A negotiation interaction decision-making method for ships with different degrees of autonomy

Abstract

The application provides a negotiation interaction decision-making method for ships with different autonomous degrees, which comprises the following steps: obtaining navigation situation awareness information of each ship, analyzing the composition of the navigation situation awareness information and negotiation interaction decision-making requirements; constructing a navigation situation awareness ontology and a negotiation interaction decision-making ontology based on a semantic interoperability standard according to the analysis results of the navigation situation awareness information composition and negotiation interaction decision-making requirements; determining a ship interaction opportunity by constructing a collision risk model based on fuzzy theory; adaptively selecting a ship interaction target based on a multi-stage ship action domain; and reaching a decision consensus by using the negotiation interaction decision-making ontology to negotiate a ship avoidance scheme. The application can realize the intention interaction between ships with different autonomous degrees and reach a decision consensus on the basis of real-time processing of navigation situation information between heterogeneous devices, thereby providing a solution for safe, efficient and collaborative avoidance of ships in autonomous navigation.

Description

Technical Field

[0001] This invention relates to the field of intelligent decision-making for autonomous ships, and more particularly to a negotiation and interactive decision-making method for ships with different levels of autonomy. Background Technology

[0002] Autonomous vessels are a hot topic in the shipping industry, representing the future of maritime transport technology by utilizing advanced sensing, cognition, decision-making, and control technologies. In 2017, the International Maritime Organization (IMO) defined four levels of autonomy for autonomous vessels (MASS). Countries worldwide have subsequently conducted research and testing on autonomous vessels and achieved phased progress. The deployment of autonomous vessels in actual production activities is an inevitable trend, and the future will see complex mixed maritime traffic scenarios where vessels with different levels of autonomy coexist.

[0003] Autonomous decision-making is a crucial process for safe navigation and collision avoidance of ships. However, most domestic and international research lacks information exchange with target vessels, failing to accurately understand their navigational intentions. This leads to overly conservative collision avoidance behaviors, with the ship, acting as an intelligent vessel, undertaking almost all collision avoidance actions. This strategy is inconsistent with COLREGs and reflects real-world maritime traffic conditions. Furthermore, with the advent of complex and ever-changing mixed maritime traffic scenarios, when ships with varying levels of autonomy encounter each other, the heterogeneity of their navigation systems makes information exchange and communication difficult through conventional methods, introducing new risks to maritime navigation safety. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the present invention provides a negotiation and interaction decision-making method for ships with different levels of autonomy. The present invention is based on an ontology semantic interoperability ship negotiation and interaction decision-making knowledge model, and aims to provide a standardized information interaction tool for negotiation and interaction decision-making between ships with different levels of autonomy.

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

[0006] A negotiation-based interactive decision-making method for ships with different levels of autonomy includes:

[0007] Acquire navigation situation awareness information of each vessel, and analyze the composition of navigation situation awareness information and the needs for consultation and interactive decision-making.

[0008] Based on the analysis results of the composition of navigation situation awareness information and the needs of negotiation and interactive decision-making, a navigation situation awareness ontology and a negotiation and interactive decision-making ontology are constructed based on semantic interoperability standards.

[0009] A collision risk model is constructed based on fuzzy theory to determine the timing of ship interactions;

[0010] Ship interaction targets are adaptively selected based on the multi-stage ship reachability domain;

[0011] The negotiation and interactive decision-making ontology is used to negotiate ship avoidance schemes and reach a consensus on the decision.

[0012] Furthermore, the ship navigation situation awareness information includes prior information from electronic charts and real-time environmental information from shipboard sensing devices; the negotiation and interactive decision-making requirements include: determining the timing of interaction, selecting interaction targets, and standardizing the negotiation and decision-making process.

[0013] Furthermore, the semantic interoperability standard is represented as a triple (C, R, I), where C is the set of navigation information categories; R is the set of navigation information attributes, including attributes describing the relationships between classes and data attributes; and I is the actual instance in the class.

