Ship collision avoidance decision triggering opportunity modeling method for simulating human social cognition process

Abstract

The invention provides a ship collision avoidance decision triggering opportunity modeling method for simulating a human social cognition process, which comprises the following steps of: acquiring ship AIS (Automatic Identification System) data, extracting longitude and latitude, course, speed and captain information of a ship in the AIS data, and constructing an eccentric elliptical ship field; calculating the remaining distance DDV before the TS of the other ship intrudes into the OS field of the ship, the remaining time TDV before the TS of the other ship intrudes into the OS field of the ship and the inter-ship distance D; constructing a drift rate function, and forming a unit time information accumulation amount and a current time information accumulation total amount; and constructing a decision boundary, and forming a decision triggering result based on the information cumulant and the intervention scene. According to the method, on the basis of a variable drift diffusion model in the field of human cognition modeling, a specific evidence accumulation mechanism in a human cognition process is used for reference, an instantaneous risk parameter is converted into an evidence accumulation amount in unit time, and a threshold judgment triggering method is replaced by a mode of judging whether the evidence amount reaches a triggering boundary or not; a social cognitive decision mode of human is simulated by introducing security decision duration.

Description

Technical Field

[0001] The present invention relates to the technical field of maritime traffic management. Specifically, it particularly relates to a method for modeling the triggering timing of ship collision avoidance decisions that mimics the human social cognitive process. Background Art

[0002] The collision avoidance decision of autonomous ships is an important task related to their navigation safety. The nature of the collision avoidance problem determines that the navigation decision-making behavior of ships is not only governed by their own rational logic, but should also consider the cognitive characteristics of other ships during the encounter process, and make the collision avoidance decision socially recognized on the basis of ensuring navigation safety. As the switch for autonomous ship collision avoidance, the decision trigger model directly determines whether to take collision avoidance actions and the timing of formulating collision avoidance strategies, which is a key issue in the collision avoidance process.

[0003] Currently, the collision avoidance decision trigger model mostly relies on the real-time calculation results of the inter-ship collision risk, and decides the decision timing by comparing with a pre-set risk threshold. This mode is easily interfered by instantaneous data fluctuations, resulting in unstable triggering of collision avoidance decisions. At the same time, the decision trigger achieved through risk threshold determination belongs to ship automation rather than autonomy, which is quite different from the human cognitive decision-making process, and the resulting collision avoidance behavior is difficult to be accepted by other ships. Summary of the Invention

[0004] In view of the above-mentioned technical problems, a method for modeling the triggering timing of ship collision avoidance decisions that mimics the human social cognitive process is provided. The present invention mimics the human social cognitive process and designs a trigger VDDM-CADT (the specific Variable Drift Diffusion Model for ship Collision Avoidance Decision Trigger). This trigger is based on the variable drift diffusion model VDDM commonly used in the field of human cognitive modeling, draws on the unique evidence accumulation mechanism in the human cognitive process, converts the instantaneous risk parameter into the evidence accumulation amount per unit time, replaces the threshold determination trigger method with the mode of whether the evidence amount reaches the trigger boundary, and enables the trigger to mimic the human social cognitive decision-making mode by introducing the concept of safe decision-making duration.

[0005] The technical means adopted by the present invention are as follows:

[0006] A method for modeling the triggering timing of ship collision avoidance decisions that mimics the human social cognitive process, comprising:

[0007] S1. Obtain ship AIS data, extract the longitude, latitude, course, speed and ship length information of the ship from the AIS data, and construct an eccentric elliptical ship domain;

[0008] S2. Based on the constructed eccentric ellipse ship domain, calculate the remaining distance DDV before the other ship's TS intrudes into the own ship's OS domain, the remaining time TDV before the other ship's TS intrudes into the own ship's OS domain, and the ship-to-ship distance D;

[0009] S3. Construct the drift rate function f[i] to form the information accumulation amount per unit time and the total information accumulation amount at the current moment;

[0010] S4. Construct the decision boundary and form the decision trigger result based on the information accumulation amount and the intervention scenario.

[0011] Furthermore, in step S1, constructing the eccentric ellipse ship domain specifically includes:

[0012] The long axis direction of the elliptical domain is parallel to the course of the own ship's OS. The semi-major axis of the elliptical domain is a, the semi-minor axis of the elliptical domain is b, the displacement of the ship from the center of the ellipse along the semi-major axis towards the stern is Δa, and the displacement of the ship from the center of the ellipse along the semi-minor axis to the left is Δb. And set the values of a, b, Δa, and Δb as follows:

[0013] a = 10l;

[0014] b = 5l;

[0015] Δa = 2.5l;

[0016] Δb = 1.25l;

[0017] Where l represents the length of the own ship.

[0018] Furthermore, step S2 specifically includes:

[0019] S21. According to the scaling factor f when the other ship's TS intrudes into the own ship's OS domain to the maximum extent min , calculate the remaining distance DDV before the other ship's TS intrudes into the own ship's OS domain. The calculation formula is as follows:

[0020] DDV = max(1 - f min , 0)

[0021] Where DDV is determined by the domain scaling factor f, and f is a quadratic function of time t, f(t). Then, in the case where f has a solution, obtain the scaling factor f when the other ship's TS intrudes into the own ship's OS domain to the maximum extent min , specifically, if f min > 1, it proves that the other ship's TS will never intrude into the own ship's OS domain; if f min = 1, it means that the other ship's TS just cuts through the domain boundary of the own ship's OS; if f min < 1, it indicates that the other ship's TS will inevitably intrude into the own ship's OS domain;

[0022] S22. Calculate the remaining time TDV before the other ship TS intrudes into the own ship OS area. The calculation formula is as follows:

[0023] f(t) = 1

[0024] Among them, TDV is obtained by solving according to the formula f(t) = 1.

