A negotiation-based ship cooperative collision avoidance method in a mixed navigation scene at sea

Abstract

The application provides a negotiation-based ship cooperative collision avoidance method in a mixed navigation scene at sea. The method comprises the following steps: S1, in the mixed navigation scene, determining negotiation participants, negotiation topics, reservation values and utility functions; S2, using a negotiation communication protocol based on the Speech Act theory, defining negotiation communication primitives and contents; S3, using a negotiation protocol based on turn-by-turn bidding, negotiating ships successively putting forward respective collision avoidance schemes; S4, autonomous ships using a negotiation strategy based on the Zeuthen strategy to decide whether to make concessions, driving the negotiation process and reaching a consensus; S5, autonomous ships applying a Bayesian learning mechanism to estimate the reservation value of a target ship and further fitting the utility function. The negotiation-based cooperative collision avoidance method provided by the application is suitable for complex marine environments where traditional ships and autonomous ships coexist, can effectively reduce the collision risk, and improve the safety and efficiency of marine traffic.

Description

Technical Field

[0001] This invention relates to the field of maritime mixed navigation scenarios and ship cooperative collision avoidance technology, and more particularly to a negotiation-based ship cooperative collision avoidance method in maritime mixed navigation scenarios. Background Technology

[0002] There is still significant room for development in the technological maturation, ship size, and fleet scale of autonomous vessels. For a considerable period, the massive technological foundation and market share of traditional vessels will remain unshaken; the coexistence of autonomous and traditional vessels is inevitable and long-term, creating a mixed maritime navigation scenario. Autonomous vessels exhibit significant differences in their level of autonomy and core decision-making algorithms. Furthermore, different operators vary considerably in their ship handling skills, understanding and adherence to rules, and navigation experience. Therefore, in this context, uncoordinated collision avoidance situations are highly likely to occur between vessels. Traditional collision avoidance methods, proven effective in single-vessel environments, may not be fully applicable in situations where autonomous and traditional vessels coexist. Collaborative collision avoidance methods, as a new direction of exploration, aim to improve collision avoidance effectiveness through information sharing and collaborative decision-making between vessels. In this approach, the intelligent computing and learning capabilities of autonomous vessels can be fully utilized, while traditional vessels only require simple information exchange, thus avoiding large-scale modifications to their operating systems.

[0003] Centralized collaborative collision avoidance methods require a central coordinator and lack widespread application scenarios. In contrast, under a distributed framework, each vessel can independently formulate its own collision avoidance decisions and then reach a consensus through communication, negotiation, and interaction to achieve collaborative collision avoidance, thereby improving collision avoidance efficiency and navigation safety. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a negotiation-based collaborative collision avoidance method for ships in mixed maritime navigation scenarios. The invention clarifies the negotiation participants and issues, defines and describes the retention value and utility functions, establishes a communication mechanism including negotiation communication primitives and content, clarifies the negotiation proposal method using a round-robin bidding protocol, and clarifies the application mechanism of the Zeuthen one-way concession strategy through maximum risk acceptance, enabling ships to ultimately reach a collision avoidance consensus. Furthermore, a Bayesian-based one-sided learning mechanism is introduced to enhance the negotiation capabilities of autonomous ships.

[0005] The technical means employed in this invention are as follows:

[0006] A negotiation-based cooperative collision avoidance method for ships in a mixed maritime navigation scenario includes:

[0007] S1. In a mixed navigation scenario, determine the participants, negotiation topics, retention value, and utility function in the negotiation;

[0008] S2. Adopt a negotiation communication protocol based on the Speech Act theory, and define the negotiation communication primitives and content;

[0009] S3. A negotiation agreement based on round-robin bidding is adopted, with the negotiating vessels submitting their respective collision avoidance plans in turn;

[0010] S4. Autonomous vessels adopt a negotiation strategy based on the Zeuthen strategy to decide whether to make concessions, drive the negotiation process and reach a consensus.

[0011] S5. Autonomous ships use a Bayesian learning mechanism to estimate the retention value of the target ship and further fit the performance function.

[0012] Further, step S1 specifically includes:

[0013] S11. Use the Automatic Identification System (AIS) to obtain the position, speed, and heading information of ships, calculate the nearest encounter distance and the nearest encounter time between ships, compare the nearest encounter distance and the nearest encounter time with a preset safety threshold, identify ship pairs that may collide, and determine them as participants in the negotiation.

[0014] S12. The topics of discussion include the probability of collision between vessels, the status of vessels, and the proposed collision avoidance actions.

