Multi-target game tracking path planning controller and method considering safety optimization
Through the combination of intelligent ship kinematics module, sequential game optimization planner module and segmented polynomial path optimization module, combined with noise interference to construct control obstacle function, the pursuit path planning of multiple dynamic target ships in restricted waters is solved. The security and accuracy issues achieve optimal path planning and robust security.
Patent Information
- Application Number
- CN202410974557.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-08
AI Technical Summary
The existing intelligent ship path planning method is difficult to effectively plan the pursuit path of multiple dynamic target ships in restricted waters, and fails to fully consider noise interference and kinematic constraints, resulting in unsafe and inaccurate planning results.
A multi-objective game pursuit path planning controller considering safety optimization was designed. Through the combination of intelligent ship kinematics module, sequential game optimization planner module, segmented polynomial path optimization module and control obstacle function safety optimization module, it uses Optimal iterative response algorithm and sensitivity analysis, combining noise interference to construct control obstacle functions, optimize path planning to ensure safety and accuracy.
On the premise of ensuring safety and collision-free, the optimal path planning is achieved in restricted waters, reducing the number of iterations of path planning, improving the planning speed and accuracy, and ensuring robust safety in the case of noise interference sex.
Smart Images

Figure CN118915455A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of intelligent ship path planning, and in particular to a multi-objective game overtaking path planning controller considering safety optimization. Background Art
[0002] The 21st century is the century of the ocean, which is rich in biological resources, oil and gas resources, and mineral resources, and is an important material basis for human survival. Among marine resources, the shipping industry has become an important part of economic development by virtue of its advantages in transportation capacity, distance, and cost compared with air and land transportation. With the continuous development of the information and intelligentization process in various industries and fields, especially the continuous advancement of technologies such as the Internet and artificial intelligence, smart ships have gradually become the main means to enhance the overall competitiveness of the shipping industry. In view of the current problems in the shipping industry, such as the long sailing time for mission execution and the high number of safety accidents caused by manual operation of crew members, new requirements have been put forward for the application of smart ships in the shipping industry.
[0003] Smart ships need to consider multiple constraints to safely and effectively plan their paths to overtake multiple target ships in restricted waters, such as the distance from the boundaries of the waterways on both sides during navigation and the minimum safe distance from the target ships. At the same time, the planning results depend on the state information of the smart ship, and it is also necessary to consider the safety issues caused by possible noise interference; as well as the timeliness of planning the optimal overtaking path for multiple target ships with uncertainty. In order to solve the problem of collision-free path planning of intelligent agents in overtaking scenarios, scholars in related fields have proposed a variety of path planning methods based on game theory. D. Sadigh proposed a two-party game theory between artificially controlled cars and intelligent cars by using the cost function in inverse reinforcement learning for intelligent cars. By training and learning the operation of drivers in different scenarios, the path planning of artificially controlled cars was realized. A. Liniger proposed a method of using a limited pre-learned operation set for the problem of collaborative path planning between two intelligent cars. Through Nash equilibrium and Stackelberg equilibrium, a dual-matrix game of two intelligent cars was established for path planning, realizing collaborative path planning of intelligent cars. D.Zhou proposed a buffered Voronoi unit graph method using position measurement for the problem of free communication of multi-agents, which achieved collision avoidance of agents against static obstacles in complex road environments. However, most of the above-mentioned related studies are based on known operation sets and training and learning, but in different restrictive waters, the previously trained and learned data sets are not applicable, which adds inaccuracy to the planning results.
[0004] In the existing path planning methods for intelligent ships in restricted waters, there are still the following problems: First: In the existing path planning methods for intelligent entities, more consideration is given to path planning for avoiding collisions with static obstacles, which cannot be extended to path planning for overtaking multiple dynamic target ships. Second, in the existing path planning methods for intelligent ships, less consideration is given to the problem of noise disturbance in state information. In the process of path planning, the accuracy of state information is relied on, resulting in the inability to guarantee safety of the planning results. Third: In the existing path planning methods for intelligent ships, the global path planning algorithm and the kinematic model are not coupled using relevant constraints, resulting in the possibility that the planned path is not optimal. Summary of the invention
[0005] In order to solve the problems of the prior art, the present invention provides a multi-objective game overtaking path planning controller and method considering safety optimization, which realizes that when the intelligent ship is disturbed by noise, the optimal overtaking path can be planned in restricted waters under the premise of ensuring safety and collision-free.
[0006] The technical means adopted by the present invention are as follows:
[0007] A multi-objective game overtaking path planning controller considering safety optimization, comprising:
[0008] An intelligent ship kinematics module, wherein the intelligent ship kinematics module receives the forward speed from the piecewise polynomial path optimization module and the bow angular velocity from the obstacle control function safety optimization module, and the intelligent ship kinematics module outputs state information to the sequential game optimization planner module and the obstacle control function safety optimization module; the intelligent ship kinematics module outputs the bow angular velocity information to the obstacle control function safety optimization module;
[0009] A sequential game optimization planner module, wherein the sequential game optimization planner module receives status information from the intelligent ship kinematics module and status information from multiple target ships; the sequential game optimization planner module outputs decision variables and path parameters in the horizontal and vertical directions to the piecewise polynomial path optimization module;
[0010] A piecewise polynomial path optimization module, wherein the piecewise polynomial path optimization module receives the output decision variables and path parameters in the horizontal and vertical directions from the sequential game optimization planner module; the piecewise polynomial path optimization module outputs the optimized forward speed to the intelligent ship kinematics module; the piecewise polynomial path optimization module outputs the optimized forward speed and the optimized bow angular velocity to the obstacle control function safety optimization module;
[0011] A control obstacle function safety optimization module receives the optimized bow angular velocity and forward velocity from the piecewise polynomial path optimization module and the position information and bow angular velocity of the intelligent ship kinematics module; the control obstacle function safety optimization module outputs the bow angular velocity to the intelligent ship kinematics module.