[0014] Furthermore, a navigation situation awareness ontology is constructed based on semantic interoperability standards, including:

[0015] The navigation scenario information is abstracted and conceptualized into a set of classes, which includes the ship class, the target ship class, and the environment class. Each class contains multiple subclass sets, and there are relational attributes and data attributes between classes and between classes themselves.

[0016] By combining real-time perception information, classes and relationships between classes are instantiated and populated with instance information in the navigation scene.

[0017] Furthermore, the negotiation and interaction decision ontology includes information sending, intent declaration, expected behavior, and negotiation response classes, and provides information interaction methods for negotiation and communication between ships by using standardized knowledge models.

[0018] Furthermore, a collision hazard model is constructed based on fuzzy theory to determine the timing of ship interactions, including:

[0019] The collision hazard level is constructed using the following formula:

[0020] CRI=ω1u d +ω2u q +ω3u k +ω4u DCPA +ω5u TCPA

[0021] Among them, u d u q u k u DCPA u TCPA These are the membership functions for inter-ship distance, hull angle, ship speed ratio, shortest meeting distance, and minimum meeting time, respectively, specifically unit real numbers from the interval [0,1], where ω is the fuzzy weight allocation coefficient;

[0022] When the Collision Risk Index (CRI) is greater than 0.5, the vessel is considered to be at risk of collision, and this is considered to be the time for interaction.

[0023] Furthermore, based on the multi-stage ship operation domain, ship interaction targets are selected, including:

[0024] The multi-stage reachable zone is defined by concentric circles representing the action zone limits for each stage of a collision, with 3-6 nautical miles considered as the active negotiation zone, where the two ships should negotiate and reach a consensus on avoidance decisions.

[0025] Furthermore, selecting ship interaction targets based on multi-stage ship action domains also includes:

[0026] When multiple ships meet, the interaction targets selected by each ship together form a ship self-organizing network with ships as nodes, which is used to select interaction targets among multiple ships.

[0027] Furthermore, the negotiation and interactive decision-making ontology is used to negotiate ship avoidance schemes and reach a consensus, including:

[0028] Set the information sending class to send ship information that is conducive to negotiation, such as ship status and meeting situation, through the navigation situation cognition ontology;

[0029] Set the intention declaration class to inform the target vessel of the evasive maneuver that this vessel intends to take;

[0030] Set the expected behavior class to express the ship's expectation of the target ship's avoidance maneuver;

[0031] The negotiation response class is set up to determine whether a consensus has been reached in the negotiation decision.

[0032] Compared with the prior art, the present invention has the following advantages:

[0033] This invention provides a negotiation and interactive decision-making method for ships with different levels of autonomy. It can fully integrate and utilize navigation situation information based on an ontology semantic interoperability negotiation and decision-making knowledge model, and provide a standardized negotiation and interactive decision-making tool for heterogeneous navigation systems of ships with different levels of autonomy. This solves practical problems such as uncoordinated avoidance caused by the lack of intent exchange in the existing ship intelligent decision-making process. Attached Figure Description

[0034] 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.

[0035] Figure 1 This is a flowchart of a negotiation and interactive decision-making method for ships with different levels of autonomy, as proposed in this invention.

[0036] Figure 2 This is a set of classes and attribute graphs within the ontology knowledge model for situational awareness and negotiation-based interactive decision-making during navigation.

[0037] Figure 3 This is an instance diagram of the navigation scenario of the present invention.

[0038] Figure 4 This is a multi-stage ship operation domain diagram of the present invention.

[0039] Figure 5 A schematic diagram for selecting interaction targets between ships.

[0040] Figure 6 Flowchart for negotiating collision avoidance between ships. Detailed Implementation

[0041] 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.

[0042] like Figure 1 The flowchart shown illustrates a negotiation and interactive decision-making method for ships with different levels of autonomy. The invention is divided into five steps: First, acquiring navigation situation awareness information for each ship and analyzing the composition of this information and the negotiation and interactive decision-making requirements; second, constructing a navigation situation awareness ontology and a negotiation and interactive decision-making ontology based on semantic interoperability standards, according to the analysis results of the navigation situation awareness information composition and negotiation and interactive decision-making requirements; third, constructing a collision risk model based on fuzzy theory to determine the timing of ship interaction; fourth, adaptively selecting ship interaction targets based on multi-stage ship action domains; and fifth, using the negotiation and interactive decision-making ontology to negotiate ship avoidance schemes and reach a decision consensus.

