avoidance decision-making method and system for ships provided in this embodiment establishes a risk assessment model by using ship operation data, ship target encounter parameters, and ship navigation history data. This can improve the timeliness and completeness of collision avoidance information, enhance the ship's collision avoidance capability in complex sea conditions, and thus improve environmental adaptability. In addition, the risk assessment model integrates an improved artificial potential field algorithm, which improves the continuity and stability of the collision avoidance path, avoids the local minima problem that traditional algorithms are prone to fall into, and thus enhances the rationality of the collision avoidance path.

[0034] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0035] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0036] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0037] Figure 1 This is a flowchart illustrating a ship intelligent collision avoidance decision-making method provided in an embodiment of the present invention. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] To facilitate understanding of this embodiment, in conjunction with Figure 1 This invention provides a detailed description of a ship intelligent collision avoidance decision-making method and system disclosed in the embodiments of the present invention.

[0040] Example 1

[0041] A ship intelligent collision avoidance decision-making method includes:

[0042] S1. Acquire ship operation data, ship target encounter parameters, and ship navigation history data.

[0043] The ship operation data includes ship position data and ship environment data. The ship target encounter parameters include the nearest encounter distance, the nearest encounter time, and visibility. The ship navigation history data includes the ship's historical safe track data, channel collision accident hotspot area data, and the ship's operation records.

[0044] Specifically, ship operating parameters are acquired in real time through ship sensor data. Ship position data includes AIS target information, navigation radar information, and GPS information, while ship environmental information includes wind speed and current speed. AIS target information and navigation radar information are used to calculate the target ship's course and speed, while GPS information is used to calculate the ship's own course and speed.

[0045] In a preferred embodiment, after collecting ship operation data, the ship operation data is preprocessed. This preprocessing includes: time alignment of the ship operation data; coordinate transformation of the ship operation data; unit standardization of the ship operation data; and handling of null and duplicate values ​​in the ship operation data.

[0046] Specifically, the process involves time alignment of ship operation data, including converting the timestamps of all ship operation data to a standard time coordinate system, which can be the UTC coordinate system; coordinate transformation of ship operation data, including converting position data from different coordinate systems in AIS target information, navigation radar information, and GPS information into position data with a unified coordinate standard; unit standardization of ship operation data, including unifying the units of physical quantities in ship operation data; and handling of missing and duplicate values ​​in ship operation data, including preliminary imputation or tagging of missing values ​​in ship operation data and removing completely duplicate ship operation data.

[0047] S2. Establish a risk assessment model based on an improved artificial potential field algorithm, using ship operation data, ship target encounter parameters, and ship navigation history data.

[0048] The expression for the risk assessment model based on the improved artificial potential field algorithm is as follows:

[0049] U(p)=U tar (p)+U hot (p)+U hist (p)

[0050] In the formula, U(p) represents the total power field of the vessel. tar (p) represents the target repulsive force, U hot (p) represents the repulsive force in the hotspot area of ​​channel collision accidents, U hist (p) represents the gravity of the historical safe course, and p is the coordinate vector of the ship.

[0051] In this expression, the expression for p can be:

[0052] p = (2, y)

[0053] Specifically, the target repulsive force U tar (p) is used to represent the potential collision risks between all target vessels and this vessel, with target repulsion force U. tar (p) The target repulsive force U is calculated using the nearest encounter distance and the nearest encounter time. tar The expression for (p) is:

[0054]

[0055] In the formula, i represents each target vessel; W tar The target repulsion force weight is used to control the repulsion force intensity for collision avoidance by the target vessel; μ i σ is the linear normalization exponent; tar The radius of influence of the target repulsive force is used to control the range of action of the target repulsive force; c i Let be the coordinate vector of the nearest point corresponding to the i-th target ship.