[0015] S13. In accordance with the International Regulations for Preventing Collisions at Sea, determine the ship's responsibility for avoiding collisions, and abstract the responsibility for avoiding collisions into the amount of change in course, which is regarded as the ship's reserve value;

[0016] S14. Construct a utility function for autonomous ships based on safety considerations and fit it to the utility function of traditional ships.

[0017] Further, step S14 specifically includes:

[0018] S141. The Sigmoid function is used to map the collision risk to the range [0,1]. A higher utility value indicates a lower collision risk. The utility function is the remaining collision risk after the collision avoidance action, and includes the following calculation methods:

[0019]

[0020] CRI = CR D ×CR T

[0021]

[0022] Among them, CR S Indicates the risk of a spatial collision between the two ships; CR TThe time-based collision risk indicates the two ships' collision risk; CRI indicates the collision risk between the two ships. denoted by ; DCPA and TCPA represent the nearest encounter distance and nearest encounter time, respectively; d1 and d2 represent the upper and lower bounds of safe encounter in space, respectively; t1 and t2 represent the upper and lower bounds of safe encounter in time, respectively; λ determines the steepness of the transition.

[0023] S142. Fit the utility function of the conventional vessel to a linear function representing risk neutrality, and determine that the first proposal of the conventional vessel has the maximum utility value, with a utility value of 0.6 at the retention value.

[0024] Further, step S2 specifically includes:

[0025] S21. Construct a negotiation communication protocol based on the Speech Act theory, and encapsulate the information required to avoid collisions into a structured information set;

[0026] S22. Based on the constructed negotiation communication protocol, define negotiation communication primitives, including message reception, expressing disagreement, confirming acceptance, explicitly rejecting, requesting clarification, verifying information, sharing information, providing suggestions, and making strong demands;

[0027] S23. Based on the constructed negotiation communication protocol, define the communication content, including the ship's navigation information, encounter situation, and collision avoidance decision.

[0028] Furthermore, in step S23, the defined communication content consists of keywords and values, as shown in the following formula:

[0029] Data = $Primitive:Content$

[0030] Communication content=Keyword(Value).

[0031] Further, step S3 specifically includes:

[0032] S31. Construct a negotiation agreement based on alternating bids to simulate the "call-response" communication pattern commonly encountered in actual maritime collision avoidance.

[0033] S32. The negotiating vessels take turns proposing their respective collision avoidance plans, promoting a consensus on collision avoidance between the two parties through continuous adjustments.

[0034] Further, step S4 specifically includes:

[0035] S41. Autonomous vessels adopt the Zeuthen one-way concession strategy, deciding whether to make concessions based on their maximum risk tolerance.

[0036] S42. A ship shall begin with the option most favorable to itself and gradually make concessions to the option most favorable to the other party, and shall never back down.

[0037] Furthermore, in step S5, the autonomous vessel continuously updates its estimate of the target vessel's utility function using a Bayesian learning mechanism, improving the accuracy of the information required for negotiation. Specifically, this includes:

[0038] S51. The initial estimate of the retention value range of autonomous vessels compared to traditional vessels is defined as RVR = {r1, r2, ..., r n The probability estimate of the retention value of autonomous ships compared to traditional ships is defined as p(r). i The conditional probability of a proposal from a traditional ship under the autonomous ship assumption is defined as p(q). i |r i ), that is, each r i The following suggestion q i The probability of;

[0039] S52. Assuming the proposal from the traditional vessel is q1, the probability distribution of the value retained by the autonomous vessel relative to the traditional vessel is updated as follows:

[0040]

[0041] S53, The estimate of the retention value of autonomous vessels relative to conventional vessels is updated as follows:

[0042]

[0043] S54. In a round of negotiations, the autonomous vessel calculates and estimates the maximum risk tolerance of both parties. The party with the smaller value will modify its proposal and make concessions. Taking the autonomous vessel as an example, the calculation of its maximum risk tolerance in the i-th round is as follows:

[0044]

[0045] in, and Let represent the utility of autonomous vessels and traditional vessels under the autonomous vessel proposal in round i, respectively. and Let U(a) and U(c) represent the utility of the traditional vessel and the autonomous vessel respectively under the traditional vessel proposal in the i-th round, and let U(a) and U(c) represent their respective utility values ​​when the negotiation fails, which are 0.