[0025] Furthermore, step S3 specifically includes:

[0026] S31. The drift rate includes the theoretical instantaneous safety decision-making duration TISDT and the theoretical standard safety decision-making duration TSSDT. Among them, K i represents the safety decision-making duration at any current moment i, represents the maximum value of the safety decision-making duration in the current TDV sub-interval. The calculation formula is as follows:

[0027] K i = h ψ (DDV i , D i )

[0028]

[0029] Among them, h ψ is the analytical function of SDT, DDV i represents the DDV value at moment i, and D i represents the distance between the two ships at moment i;

[0030] S32. Calculate the evidence accumulation direction g[i] at moment i. The calculation formula is as follows:

[0031]

[0032] Among them, DDV i crit represents the corresponding DDV value;

[0033] S33. Based on step S31 and step S32, define the drift rate function f[i] as follows:

[0034]

[0035] Among them, Δt is the time step, λ is the tuning factor, ε is a fixed constant, ∑t i is the cumulative decision-making time at the current moment i, and |∑t i - K i | reflects the rate of evidence accumulation;

[0036] S34. Construct the information accumulation model of VDDM-CADT. The formula is as follows:

[0037] ΔX[i] = X[i] - X[i - 1]

[0038] = ωf[i] + W[i]

[0039] Wherein, X[i] represents the total amount of information accumulated at the current time i, ΔX[i] represents the unit-time information accumulation amount with Δt as the time step, the unit of Δt is seconds, and W[i] represents diffusion noise; it is assumed that the increment of the diffusion part follows a standard Wiener process, indicating that the change of the diffusion increment in any finite time follows a normal distribution, expressed as W[i] ~ N(0, Δtσ 2 ); ω represents the scale parameter, reflecting the proportion of the remaining information accumulation amount from the distance trigger decision threshold to the current departure trigger threshold. When the information accumulation value is closer to the trigger threshold at the current time, the calculated ΔX[i] will be given a smaller scale factor; the calculation formula of the scale parameter ω is as follows:

[0040]

[0041] Wherein, B i represents the information accumulation boundary at time i, α is the power function parameter, and α ∈ (0, 1);

[0042] S35. When setting T 0 as the initial time of information accumulation, with Δt as the unit time length, calculate the information accumulation amount at time T d The calculation formula is as follows:

[0043]

[0044] Furthermore, both the theoretical instantaneous safety decision time TISDT and the theoretical standard safety decision time TSSDT in step S31 are related to the safety decision time SDT, where SDT represents the upper limit of the collision avoidance decision trigger time for the ship's driver on the premise of ensuring navigation safety, that is, the longest time to determine whether to make a collision avoidance decision. The setting of the safety decision time SDT is related to the remaining distance DDV, the distance between ships D, and TDV before the other ship TS intrudes into the own ship OS area. Specifically as follows:

[0045] The remaining distance DDV and the distance between ships D before the other ship TS intrude into the own ship OS area represent the urgency of the current situation, that is, the clearer the urgency of the current encounter situation, the relatively shorter the duration of SDT. On the contrary, the safety decision time SDT is longer; the larger the remaining time TDV before the other ship TS intrudes into the own ship OS area, the more sufficient the safety decision time for the human driver, and the relatively larger the safety decision time SDT will be.

[0046] Furthermore, the specific calculation process of the safety decision time SDT includes:

[0047] Define the value ranges and intervals for the remaining time TDV before the other ship TS enters the own ship OS area, the remaining distance DDV before the other ship TS enters the own ship OS area, and the distance D. Here, the unit of the remaining time TDV before the other ship TS enters the own ship OS area is hours, and the unit of D is nautical miles;

[0048] Divide the remaining time TDV before the other ship TS enters the own ship OS area into six sub - intervals [TDV j , TDV j+1 ). For each group of the remaining time TDV before the other ship TS enters the own ship OS area, according to the investigation and analysis results of ship drivers, respectively count the proportion of the remaining time TDV before the other ship TS enters the own ship OS area under different remaining distances DDV and ship - to - ship distances D;

[0049] Obtain the analytical formula of the safe decision - making duration SDT through data fitting, and calculate the safe decision - making duration SDT according to the actual independent variable values (specific values of DDV and D), as follows:

[0050]

[0051] Among them, TDV j+1 represents the TDV value at the (j + 1) - th moment, TDV j represents the TDV value at the j - th moment, DDV j+1 represents the DDV value at the (j + 1) - th moment, DDV j represents the DDV value at the j - th moment, D j+1 represents the distance between the two ships at the (j + 1) - th moment, D j represents the distance between the two ships at the j - th moment.

[0052] Furthermore, assume that the numerical expression of the safe decision - making duration SDT is a special Gaussian - type smooth function, that is, a bump function, defined as:

[0053]

[0054] Among them, A represents the amplitude of the bump function, that is, the highest value at the center point, r represents the central radius of the bump function, (x 0 , y 0 ) represents the center point of the bump function, and d represents the minimum margin of the bump function;

[0055] Based on the numerical expression of the safe decision - making duration SDT, obtain the SDT calculation formula SDT(DDV i , D i ), as follows:

[0056] When then there is:

[0057]

[0058] When , then there is:

[0059] SDT(DDV i , D i ) = SDT min

[0060] where DDV 0 and D 0 represent the coordinates of the center point of the bulge function; SDT max represents the ratio of the safe decision-making time SDT to the remaining time TDV before the most entangled other ship TS invades the own ship OS area under the combination of the remaining distance DDV and the ship-to-ship distance D; SDT min represents the ratio of the minimum decision trigger time for ship collision avoidance to the remaining time TDV before the other ship TS invades the own ship OS area; r SDT represents the radius of the bulge function; DDV e and D e ' are the center and radius of the bulge function respectively;

[0061] Since the interval range of the ship-to-ship distance D does not coincide with the remaining distance DDV before the other ship TS invades the own ship OS area, it is necessary to map the value of the ship-to-ship distance D from the interval [2, 6] to the interval [0, 1], specifically:

[0062] D' = (D - 2) / 4.