[0056] In this expression, W tar The value range is 0.1 to 5.0, with a preferred value of 1.0; σ tar The value range is 0.5 to 3×d safe The preferred value is 1.5×d safe ;c i The expression can be:

[0057]

[0058] Repulsive force U in hotspot areas of channel collision accidents hot (p) Used to guide the vessel away from high-risk waters where collision hotspots are likely to occur, repulsive force U in the channel collision hotspot area. hot The expression for (p) is:

[0059]

[0060] In the formula, j represents each channel collision hotspot area; W hot The repulsive force weight for hotspot areas of channel collision accidents is used to control the repulsive force intensity in these areas. σ is the hazard coefficient of the repulsive force in the hotspot area of ​​a collision accident in the j-th waterway, used to represent the hazard level of the hotspot area of ​​a collision accident in the waterway; hot h represents the radius of influence of the repulsive force in the hotspot area of ​​a channel collision accident, used to control the range of action of the repulsive force in the hotspot area of ​​a channel collision accident; j This is the j-th channel collision hotspot area.

[0061] In this expression, W hot The value range is 0.1 to 3.0, with a preferred value of 0.8; The value range of σ is 0 to 1; hot The value range is 50–500 m, with a preferred value of 200 m; h j The expression can be:

[0062] h j =(x j y j )

[0063] Historical safe flight path gravity U hist (p) Used to guide this vessel back to its historical safe course, historical safe course gravity U hist The expression for (p) is:

[0064]

[0065] In the formula, k represents each historical safe flight path; W hist η is the gravitational weight for the historical safe course, used to control the attraction strength for the vessel to return to its historical safe course; k d represents the weight of the k-th historical safe flight path, indicating its importance; k (p) represents the shortest distance between this vessel and the kth historical safe track; σ hist The attraction radius of a historical safe flight path is used to control the attraction bandwidth of that path.

[0066] In this expression, W hist The value of η ranges from 0 to 2.0, with a preferred value of 0.6; k The value range of σ is 0 to 1, with a preferred range of 0.4 to 1; hist The value range is 100 to 1000m, and the preferred value is 2×d. safe .

[0067] Furthermore, in the target repulsive force U tar In the expression for (p), the linear normalization exponent μ i The expression is:

[0068]

[0069] In the formula, CPA is the nearest encounter distance; TCPA is the nearest encounter time; λ is the hazard coefficient, used to reflect the degree of danger of the nearest encounter distance, the nearest encounter time, and visibility; α d α is a sensitivity coefficient representing the nearest encounter distance, used to control the effect of the nearest encounter distance on the attenuation of the danger level; t d is a sensitivity coefficient representing the nearest encounter distance, used to control the impact of the nearest encounter time on the attenuation of the degree of danger; safe Safety distance, a distance indicator used to measure the risk of ship collision; T ref This is the normalized time of the most recent meeting time.

[0070] In this expression, λ ranges from 0 to 1; α d The value range is 0.5 to 5.0, with a preferred value of 1.0; α t The value range is 0.1 to 2.0, with a preferred value of 0.5; d safe The value range is 100–500m, with a preferred value of 150m; T ref The value range is 60 to 600s, with 300s being the preferred value.

[0071] Furthermore, in the linear normalization exponent μ i In the expression, the expression for the risk factor λ is:

[0072] λ = 0.4λ CPA +0.4λ TCPA +0.2λ V

[0073] In the formula, λ CPA λ is the nearest encounter distance score; TCPA The score for the most recent encounter time; λ V This represents the visibility score.

[0074] In this expression, the nearest encounter distance score is λ. CPA The most recent time score λ TCPA Visibility score λ V The values ​​are all set through risk status, which includes low risk, medium risk and high risk.

[0075] Among them, the values ​​of the nearest encounter distance score λCPA, the nearest encounter time score λTCPA, and the visibility score λV are all determined through risk status settings, including:

[0076] 1. When the nearest encounter distance (CPA) is greater than or equal to twice the overall length L of the vessel, the nearest encounter distance is considered to be in a risk-free state, and a nearest encounter distance score λ is set. CPA The value is 0;

[0077] If the nearest encounter distance (CPA) is greater than or equal to the vessel's overall length (L) but less than twice the vessel's overall length, then the nearest encounter distance is considered a potential risk condition, and a nearest encounter distance score (λ) is assigned. CPA The value is 0.5;

[0078] When the nearest encounter distance (CPA) is less than the overall length L of the vessel, the nearest encounter distance is considered a high-risk condition, and a nearest encounter distance score λ is set. CPA The value is 1.