[0046] Compared with the prior art, the present invention has the following advantages:

[0047] 1. The present invention provides a negotiation-based ship collaborative collision avoidance method in mixed navigation scenarios at sea, which effectively solves the problem of ship collaborative collision avoidance in mixed navigation scenarios, improves the current situation of autonomous ships actively avoiding collisions, realizes two-way communication and cooperation between autonomous ships and traditional ships, and jointly formulates collision avoidance plans. This not only avoids the occurrence of collision accidents, but also reduces the amount of change in the ship's course and the magnitude of deceleration, thereby improving navigation efficiency.

[0048] 2. The present invention provides a negotiation-based ship cooperative collision avoidance method in a mixed navigation scenario at sea. The proposed communication mechanism ensures the effectiveness and accuracy of information transmission, the negotiation protocol makes the negotiation process smoother and more efficient, and the negotiation strategy not only ensures the convergence of the negotiation process, but also balances the interests of both parties.

[0049] 3. The present invention provides a negotiation-based ship cooperative collision avoidance method in a mixed maritime navigation scenario. By applying a Bayesian learning mechanism, it can continuously update the estimate of the retention value of the target ship and fit the utility function of the target ship, thereby predicting its subsequent behavior, providing a reference for the negotiation process, and improving the negotiation efficiency and success rate.

[0050] Based on the above reasons, this invention can be widely applied in areas such as mixed navigation scenarios at sea and collaborative collision avoidance between ships. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the 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 based on these drawings without creative effort.

[0052] Figure 1 This is a schematic diagram illustrating the negotiation formation process of a negotiation-based ship cooperative collision avoidance method in a mixed maritime navigation scenario according to the present invention.

[0053] Figure 2 This is a schematic diagram of the negotiation agreement and negotiation process described in this invention.

[0054] Figure 3 This is the negotiation protocol based on round-robin bidding as described in this invention.

[0055] Figure 4 The application process of the Zeuthen strategy in ship collision avoidance negotiation as described in this invention.

[0056] Figure 5 The present invention relates to a one-sided learning model for ship collision avoidance negotiation based on the Zeuthen strategy. Detailed Implementation

[0057] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0058] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0059] This invention provides a negotiation-based cooperative collision avoidance method for ships in mixed maritime navigation scenarios, including:

[0060] S1. In a mixed navigation scenario, determine the participants, negotiation topics, retention value, and utility function in the negotiation;

[0061] S2. Adopt a negotiation communication protocol based on the Speech Act theory, and define the negotiation communication primitives and content;

[0062] S3. A negotiation agreement based on round-robin bidding is adopted, with the negotiating vessels submitting their respective collision avoidance plans in turn;

[0063] S4. Autonomous vessels adopt a negotiation strategy based on the Zeuthen strategy to decide whether to make concessions, drive the negotiation process and reach a consensus.

[0064] S5. Autonomous ships use a Bayesian learning mechanism to estimate the retention value of the target ship and further fit the performance function.

[0065] In a specific implementation, as a preferred embodiment of the present invention, step S1 specifically includes:

[0066] S11. Use the Automatic Identification System (AIS) to obtain the position, speed, and heading information of ships, calculate the nearest encounter distance (DCPA) and nearest encounter time (TCPA) between ships, compare the nearest encounter distance and nearest encounter time with preset safety thresholds, identify ship pairs that may collide, and determine them as negotiation participants.

[0067] S12. The topics of discussion include the probability of collision between vessels, the status of vessels, and the proposed collision avoidance actions.

[0068] S13. In accordance with the International Regulations for Preventing Collisions at Sea (COLREGs), determine the ship's responsibility for avoiding collisions, and abstract the responsibility for avoiding collisions into the amount of change in course, as the ship's reserve value;

[0069] S14. Construct a utility function for autonomous ships based on safety considerations and fit it to the utility function of traditional ships.

[0070] In this embodiment, the mixed navigation scenario at sea specifically refers to a maritime traffic scenario involving both conventional vessels and autonomous vessels. The vessels, based on COLREGs, determine that a collision hazard exists and an uncoordinated collision avoidance scenario has been created; therefore, the vessels begin to negotiate on this issue. Specifically: [Example:] Figure 1 As shown, for autonomous vessels, the conventional vessel is identified and tracked, the collision risk is calculated, and an initial decision is made before negotiations are initiated, which include established negotiation protocols and strategies. For conventional vessels, the watch officer identifies and tracks the autonomous vessel, assesses the current collision risk, makes appropriate decisions, and participates in the negotiations. Through multiple rounds of negotiations, both parties reach a consensus and conclude the negotiations.