[0063] Furthermore, step S4 specifically includes:

[0064] S41. Construct a decision boundary, which is a function of TDV. When designing, refer to the Sigmoid activation function, whose characteristic is to compress a large numerical range into the interval (0, 1), and is defined as follows:

[0065]

[0066] where b 0 is a parameter for adjusting the boundary size, k is the sensitivity of the boundary to TDV, μ is the critical value of the boundary at the baseline value, and k 0 is the minimum value for restricting the boundary;

[0067] S42. Calculate the decision trigger result based on the information accumulation amount The calculation formula is as follows:

[0068]

[0069] Among them, is the decision trigger result after intervention, X i represents the total amount of information accumulated at the current moment, B(TDV i ) represents the information boundary size at the current moment, μ is the proportionality factor, C 1 and C 2 are making collision avoidance decisions and not making collision avoidance decisions respectively;

[0070] S43. Based on the intervention scenario ξ i , calculate the decision trigger result, and the calculation formula is as follows:

[0071]

[0072] Among them, the intervention scenario ξ i includes Condi_1 and Condi_2. Condi_1 is defined as the total duration of information accumulation at the current moment being greater than or equal to the current TISDT, Condi_1: Σt i ≥K i ; Condi_2 is defined as the total change in the α sampling moments before time i and the current moment TISDT being less than or equal to the actual cumulative time, Condi_2: K i-α -K i ≤αΔt;

[0073] S44. When the intervention scenario ξ i satisfies Condi_1 and Condi_2, execute C 1 , otherwise execute C 2 .

[0074] Compared with the prior art, the present invention has the following advantages:

[0075] 1. A method for modeling the triggering timing of ship collision avoidance decisions that mimics the human social cognitive process provided by the present invention. The constructed decision trigger model does not rely on the instantaneous value of the ship - to - ship collision risk parameter, but models according to the collision risk change trend within the safe decision time interval. The decision trigger timing calculated thereby is a stable output obtained through a certain observation period, thus avoiding the contingency and volatility problems in previous trigger calculations.

[0076] 2. A method for modeling the triggering timing of ship collision avoidance decisions that mimics the human social cognitive process provided by the present invention. For the first time, a variable drift diffusion model VDDM is used to model the triggering mechanism of autonomous ship collision avoidance decisions. The decision trigger model integrating the VDDM framework can approximately mimic the cognitive decision - making process of human drivers.

[0077] 3. The ship collision avoidance decision-making trigger timing modeling method provided by the present invention, which imitates the human social cognitive process, explores a feasible path for integrating cognitive models in autonomous navigation systems, considers the necessity of human-like decision-making for autonomous ships in mixed scenarios, and provides certain theoretical and technical support for the social operation of autonomous ships.

[0078] Based on the above reasons, the present invention can be widely promoted in the fields of marine ship traffic and the like. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0080] Figure 1 It is a flowchart of the method of the present invention.

[0081] Figure 2 It is a schematic diagram of the eccentric elliptical ship domain constructed by the present invention.

[0082] Figure 3 It is an SDT function graph under different TDV intervals provided by the present invention.

[0083] In the figure: (a) TDV ∈ [0.1, 0.2); (b) TDV ∈ [0.2, 0.3); (c) TDV ∈ [0.3, 0.4); (d) TDV ∈ [0.4, 0.5); (e) TDV ∈ [0.5, 0.6); (f) TDV ∈ [0.6, 0.7);

[0084] Figure 4 It is a schematic diagram of the ship encounter situation provided by the present invention.

[0085] In the figure: (a) Scenario I; (b) Scenario II; (c) Scenario III;

[0086] Figure 5 It is a schematic diagram of the initial parameters of the ship provided by the present invention.

[0087] Figure 6 It is a schematic diagram of the change of domain parameters under Scenario I provided by the present invention.

[0088] In the figure: (a) DDV change record of TS-1; (b) TDV change record of TS-1;

[0089] Figure 7 It is a schematic diagram of the change of domain parameters under Scenario II provided by the present invention.

[0090] In the figure: (a) Record of the DDV change of TS-1; (b) Record of the TDV change of TS-1; Record of the DDV change of TS-2; Record of the TDV change of TS-2;

[0091] Figure 8 This is a schematic diagram of the domain parameter change under Scenario III provided by the present invention.

[0092] In the figure: (a) Record of the DDV change of TS-1; (b) Record of the TDV change of TS-1; (c) Record of the DDV change of TS-2; (d) Record of the TDV change of TS-2; (e) Record of the DDV change of TS-3; (f) Record of the TDV change of TS-3;

[0093] Figure 9 This is a record diagram of the model information accumulation and boundary change of Scenario I provided by the present invention.

[0094] Figure 10 This is a record diagram of the model information accumulation and boundary change of Scenario II provided by the present invention.

[0095] In the figure: Diagram of the information accumulation and triggering process of TS-1; Diagram of the information accumulation and triggering process of TS-2;

[0096] Figure 11 This is a record diagram of the model information accumulation and boundary change of Scenario III provided by the present invention.