[0079] 2. When the value of the most recent encounter time TCPA is greater than 300s, it is determined that the most recent encounter time is in a risk-free state, and the value of the most recent encounter time is set to 0.

[0080] When the nearest encounter time (TCPA) is greater than or equal to 120s and less than 300s, the nearest encounter distance is determined to be a potential risk state, and the value of the nearest encounter time is set to 0.5.

[0081] If the value of the most recent encounter time (TCPA) is less than 120 seconds, then the most recent encounter time is determined to be in a high-risk state, and the value of the most recent encounter time is set to 1.

[0082] 3. When the visibility V value is greater than or equal to 3 nautical miles, the visibility is determined to be in a risk-free state, and the visibility score is set to 0.

[0083] When the visibility V value is greater than or equal to 1 nautical mile and less than 3 nautical miles, the visibility is determined to be in a potential risk state, and the visibility score is set to 0.5.

[0084] When the visibility V value is less than 1 nautical mile, the visibility is determined to be in a high-risk state, and the visibility score is set to 1.

[0085] This embodiment provides Table 1, which can intuitively display the relationship between the nearest encounter distance score, the nearest encounter time score, the visibility score, and the risk status in the risk factor.

[0086] Table 1

[0087]

[0088]

[0089] As an example, assuming the vessel's length L is 50m, when the nearest encounter distance (CPA) is 40m (40m < 50m), the nearest encounter time (TCPA) is 100s (100s < 120s), and the visibility V is 2 nautical miles (1 nautical mile ≤ 2 nautical miles < 3 nautical miles),

[0090] The nearest encounter will be the distance score λ. CPA The value is 1, and the most recent time score λ is encountered. TCPA The value is 1, and the visibility score is λ. V The value is 0.5, therefore, the risk factor is:

[0091] λ=0.4×1+0.4×1+0.2×0.5=0.9.

[0092] S3. Calculate the potential field composite force vector based on the risk assessment model, and generate short-term collision avoidance instructions and long-term path planning strategies in real time based on the potential field composite force vector.

[0093] The potential field composite force vector represents the collision avoidance control force along the potential field gradient direction, indicating the collision avoidance direction of the vessel. The expression for the potential field composite force vector is:

[0094]

[0095] Short-term collision avoidance instructions are used to instruct the vessel to perform collision avoidance maneuvers within a short period of time. Short-term collision avoidance instructions include, but are not limited to, "turn 30° to starboard" and "reduce to 10 knots".

[0096] Furthermore, the potential field composite force vector can also generate a potential field thermal distribution, and the potential field thermal distribution and long-term path planning strategy can be superimposed and displayed on the electronic nautical chart.

[0097] Example 2

[0098] The present invention discloses a ship intelligent collision avoidance decision-making system, comprising:

[0099] Data acquisition module: used to acquire ship operation data, ship-target encounter parameters, and ship navigation history data.

[0100] Risk assessment module: used to establish a risk assessment model based on an improved artificial potential field algorithm, based on ship operation data, ship-target encounter parameters, and ship navigation history data.

[0101] Decision output module: used to calculate the potential field composite force vector based on the risk assessment model, and generate short-term collision avoidance instructions and long-term path planning strategies in real time based on the potential field composite force vector.

[0102] The decision output module has a human-computer interaction function. When the decision output module is manually operated, the short-term collision avoidance command is automatically frozen.

[0103] It should be noted that the intelligent collision avoidance decision-making system for ships disclosed in this embodiment can realize all the contents of the intelligent collision avoidance decision-making method for ships disclosed in Embodiment 1 of this application. Further details will not be elaborated here.

[0104] It should also be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0105] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for intelligent collision avoidance decision of a ship, characterized in that, The application relates to a ship collision risk assessment and avoidance method based on an improved artificial potential field algorithm. Ship operation data, ship target meeting parameters and ship navigation history data are acquired. A risk assessment model based on the improved artificial potential field algorithm is established according to the ship operation data, the ship target meeting parameters and the ship navigation history data. A potential field resultant force vector is calculated based on the risk assessment model, and short-term collision avoidance instructions and long-term path planning strategies are generated in real time according to the potential field resultant force vector.