[0071] In a specific implementation, as a preferred embodiment of the present invention, step S14 specifically includes:

[0072] S141. The Sigmoid function is used to map the collision risk to the range [0,1]. A higher utility value indicates a lower collision risk. The utility function is the remaining collision risk after the collision avoidance action, and includes the following calculation methods:

[0073]

[0074]

[0075] Among them, CR S Indicates the risk of a spatial collision between the two ships; CR T The time-based collision risk indicates the two ships' collision risk; CRI indicates the collision risk between the two ships. denoted by ; DCPA and TCPA represent the nearest encounter distance and nearest encounter time, respectively; d1 and d2 represent the upper and lower bounds of safe encounter in space, respectively; t1 and t2 represent the upper and lower bounds of safe encounter in time, respectively; λ determines the steepness of the transition.

[0076] S142. Fit the utility function of the conventional vessel to a linear function representing risk neutrality, and determine that the first proposal of the conventional vessel has the maximum utility value, with a utility value of 0.6 at the retention value.

[0077] In a specific implementation, as a preferred embodiment of the present invention, step S2 specifically includes:

[0078] S21, such as Figure 2 As shown, a negotiation communication protocol based on the Speech Act theory is constructed, which encapsulates the information required to avoid collisions into a structured information set;

[0079] S22. Based on the constructed negotiation communication protocol, define negotiation communication primitives, including message reception (Ack), expressing disagreement (Disagree), confirming acceptance (Accept), explicitly rejecting (Reject), requesting clarification (Doubt), verifying information (Verify), sharing information (Inform), providing advice (Advise), and making a strong request (Request).

[0080] S23. Based on the constructed negotiation communication protocol, define the communication content, including the ship's navigation information, encounter situation, and collision avoidance decision.

[0081] In a specific implementation, as a preferred embodiment of the present invention, in step S23, the defined communication content consists of keywords and values, as shown in the following formula:

[0082] Data = $Primitive:Content$

[0083] Communication content=Keyword(Value).

[0084] In a specific implementation, as a preferred embodiment of the present invention, step S3 specifically includes:

[0085] S31, such as Figure 3 As shown, a negotiation agreement based on alternating bids is constructed to simulate the "call-response" communication pattern commonly encountered in actual maritime collision avoidance.

[0086] S32. The negotiating vessels take turns proposing their respective collision avoidance plans, promoting a consensus on collision avoidance between the two parties through continuous adjustments.

[0087] In a specific implementation, as a preferred embodiment of the present invention, step S4 specifically includes:

[0088] S41, such as Figure 4 As shown, autonomous vessels adopt the Zeuthen one-way concession strategy, deciding whether to make concessions based on their maximum risk tolerance; (based on the Zeuthen one-way concession strategy, during the negotiation process, vessels will decide whether to make concessions based on their expected gains and tolerable risks).

[0089] S42. A ship shall begin with the option most favorable to itself and gradually make concessions to the option most favorable to the other party, and shall never back down.

[0090] In a preferred embodiment of the present invention, in step S5, the autonomous vessel continuously updates the estimate of the target vessel's utility function using a Bayesian learning mechanism to improve the accuracy of the information required for negotiation. This specifically includes:

[0091] S51. The initial estimate of the retention value range of autonomous vessels compared to traditional vessels is defined as RVR = {r1, r2, ..., r n The probability estimate of the retention value of autonomous ships compared to traditional ships is defined as p(r). i The conditional probability of a proposal from a traditional ship under the autonomous ship assumption is defined as p(q). i |r i ), that is, each r i The following suggestion q i The probability of;

[0092] S52. Assuming the proposal from the traditional vessel is q1, the probability distribution of the value retained by the autonomous vessel relative to the traditional vessel is updated as follows:

[0093]

[0094] S53, The estimate of the retention value of autonomous vessels relative to conventional vessels is updated as follows:

[0095]

[0096] S54. In a round of negotiations, the autonomous vessel calculates and estimates the maximum risk tolerance of both parties. The party with the smaller value will modify its proposal and make concessions. Taking the autonomous vessel as an example, the calculation of its maximum risk tolerance in the i-th round is as follows:

[0097]

[0098] in, and Let represent the utility of autonomous vessels and traditional vessels under the autonomous vessel proposal in round i, respectively. and Let U(a) and U(c) represent the utility of the traditional vessel and the autonomous vessel respectively under the traditional vessel proposal in the i-th round, and let U(a) and U(c) represent their respective utility values ​​when the negotiation fails, which are 0.