[0097] In the figure: (a) Diagram of the information accumulation and triggering process of TS-1; (b) Diagram of the information accumulation and triggering process of TS-2; (c) Diagram of the information accumulation and triggering process of TS-3. Detailed implementation manners

[0098] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0099] It should be noted that the terms "first", "second", etc. in the description, claims and the above-mentioned drawings of the present invention are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily limit to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0100] As Figure 1 shown, the present invention provides a method for modeling the triggering timing of ship collision avoidance decision-making that imitates the human social cognitive process, including:

[0101] S1. Obtain ship AIS data, extract the longitude, latitude, course, speed and ship length information of the ship in the AIS data, and construct an eccentric elliptical ship domain;

[0102] S2. Based on the constructed eccentric elliptical ship domain, calculate the remaining distance DDV before the other ship TS intrudes into the own ship OS domain, the remaining time TDV before the other ship TS intrudes into the own ship OS domain, and the ship-to-ship distance D;

[0103] S3. Construct a drift rate function f[i] to form the information accumulation amount per unit time and the total information accumulation amount at the current moment;

[0104] S4. Construct a decision boundary and form a decision trigger result based on the information accumulation amount and the intervention scenario.

[0105] In specific implementation, as a preferred implementation manner of the present invention, in step S1, constructing an eccentric elliptical ship domain specifically includes:

[0106] The long axis direction of the elliptical domain is parallel to the course of the own ship OS. The major semi-axis of the elliptical domain is a, the minor semi-axis of the elliptical domain is b, the displacement of the ship from the center of the ellipse along the semi-major axis to the stern is Δa, and the displacement of the ship from the center of the ellipse along the minor semi-axis to the left is Δb. And set the values of a, b, Δa and Δb, as Figure 2 shown, specifically as follows:

[0107] a = 10l;

[0108] b = 5l;

[0109] Δa = 2.5l;

[0110] Δb = 1.25l;

[0111] Among them, l represents the length of the ship itself.

[0112] In specific implementation, as a preferred implementation manner of the present invention, step S2 specifically includes:

[0113] S21. According to the scaling factor f when the other ship TS intrudes into the OS area of the own ship to the maximum extent min , calculate the remaining distance DDV before the other ship TS intrudes into the OS area of the own ship. The calculation formula is as follows:

[0114] DDV = max(1 - f min , 0)

[0115] Among them, DDV is determined by the area scaling factor f, and f is a quadratic function of time t, f(t). Then, in the case where f has a solution, the scaling factor f when the other ship TS intrudes into the OS area of the own ship to the maximum extent min is obtained. Specifically, if f min > 1, it proves that the other ship TS will never intrude into the OS area of the own ship; if f min = 1, it means that the other ship TS just cuts across the area boundary of the own ship OS; if f min < 1, it indicates that the other ship TS will inevitably intrude into the OS area of the own ship;

[0116] S22. Calculate the remaining time TDV before the other ship TS intrudes into the OS area of the own ship. The calculation formula is as follows:

[0117] f(t) = 1

[0118] Among them, TDV is obtained by solving according to the formula f(t) = 1.

[0119] In specific implementation, as a preferred implementation manner of the present invention, step S3 specifically includes:

[0120] S31. The drift rate includes the theoretical instantaneous safety decision-making duration TISDT and the theoretical standard safety decision-making duration TSSDT. Among them, K i represents the safety decision-making duration at any current moment i, and represents the maximum value of the safety decision-making duration in the current TDV sub-interval. The calculation formula is as follows:

[0121] K i = h ψ (DDV i , D i )

[0122]

[0123] Among them, h ψ is an analytical function of SDT, DDV i represents the DDV value at moment i, Di represents the distance between two ships at time i.

[0124] S32. Calculate the evidence accumulation direction g[i] at time i, and the calculation formula is as follows:

[0125]

[0126] where DDV i crit represents the corresponding DDV value;

[0127] S33. Based on steps S31 and S32, define the drift rate function f[i] as follows:

[0128]

[0129] where Δt is the time step, with the unit of seconds, λ is the tuning factor, ε is a fixed constant, ∑t i is the cumulative decision time at the current time i, and |∑t i -K i | reflects the rate of evidence accumulation;

[0130] S34. Construct the information accumulation model of VDDM-CADT, and the formula is as follows:

[0131] ΔX[i] = X[i] - X[i - 1]

[0132] = ωf[i] + W[i]

[0133] where X[i] represents the total amount of information that has been accumulated at the current time i, ΔX[i] represents the unit-time information accumulation amount with Δt as the time step, the unit of Δt is seconds, and W[i] represents the diffusion noise; it is assumed that the increment of the diffusion part follows a standard Wiener process, indicating that the change of the diffusion increment in any finite time follows a normal distribution, expressed as W[i] ∼ N(0, Δtσ 2 ); ω represents the scale parameter, reflecting the proportion of the remaining information accumulation amount from the distance trigger decision threshold to the current departure trigger threshold. When the information accumulation value is closer to the trigger threshold at the current time, the calculated ΔX[i] will be given a smaller scale factor; the calculation formula of the scale parameter ω is as follows:

[0134]

[0135] where B i represents the information accumulation boundary at time i, α is the power function parameter, and α ∈ (0, 1);

[0136] S35. When setting T 0At the initial moment of information accumulation, with Δt as the unit time length, calculate T d The information accumulation amount at the moment, and the calculation formula is as follows:

[0137]