[0043] like Figure 2The diagram shows the class sets and attribute graphs within the knowledge model of the navigation situation cognition and negotiation interaction decision-making ontology of this invention. Specifically, the navigation situation cognition ontology comprises three class sets: OwnShip, TargetShip, and Environment, each of which can be further subdivided. Analyzing the maritime avoidance negotiation decision-making process, a negotiation interaction decision-making ontology is constructed, specifically including: Sent_Information, Statement_Intention, Expectation_Behavior, and Negotiation_Answer, with each step containing multiple negotiation measures. Creating attributes for classes in the ontology provides a clearer semantic description of the classes; attributes can be categorized into relational attributes and data attributes. Relationship attributes are used to describe the relationships between classes, including positional relationships, subordinate relationships, etc., and are described by a domain and a range. Data attributes are the characteristics of a class itself, such as the draft, speed, and course of a ship. The relationships can also be constrained by a domain and a range, where the domain is the class type and the range is the basic data type.

[0044] like Figure 3 This is an instantiation diagram of the navigation scenario of the present invention. The preceding description of the modeling process of the navigation situation cognition ontology and the negotiation and interaction decision ontology is equivalent to filling in the background knowledge that makes up the ontology knowledge model. Contextual knowledge, on the other hand, needs to be re-expressed using the pre-defined conceptual model in the background knowledge, based on the actual scene elements and real-time information in the navigation scenario; this is the instantiation of the ontology model. The following example of an encounter scenario illustrates this process, such as... Figure 3 As shown, through multi-source sensor data acquisition and automated information processing technology, in Figure 3 The example of a Chinese ship is yukun, which navigates in open waters and encounters the target ship yupeng. At the same time, the static and dynamic information of yukun can be viewed at any time.

[0045] like Figure 4 This paper presents a multi-stage ship action domain diagram. Specifically, a ship collision situation typically progresses through four stages: free maneuver stage, collision hazard formation stage, imminent situation stage, imminent danger stage, and finally, collision. Based on navigational practice and COLREGs requirements, ships may make different behavioral decisions at different encounter stages. Therefore, this paper proposes a multi-stage ship action domain. For ease of calculation and axiomatic expression, concentric circles are used to represent the boundaries of each stage's action domain, centered on the ship. Figure 4As shown. When two vessels approach each other and pose a collision hazard, maritime practice generally considers a distance of 3-6 nautical miles as appropriate, and the two vessels should engage in active and effective communication and coordination.

[0046] like Figure 5 The diagram illustrates the selection of interaction targets for ships. Specifically, assuming multiple ships meet, ship OS will select ship TS1 and ship TS2 as its interaction targets. At the same time, ship TS2 will select ship OS and ship TS3 as its interaction targets. After each ship selects interaction targets according to the above criteria, these interaction targets together form a ship self-organizing network with ships as nodes for negotiation and interaction.

[0047] like Figure 6 The flowchart for the ship-to-ship collision avoidance negotiation process is as follows: Within the ship's perception range, the target ship's situational information is acquired in real time. Collision risk is calculated through situational awareness to determine if an interaction opportunity has been reached. Once the interaction opportunity is reached, a multi-stage ship action domain is used to select an interaction target for information exchange. The higher the ship's autonomy, the stronger its ability to cope with complex navigation environments. Therefore, it is believed that sending an interaction request from a ship with a relatively high level of autonomy to a ship with a relatively low level of autonomy is more conducive to reaching a consensus decision. Based on situational awareness and recommended avoidance behavior, active communication and negotiation are conducted with the target ship via Sent_Information, Statement_Intention, and Expectation_Behavior. If a Permission response is received, a consensus decision is reached, and the decision information is executed by the ship's control system. If a Refuse response is received, negotiation continues until a consensus is reached between the two parties.