[0099] In this embodiment, the autonomous vessel uses a Bayesian learning mechanism to continuously update its estimate of the target ship's retention value, such as... Figure 5 As shown. Based on the target ship's proposals and actions, conditional probabilities are calculated, the posterior probability of the target ship's retention value is updated, the target ship's utility function is further fitted, and the target ship's subsequent behavior is predicted to improve negotiation capabilities. Specifically: the autonomous ship pre-defines its own utility function and has prior knowledge about itself. During negotiation, when the autonomous ship receives a proposal from the target ship, it is considered to have acquired posterior knowledge. It updates its prior knowledge according to a Bayesian learning mechanism and applies a Zeuthen strategy to calculate the maximum risk acceptance level, deciding whether to make concessions. The autonomous ship will then decide whether to make new proposals based on the other party's proposals. The negotiation process can be viewed as a dynamic sequence of proposals, with the last set of proposals being the negotiation solution.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A negotiation-based cooperative collision avoidance method for ships in a mixed maritime navigation scenario, characterized in that, include: S1. In a mixed navigation scenario, determine the participants in the negotiation, the negotiation topics, the retention value, and the utility function, including: S11. Use the Automatic Identification System (AIS) to obtain the position, speed, and heading information of ships, calculate the nearest encounter distance and the nearest encounter time between ships, compare the nearest encounter distance and the nearest encounter time with a preset safety threshold, identify ship pairs that may collide, and determine them as participants in the negotiation. S12. The topics of discussion include the probability of collision between vessels, the status of vessels, and the proposed collision avoidance actions. S13. In accordance with the International Regulations for Preventing Collisions at Sea, determine the ship's responsibility for avoiding collisions, and abstract the responsibility for avoiding collisions into the amount of change in course, which is taken as the ship's reserve value; S14. Construct a utility function for autonomous ships based on safety considerations, and fit it to the utility function of traditional ships, including: S141. The Sigmoid function is used to map the collision risk to the range [0, 1]. A higher utility value indicates a lower collision risk. The utility function is the remaining collision risk after the collision avoidance action, and includes the following calculation methods: in, This indicates the risk of a spatial collision between the two ships; This indicates the risk of a time-based collision between the two ships; This indicates the risk of a collision between the two ships; Represents the utility function of autonomous ships; and These represent the nearest meeting distance and the nearest meeting time, respectively. and These represent the upper and lower limits of safe encounters between ships in space, respectively. and These represent the upper and lower bounds of the safe encounter time for ships, respectively; Determines the steepness of the transition; S142. Fit the utility function of the conventional ship with a linear function representing risk neutrality, and determine that the first proposal of the conventional ship has the maximum utility value, with a utility value of 0.6 at the retention value. S2. Adopt a negotiation communication protocol based on the Speech Act theory, and define the negotiation communication primitives and content; S3. A negotiation agreement based on round-robin bidding is adopted, with the negotiating vessels submitting their respective collision avoidance plans in turn; S4. Autonomous vessels adopt a negotiation strategy based on the Zeuthen strategy to decide whether to make concessions, drive the negotiation process and reach a consensus. S5. Autonomous ships use a Bayesian learning mechanism to estimate the retention value of the target ship and further fit the performance function.

2. The method for cooperative collision avoidance of ships based on negotiation in a mixed maritime navigation scenario according to claim 1, characterized in that, Step S2 specifically includes: S21. Construct a negotiation communication protocol based on the Speech Act theory, and encapsulate the information required to avoid collisions into a structured information set; S22. Based on the constructed negotiation communication protocol, define negotiation communication primitives, including message reception, expressing disagreement, confirming acceptance, explicitly rejecting, requesting clarification, verifying information, sharing information, providing suggestions, and making strong demands; S23. Based on the constructed negotiation communication protocol, define the communication content, including the ship's navigation information, encounter situation, and collision avoidance decision.

3. The method for cooperative collision avoidance of ships based on negotiation in a mixed navigation scenario at sea, as described in claim 1, is characterized in that... Step S3 specifically includes: S31. Construct a negotiation agreement based on alternating bids to simulate the call-response communication pattern commonly encountered in actual maritime collision avoidance. S32. The negotiating vessels take turns proposing their respective collision avoidance plans, promoting a consensus on collision avoidance between the two parties through continuous adjustments.

4. The method for cooperative collision avoidance of ships based on negotiation in a mixed maritime navigation scenario according to claim 1, characterized in that, Step S4 specifically includes: S41. Autonomous vessels adopt the Zeuthen one-way concession strategy, deciding whether to make concessions based on their maximum risk tolerance. S42. A ship shall begin with the option most favorable to itself and gradually make concessions to the option most favorable to the other party, and shall never back down.

Images