[0138] In specific implementation, as a preferred implementation manner of the present invention, both the theoretical instantaneous safety decision time TISDT and the theoretical standard safety decision time TSSDT in step S31 are related to the safety decision time SDT, where SDT represents the upper limit of the collision avoidance decision trigger time of the ship's driver on the premise of ensuring navigation safety, that is, the longest time to determine whether to make a collision avoidance decision. The setting of the safety decision time SDT is related to the remaining distance DDV, the distance D between ships, and TDV before the other ship TS intrudes into the own ship OS area, specifically as follows:

[0139] The remaining distance DDV and the distance D between ships before the other ship TS intrudes into the own ship OS area represent the urgency of the current situation, that is, the clearer the urgency of the current encounter situation, the relatively shorter the duration of SDT. On the contrary, the safety decision time SDT is longer; the larger the remaining time TDV before the other ship TS intrudes into the own ship OS area, the more sufficient the safety decision time of the human driver, and the relatively larger the safety decision time SDT will be.

[0140] In specific implementation, as a preferred implementation manner of the present invention, the specific calculation process of the safety decision time SDT includes:

[0141] Limit the value range and interval of the remaining time TDV before the other ship TS intrudes into the own ship OS area, the remaining distance DDV before the other ship TS intrudes into the own ship OS area, and the distance D. Among them, the unit of the remaining time TDV before the other ship TS intrudes into the own ship OS area is hours, and the unit of D is nautical miles;

[0142] Divide the remaining time TDV before the other ship TS intrudes into the own ship OS area into six sub-intervals [TDV j , TDV j+1 ). For each group of the remaining time TDV before the other ship TS intrudes into the own ship OS area, according to the investigation and analysis results of the ship's driver, respectively count the proportion of the remaining time TDV before the other ship TS intrudes into the own ship OS area under different remaining distances DDV and distances D between ships;

[0143] Obtain the analytical formula of the safety decision time SDT through data fitting, and calculate the safety decision time SDT according to the actual independent variable values (specific values of DDV and D), as Figure 3 shown, specifically as follows:

[0144]

[0145] Among them, TDV j+1 represents the TDV value at time j + 1, and TDV j represents the TDV value at time j, and DDV j+1 represents the DDV value at time j + 1, and DDV j represents the DDV value at time j, and D j+1 represents the distance between two ships at time j + 1, and D j represents the distance between two ships at time j.

[0146] In specific implementation, as a preferred implementation manner of the present invention, in this embodiment, considering that the cognitive decision-making duration of humans reflects the degree of entanglement in judging things, therefore, the closer to the intermediate state of the independent variable, the larger the value of SDT should be. In addition, for independent variable values exceeding a certain limit, SDT should also consider the shortest decision-making time of humans, that is, the shortest duration for immediately making a collision avoidance decision trigger judgment. The present invention assumes that the numerical expression of the safe decision-making duration SDT is a special Gaussian-type smooth function, that is, a bump function, which is defined as:

[0147]

[0148] Among them, A represents the amplitude of the bump function, that is, the highest value at the center point, r represents the central radius of the bump function, (x 0 , y 0 ) represents the center point of the bump function, and d represents the minimum margin of the bump function;

[0149] Based on the numerical expression of the safe decision-making duration SDT, the SDT calculation formula SDT(DDV i , D i ) is as follows:

[0150] When , then there is:

[0151]

[0152] When , then there is:

[0153] SDT(DDV i , D i ) = SDT min

[0154] Among them, DDV 0 and D 0 represent the coordinates of the center point of the bump function; SDT maxRepresents the proportion of the safe decision-making time SDT in the remaining time TDV before the other ship TS intrudes into the domain of the own ship OS under the combination of the remaining distance DDV and the distance between ships D at the most entangled situation; SDT min Represents the proportion of the minimum decision trigger time for ship collision avoidance in the remaining time TDV before the other ship TS intrudes into the domain of the own ship OS; r SDT Represents the radius of the bump function; DDV e and D e ′ are the center and radius of the bump function respectively;

[0155] Since the interval range of the distance between ships D does not coincide with the remaining distance DDV before the other ship TS intrudes into the domain of the own ship OS, it is necessary to map the value of the distance between ships D from the interval [2, 6] to the interval [0, 1], specifically:

[0156] D′ = (D - 2) / 4.

[0157] In specific implementation, as a preferred implementation manner of the present invention, step S4 specifically includes:

[0158] S41. Construct a decision boundary, which is a function of TDV. When designing, refer to the Sigmoid activation function, and its characteristic is to compress a large numerical range into the interval (0, 1), and it is defined as follows:

[0159]

[0160] Among them, b 0 is a parameter for adjusting the boundary size, k is the sensitivity of the boundary to TDV, μ is the critical value of the boundary at the baseline value, k 0 is the minimum value for restricting the boundary;

[0161] S42. Calculate the decision trigger result based on the information accumulation amount The calculation formula is as follows:

[0162]

[0163] Among them, is the decision trigger result after intervention, X i represents the total amount of information accumulated at the current moment, B(TDV i ) represents the information boundary size at the current moment, μ is the proportionality factor, C 1 and C 2 are to make a collision avoidance decision and not to make a collision avoidance decision respectively;

[0164] S43. Calculate the decision trigger result based on the intervention scenario ξ i , and the calculation formula is as follows:

[0165]

[0166] Among them, the intervention scenario ξ i includes Condi_1 and Condi_2. Condi_1 is defined as the cumulative total duration of information at the current moment being greater than or equal to the current TISDT. Condi_1: Σt i ≥K i ; Condi_2 is defined as the total change amount between the α sampling moments before moment i and the current moment TISDT being less than or equal to the actual cumulative time. Condi_2: K i-α -K i ≤αΔt;

[0167] S44. When the intervention scenario ξ i satisfies Condi_1 and Condi_2, execute C 1 , otherwise execute C 2 .