2. The intelligent collision avoidance decision-making method for a ship according to claim 1, characterized in that, The ship operation data includes ship position data and ship environment data, the ship target meeting parameters include a closest point of approach distance, a closest point of approach time and visibility, and the ship navigation history data includes historical safe track data of the ship, channel collision accident hotspot area data and operation records of the ship.

3. The intelligent collision avoidance decision-making method for a ship according to claim 1, characterized in that, After the ship operation data is collected, the ship operation data is preprocessed.

4. The intelligent collision avoidance decision method for a ship according to claim 1, characterized in that, The expression of the risk assessment model based on the improved artificial potential field algorithm is as follows: U(p) = U tar (p) + U hot (p) + U hist (p) where U(p) is the total potential field of the own ship, U tar (p) is the target repulsive force, U hot (p) is the channel collision accident hotspot area repulsive force, U hist (p) is the historical safe track attractive force, and p is the coordinate vector of the own ship.

5. The intelligent collision avoidance decision method for a ship according to claim 4, characterized in that, Target repulsive force U tar The expression of (p) is: where i represents each target ship; W tar is a target repulsion weight, used to control the strength of the target ship collision avoidance repulsion; μ i is a linear normalization index; σ tar is a target repulsion influence radius, used to control the range of the target repulsion; c i is the closest point coordinate vector corresponding to the i-th target ship; Channel collision accident hot area repulsive force U hot The expression of (p) is: In the formula, j represents each channel collision accident hotspot area; W hot represents the repulsive force weight of the channel collision accident hotspot area, which is used to control the strength of the repulsive force of the channel collision accident hotspot area; represents the risk coefficient of the repulsive force of the jth channel collision accident hotspot area, which is used to represent the risk level of the channel collision accident hotspot area;σ hot represents the influence radius of the repulsive force of the channel collision accident hotspot area, which is used to control the range of the repulsive force of the channel collision accident hotspot area; h j represents the jth channel collision accident hotspot area; Historical safe path gravity U hist The expression for (p) is: In the formula, k is used to represent each historical safe track; W hist is the historical safe track gravity weight, used to control the attraction strength of the ship returning to the historical safe track. η k is the weight of the kth historical safe track, which is used to represent the importance of the historical safe track;d k (p) is the shortest distance between the current ship and the kth historical safe track;σ hist is the attractive radius of the historical safe track, which is used to control the attractive bandwidth of the historical safe track.

6. The intelligent collision avoidance decision method for a ship according to claim 5, characterized in that, Linear normalization exponent μ i The expression for μ is: In the formula, CPA is the closest point of approach distance; TCPA is the closest point of approach time; λ is a danger coefficient, used to reflect the danger degree of the closest point of approach distance, the closest point of approach time and the visibility; α d is a sensitive coefficient of the closest point of approach distance, used to control the influence of the closest point of approach distance on the danger degree attenuation; α t is a sensitive coefficient of the closest point of approach time, used to control the influence of the closest point of approach time on the danger degree attenuation; d safe is a safety distance, used to measure the distance index of the ship collision risk; T ref is a normalized time of the closest point of approach time.

7. The intelligent collision avoidance decision method for a ship according to claim 6, characterized in that, The expression of a danger coefficient lambda is as follows: λ = 0.4λ CPA +0.4λ TCPA +0.2λ V where λ CPA is a closest approach distance score; λ TCPA is a closest approach time score; λ V is a visibility score; a recent encounter distance score λ CPA a recent encounter time score λ TCPA a visibility score λ V The values of the scores are set by a risk state, which includes a low risk state, a medium risk state, and a high risk state.

8. A ship intelligent collision avoidance decision system, characterized in that, The application relates to a ship collision risk assessment and avoidance method based on an improved artificial potential field algorithm. A data acquisition module is used for acquiring ship operation data, ship target meeting parameters and ship navigation history data. A risk assessment module is used for establishing a risk assessment model based on the improved artificial potential field algorithm according to the acquired ship operation data, ship target meeting parameters and ship navigation history data. A decision output module is used for calculating a potential field resultant force vector based on the risk assessment model, and generating short-term collision avoidance instructions and long-term path planning strategies in real time according to the potential field resultant force vector.

Images