[0048] 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 therein. Such 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 and interactive decision-making method for ships with different levels of autonomy, characterized in that, include: Acquire navigation situation awareness information of each vessel, and analyze the composition of navigation situation awareness information and the needs for consultation and interactive decision-making. Based on the analysis results of the composition of navigation situation awareness information and the needs of negotiation and interactive decision-making, a navigation situation awareness ontology and a negotiation and interactive decision-making ontology are constructed based on semantic interoperability standards. A collision risk model is constructed based on fuzzy theory to determine the timing of ship interactions; Adaptive selection of ship interaction targets based on multi-stage ship action domain; The negotiation and interactive decision-making ontology is used to negotiate ship avoidance schemes and reach a consensus on the decision-making process. Based on fuzzy theory, a collision hazard model is constructed to determine the timing of ship interactions, including: The collision hazard level is constructed using the following formula: in, , , , , These are the membership functions for ship-to-ship distance, hull angle, speed ratio, shortest meeting distance, and minimum meeting time, respectively, specifically unit real numbers from the interval [0,1]. Assign coefficients to fuzzy weights; When the Collision Risk Index (CRI) is greater than 0.5, the vessel is considered to be at risk of collision, and this is considered to be the time for interaction.

2. The negotiation and interactive decision-making method for ships with different levels of autonomy according to claim 1, characterized in that, The ship navigation situation awareness information includes prior information from electronic charts and real-time environmental information from shipborne sensing devices. The negotiation and interactive decision-making requirements include: determining the timing of the interaction, selecting the interaction target, and standardizing the negotiation and decision-making process.

3. The negotiation and interactive decision-making method for ships with different levels of autonomy according to claim 2, characterized in that, The semantic interoperability standard is represented as a triple (C, R, I), where C is the set of navigation information categories; R is the set of navigation information attributes, including attributes describing the relationships between classes and data attributes; and I is the actual instance in the class.

4. The negotiation and interactive decision-making method for ships with different levels of autonomy according to claim 2, characterized in that, A navigation situation awareness ontology is constructed based on semantic interoperability standards, including: The navigation scenario information is abstracted and conceptualized into a set of classes, which includes the ship class, the target ship class, and the environment class. Each class contains multiple subclass sets, and there are relational attributes and data attributes between classes and between classes themselves. By combining real-time perception information, classes and relationships between classes are instantiated and populated with instance information in the navigation scene.

5. A negotiation and interactive decision-making method for ships with different levels of autonomy, as described in claim 4, is characterized in that... The negotiation and interaction decision ontology includes information sending, intent declaration, expected behavior, and negotiation response classes. It provides information exchange methods for negotiation and communication between ships by using standardized knowledge models.

6. A negotiation and interactive decision-making method for ships with different levels of autonomy, as described in claim 4, is characterized in that... Based on the selection of ship interaction targets in a multi-stage ship action domain, including: The multi-stage vessel action zone is defined by concentric circles centered on the vessel during the different stages of a collision. 3-6 nautical miles is considered the active negotiation action zone, where the two vessels should negotiate and reach a consensus on avoidance decisions.

7. A negotiation and interactive decision-making method for ships with different levels of autonomy, as described in claim 6, is characterized in that... Selecting ship interaction targets based on multi-stage ship action domains also includes: When multiple ships meet, the interaction targets selected by each ship together form a ship self-organizing network with ships as nodes, which is used to select interaction targets among multiple ships.

8. A negotiation and interactive decision-making method for ships with different levels of autonomy, as described in claim 5, is characterized in that... The negotiation and interactive decision-making ontology is used to negotiate ship avoidance schemes and reach a consensus, including: The information transmission class is set to transmit ship status information and encounter situation information that are conducive to negotiation through the navigation situation cognition ontology; Set the intention declaration class to inform the target vessel of the evasive maneuver that this vessel intends to take; Set the expected behavior class to express the ship's expectation of the target ship's avoidance maneuver; The negotiation response class is set up to determine whether a consensus has been reached in the negotiation decision.

Images