[0168] Embodiment

[0169] Based on this verification platform, the present invention designs three different ship encounter scenarios to verify the rationality of the decision-making timing output by the VDDM-CADT model, specifically as Figure 4 shown, where:

[0170] Scenario I is a typical head-on encounter situation between two ships. The present invention sets TS-1 to continuously turn left during the encounter process to cross the starboard angle of OS;

[0171] Scenario II is a three-ship encounter scenario. OS forms crossing and head-on encounter situations with TS-1 and TS-2 respectively, but no action is taken by any of the three ships in this scenario;

[0172] Scenario III is a more complex four-ship encounter situation. In this scenario, OS forms right crossing, head-on, and left crossing encounter situations with TS-1, TS-2, and TS-3 respectively. Except for OS, the other three ships have taken different actions. TS-1 first turns left and then continuously turns right, and finally turns left to near the original course after passing and clearing. TS-2 first turns left and then turns right, and finally turns left back to the original course. TS-3 turns right initially and then turns left to resume the original course.

[0173] These three ship encounter scenarios respectively involve two ships, three ships, and four ships. Considering the experimental purpose of triggering the collision avoidance decision-making timing, the experimental design logic of the present invention based on the scenarios is as follows:

[0174] OS always sails at a constant course and speed, and its TS ships are operated by the driver (either taking actions or not taking actions). The initial parameter information of the ships in these three scenarios is asFigure 5 as shown Figure 5 It not only shows the initial motion information of each ship, but also records the duration of three encounter scenarios and the ship steering process.

[0175] Figures 6 - 8 They are the DDV and TDV parameter change diagrams for each scenario.

[0176] For Scenario I, the parameter change record is as Figure 6 shown. Since TS-1 took two actions during the head-on encounter, the DDV and TDV fluctuated, and the maximum DDV between the two ships was close to 1 when it was the largest;

[0177] For Scenario II, the parameter change record is as Figure 7 shown. Since neither TS-1, TS-2 nor OS took any actions, the fluctuation amplitude of the parameter change was not large. An interesting phenomenon is that the TDV value of TS-1 changed abruptly at the last moment because there was a second DDV due to the instability of the ship motion parameters, that is, there were two ship positions with DDV for TS-1 relative to OS during the encounter of the two ships;

[0178] For Scenario III, the parameter change record is as Figure 8 shown. Although OS did not take any actions, TS-1, TS-2 and TS-3 all took additional actions, so the parameter change curve was significantly affected.

[0179] In summary, through the analysis of the domain-based parameter changes between ships in the three scenarios, the following important conclusions can be obtained: when the ship does not take any actions (steering / speed change), due to the influence of the ocean environment, the ship's heading and speed will still have small disturbances, so the domain-based parameters between ships will also have small vibrations; when the ship takes actions, the domain intrusion parameters between ships will change significantly, and the smoothness of this change is related to the frequency of taking actions. If the interval is short, the parameter change is sharp, and if the interval is long, the parameter change is slow.

[0180] Figures 9 - 11 They are the information accumulation and boundary change record diagrams for each scenario.

[0181] For TS-1 in Scenario I, its information cumulative change curve is as Figure 9As shown. In the initial stage of information accumulation, the evidence accumulation between the ships quickly shifted in the direction of "not triggering a decision" and reached the trigger boundary for the first time. After that, the total amount of evidence accumulation between the ships returned to zero and began to be collected again. Since the OS had not taken any action, its information accumulation process continued to shift in the negative direction. During this process, TS-1 took a left turn to avoid, which caused the DDV between the two ships to increase sharply, so that the evidence accumulation value between the ships turned from negative to positive and shifted in the direction of "triggering a decision". Unfortunately, due to TS-1's left turn, the DDV between the two ships first increased and then decreased. Therefore, when the DDV decreased and stabilized, the amount of decision evidence accumulation between the ships also began to change slowly, and did not reach the trigger boundary until the last moment of the scene.

[0182] For scenario II, during the entire ship encounter process, OS, TS-1 and TS-2 did not take any action. The information accumulation change curves in this scenario are as follows: Figure 10 As shown in (a) and (b) in the figure. As can be seen from the figure, since all ships maintain direction and speed during the encounter, the information accumulation between ships is always biased towards the negative side. For TS-1 and TS-2, due to the difference in risk parameters relative to the OS, the frequency of evidence accumulation triggering boundaries is also different. Since TS-2 has a higher DDV value relative to TS-1, the evidence accumulation speed of TS-2 is slightly faster, that is, TS-1 is triggered three times, while TS-2 is only triggered twice. After the collision risk parameters tend to stabilize, the information accumulation process of the two ships relative to the OS also tends to stabilize.

[0183] Scenario III is a complex ship encounter situation. The information accumulation change curves in this scenario are as follows: Figure 11As shown in (a), (b), and (c) thereof. The OS encounters three TSSs, and the duration of the entire ship encounter scenario is relatively long (more than 30 minutes). The three TSSs all take a series of different avoidance maneuvers by the driver according to their personal cognitive results, including both compliance and non-compliance with the COLREGS. Among them, the longest natural decision trigger time is for TS-1. This is because during the movement process, TS-1 first slowly turns left and then accelerates to turn left, causing the DDV between the two ships to continuously increase, and the evidence accumulation also continuously increases. After that, TS-1 starts to turn right continuously, causing the domain parameter between the ships to rapidly decrease, and the information accumulation rapidly expands in the negative direction and reaches the negative decision boundary continuously and quickly for many times. In addition to TS-1, TS-2 is in a typical crossing situation with the OS at the initial moment. Due to the extremely large value of the inter-ship domain parameter, the information accumulation curve of TS-2 triggers the decision upper boundary continuously for many times at the initial moment. After that, TS-2 first takes a left-turn avoidance action, which causes the inter-ship domain parameter to decrease slightly. Subsequently, because TS-2 needs to avoid TS-1 on its left, it takes a right-turn maneuver. It should be noted that during this turning maneuver, the parameter value between TS-2 and the OS shows a short-term rapid increase process, and then quickly drops to 0 and remains at this value continuously. And because the model of the present invention adopts an information accumulation mechanism, there is no jump in the evidence accumulation amount during the information accumulation process. For TS-3, at the initial moment, it is on the port side of the OS and is the give-way vessel according to the provisions of the COLREGS. Therefore, TS-3 takes reasonable avoidance actions actively during the encounter process, that is, turns right to pass behind the OS. Since the initial domain parameter value between the OS and TS-3 is small, and TS-3 takes effective avoidance actions, its information accumulation value increases in the negative direction at the initial moment, and the information accumulation process is in a stable fluctuation stage after the parameter values tend to be stable.

[0184] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements 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 ship collision avoidance decision triggering timing modeling method that simulates the human social cognitive process, characterized in that: include: S1. Obtain ship AIS data, extract the latitude and longitude, heading, speed and length information of the ship in the AIS data, and construct an eccentric elliptical ship field; S2. Based on the constructed eccentric elliptical ship field, calculate the remaining distance DDV before the other ship TS invades the OS field of the own ship, the remaining time TDV before the other ship TS invades the OS field of the own ship, and the distance D between ships; S3, constructing a drift rate function f[i] to form the accumulated information per unit time and the accumulated total information at the current moment; S4. Construct a decision boundary and form a decision trigger result based on the accumulated information and intervention scenarios.

2. According to claim 1, a ship collision avoidance decision triggering timing modeling method that simulates the human social cognitive process is characterized by: In step S1, an eccentric elliptical ship field is constructed, which specifically includes: The major axis direction of the ellipse is parallel to the course of the ship's OS, the major semi-axis of the ellipse is a, the minor semi-axis of the ellipse is b, the displacement of the ship from the center of the ellipse along the semi-major axis to the stern is Δa, and the displacement of the ship from the center of the ellipse along the minor semi-axis to the left is Δb, and the values ​​of a, b, Δa and Δb are set as follows: a＝10l； b＝5l； Δa=2.5l; Δb=1.25l; Among them, l represents the captain of this ship.

3. The ship collision avoidance decision triggering timing modeling method that simulates the human social cognitive process according to claim 1 is characterized in that: Step S2 specifically includes: S21, based on the scaling factor f when the other ship TS intrudes into the OS area of ​​the ship to the maximum extent min , calculate the remaining distance DDV before the other ship TS invades the OS area of ​​the ship, the calculation formula is as follows: DDV=max(1-f min ,0) Among them, DDV is determined by the domain scaling factor f, and f is a quadratic function f(t) about time t. If f has a solution, the scaling factor f when the other ship TS invades the OS domain to the maximum extent is obtained: min , specifically, f min >1, it proves that the TS of other ships will never invade the OS area of ​​our ship; if f min =1, it means that the other ship TS just cuts across the boundary of the OS area; if f min <1 indicates that the TS of other ships will inevitably invade the area of ​​​​the OS of our ship; S22. Calculate the remaining time TDV before the other ship TS invades the OS area of ​​the own ship. The calculation formula is as follows: f(t)=1 Wherein, TDV is obtained by solving the formula f(t)=1.

4. The ship collision avoidance decision triggering timing modeling method according to claim 1 that simulates the human social cognitive process is characterized in that: Step S3 specifically includes: S31, drift rate includes theoretical instantaneous safety decision time TISDT and theoretical standard safety decision time TSSDT, where K i represents the safety decision-making time at any time i, Indicates the maximum safety decision time of the current TDV sub-interval. The calculation formula is as follows: K i =h ψ (DDV i ,D i ) Among them, h ψ is the analytical function of SDT, DDV i represents the DDV value at time i, D i represents the distance between the two ships at time i; S32. Calculate the evidence accumulation direction g[i] at time i. The calculation formula is as follows: Among them, DDV i crit express The corresponding DDV value; S33. Based on step S31 and step S32, the drift rate function f[i] is defined as follows: Where Δt is the time step, λ is the tuning factor, ε is a fixed constant, ∑t i is the cumulative decision time at the current moment i, |∑t i -K i | reflects the rate at which evidence accumulates; S34. Construct the information accumulation model of VDDM-CADT. The formula is as follows: ΔX[i]=X[i]-X[i-1] =ωf[i]+W[i] Where X[i] represents the total amount of information accumulated at the current time i, ΔX[i] represents the accumulated amount of information per unit time with a time step of Δt, where the unit of Δt is seconds, and W[i] represents the diffusion noise; assuming that the increment of the diffusion part obeys the standard Wiener process, it means that the change of the diffusion increment in any finite time obeys the normal distribution, which is expressed as W[i]~N(0,Δtσ 2 );ω represents the scale parameter, which reflects the proportion of the remaining information accumulation of the distance trigger decision threshold to the current departure trigger threshold. When the information accumulation value is closer to the trigger threshold at the current moment, the calculated ΔX[i] will be assigned a smaller scale factor; the calculation formula of the scale parameter ω is as follows: Among them, B i represents the information accumulation boundary at time i, α is the power function parameter, and α∈(0,1); S35, when T0 is set as the initial time of information accumulation, T is calculated with Δt as the unit time length. d The information accumulation at a certain moment is calculated as follows:

5. The ship collision avoidance decision triggering timing modeling method according to claim 4 that simulates the human social cognitive process is characterized in that: The theoretical instantaneous safety decision time TISDT and the theoretical standard safety decision time TSSDT in step S31 are both related to the safety decision time SDT, where SDT represents the upper limit of the collision avoidance decision triggering time of the ship driver under the premise of ensuring navigation safety, that is, the maximum time to determine whether to make a collision avoidance decision. The setting of the safety decision time SDT is related to the remaining distance DDV, the inter-ship distance D and TDV before the other ship TS invades the OS area of ​​the own ship, as follows: The remaining distance DDV before the other ship TS invades the OS area of ​​this ship and the distance between ships D indicate the urgency of the current situation. That is, the clearer the urgency of the current situation, the shorter the SDT will be. Conversely, the longer the safety decision time SDT will be. The longer the remaining time TDV before the other ship TS invades the OS area of ​​this ship, the more sufficient the safety decision time for the human driver will be, and the safety decision time SDT will also be relatively longer.

6. The ship collision avoidance decision triggering timing modeling method that simulates the human social cognitive process according to claim 5 is characterized in that: The specific calculation process of the safety decision time SDT includes: The remaining time TDV before the other ship TS invades the OS area of ​​the own ship, the remaining distance DDV before the other ship TS invades the OS area of ​​the own ship and the distance D limit the value range and interval, among which the remaining time TDV before the other ship TS invades the OS area of ​​the own ship is in hours, and the unit of D is in nautical miles; The remaining time TDV before the other ship TS invades the OS area of ​​the own ship is divided into six groups of sub-intervals [TDV j ,TDV j+1 ), for each group of the remaining time TDV before the TS of another ship invades the OS area of ​​the own ship, according to the survey and analysis results of the ship drivers, the remaining distance DDV before the TS of another ship invades the OS area of ​​the own ship and the proportion of the remaining time TDV before the TS of another ship invades the OS area of ​​the own ship under different ship-to-ship distances D are counted respectively; The analytical formula of the safety decision time SDT is obtained through data fitting, and the safety decision time SDT is calculated according to the actual value of the independent variable, as follows: Among them, TDV j+1 represents the TDV value at time j+1, TDV j Indicates the TDV value at time j, DDV j+1 represents the DDV value at time j+1, DDV j represents the DDV value at time j, D j+1 represents the distance between the two ships at time j+1, D j represents the distance between the two ships at time j.

7. The ship collision avoidance decision triggering timing modeling method according to claim 6 that simulates the human social cognitive process is characterized in that: Assume that the numerical expression of the safety decision time SDT is a special Gaussian smooth function, namely, the bulge function, which is defined as: Among them, A represents the amplitude of the bulge function, that is, the highest value at the center point, r represents the central radius of the bulge function, (x0, y0) represents the center point of the bulge function, and d represents the minimum margin of the bulge function; Based on the numerical expression of the safety decision time SDT, the SDT calculation formula SDT(DDV i ,D i ),as follows: when When , we have: when When , we have: SDT(DDV i ,D i )=SDT min Among them, DDV0 and D0 represent the coordinates of the center point of the bulge function; SDT max It indicates the ratio of the safe decision time SDT to the remaining time TDV before the other ship TS invades the OS area of ​​the own ship under the most entangled combination of the remaining distance DDV before the other ship TS invades the OS area of ​​the own ship and the distance between ships D; SDT min It indicates the ratio of the minimum decision triggering time for ship collision avoidance to the remaining time TDV before the other ship TS invades the OS area of ​​the own ship; r SDT Indicates the radius of the bulge function; DDV e and D e ′ are the center and radius of the bulge function respectively; Since the interval range of the inter-ship distance D is not consistent with the remaining distance DDV before the other ship TS invades the OS area of ​​the own ship, it is necessary to map the value of the inter-ship distance D from the interval [2,6] to the interval [0,1], specifically: D′=(D-2) / 4.

8. The ship collision avoidance decision triggering timing modeling method that simulates the human social cognitive process according to claim 1 is characterized in that: Step S4 specifically includes: S41. Construct a decision boundary, which is a function of TDV. The design refers to the Sigmoid activation function, which is characterized by compressing a large numerical range to the interval (0, 1). It is defined as follows: Among them, b0 is the parameter for adjusting the size of the boundary, k is the sensitivity of the boundary to TDV, μ is the critical value of the boundary at the baseline value, and k0 is the minimum value of the boundary limit; S42. Calculate decision trigger results based on information accumulation The calculation formula is as follows: in, is the decision trigger result after intervention, X i Indicates the total amount of information accumulated at the current moment, B(TDV i ) represents the size of the information boundary at the current moment, μ is the scale factor, C1 and C2 are respectively making collision avoidance decisions and not making collision avoidance decisions; S43, based on intervention scenario i , calculate the decision trigger result, the calculation formula is as follows: Among them, the intervention scenario i It includes Condi_1 and Condi_2. Condi_1 is defined as the total accumulated time of information at the current moment is greater than or equal to the current TISDT. Condi_1:∑t i ≥K i ; Condi_2 is defined as the total change between the α sampling moments before time i and the current moment TISDT is less than or equal to the actual accumulated time, Condi_2:K i-α -K i ≤αΔt; S44, when the intervention scene i When Condi_1 and Condi_2 are satisfied, C1 is executed, otherwise C2 is executed.

Images