[0012] Furthermore, in the intelligent ship kinematics module:
[0013] The input of the intelligent ship kinematics module is the forward velocity U i , yaw angular velocity r i And the noise interference ω to which the smart ship is subjected; the kinematic module of the smart ship is as follows:
[0014]
[0015] Among them, [x i ,y i ,ψ i ] T are the longitudinal position, transverse position and bow angle of the intelligent ship respectively; the output of the intelligent ship kinematics module is the state information of the intelligent ship [x i ,y i ,ψ i ] T and the yaw angular velocity r i .
[0016] Furthermore, in the sequential game optimization planner module:
[0017] The input of the sequential game optimization planner module is the state information of the intelligent ship [x i ,y i ,ψ i ] T and the status information of multiple target ships to be overtaken [x j ,y j ,ψ j ] T ;[x j ,y j ,ψ j ] T are the longitudinal position, transverse position and pitch angle of the target ship respectively, the subscript j=1,...,k is the sequence index of k different target ships to be overtaken, and k is the total number index of target ships; the sequential game optimization planner module is as follows:
[0018]
[0019] sth eq (Θ i ,A i ,Bi )=0,
[0020] g ineq (Θ i ,A i ,B i )≤0,
[0021] γ(Θ i ,Θ j )≤0. (2)
[0022] in, β ij is an adjustable parameter reflecting the degree of coupling, λ ij The Lagrange multiplier representing the degree of association in the constraint condition; n = 1, 2..., N is the step index of the global planning, and N is the total number of steps of the global planning; is the decision variable of the smart ship, is the position of the smart ship under different step indexes, is the speed of the smart ship at different indexes; is the decision variable of the target ship, is the position of the jth target ship to be overtaken under different step indexes, is the speed of the jth target ship to be overtaken at different indexes; represents the polynomial path fitting parameters in the longitudinal direction, represents the polynomial path fitting parameters in the horizontal direction, m represents the order of the polynomial; h eq (Θ i ,A i ,B i ) is an equality constraint, which is equivalent to Indicates the relationship between the position and speed corresponding to the step index of the current and previous step, Δt is the step length; g ineq (Θ i ,A i ,B i ) are inequality constraints including: and τ is the central channel of the second-order parametric fitting, n = τ″ / ||τ″|| is the unit normal vector of the central channel, W represents the distance between the central channel and the channel boundary, is the shortest distance between the smart ship and the central channel under the current step index, U max represents the maximum speed limit of the smart ship, r max is the maximum bow angular velocity of the intelligent ship; γ(Θ i ,Θ j ) is the interaction inequality constraint, which is equivalent to the minimum safe collision avoidance constraint between the intelligent ship and the jth target ship to be overtaken dmin represents the minimum safe collision avoidance distance between the smart ship and the jth target ship to be overtaken; the output of the sequential game optimization planner module is the decision variable Θ of the smart ship i and the horizontal and vertical planning path parameters A i ,B i .
[0023] Furthermore, in the piecewise polynomial path optimization module:
[0024] The input of the piecewise polynomial path optimization module is the decision variable Θ of the intelligent ship i and the horizontal and vertical planning path parameters A i ,B i , the piecewise polynomial path optimization module is as follows:
[0025]
[0026] Among them, t∈[(n-1)Δt,nΔt] is the time domain interval corresponding to the step index; the output of the piecewise polynomial path optimization module is the forward speed U i and the optimized yaw angular velocity r opt .
[0027] Furthermore, in the control barrier function safety optimization module:
[0028] The input of the control obstacle function safety optimization module is the optimized bow angular velocity r opt , the status information of the smart ship[x i ,y i ,ψ i ] T and the yaw angular velocity r i , forward speed U i , and the status information of the jth target ship to be overtaken [x j ,y j ,ψ j ] T and speed U j , and the existing noise disturbance ω, the control barrier function is safely optimized as follows:
[0029]
[0030] r i ≤|r max |. (7)
[0031] in, The control obstacle function constructed considering noise disturbance, η i =[x i ,y i ,ψ i ]T is the status information of the smart ship, p i =[x i ,y i ] T is the current position of the smart ship, p j =[x j ,y j ] T is the current position of the jth target ship to be overtaken, ω≤||ω max || is the noise disturbance, ω max is the maximum value of the noise disturbance, α(·) is Class equation, Respectively In (p i ,p j ) on f(p i ,p j ) and g(p i ,p j )’s Lie derivative, U j represents the forward combined velocity of the target ship, ψ j represents the pitch angle of the target ship; the output of the safety optimization of the control obstacle function is the pitch angular velocity r i .
[0032] The present invention also provides a multi-objective game overtaking path planning method considering safety optimization, which is implemented based on any one of the multi-objective game overtaking path planning controllers considering safety optimization, including the following steps:
[0033] S1, the sequential game optimization planner module receives the state information of the intelligent ship [x i ,y i ,ψ i ] T and the status information of multiple target ships to be overtaken [x j ,y j ,ψ j ] T After performing mathematical processing of secondary optimization and game iteration, the decision variable Θ is generated. i and the path parameters A in both horizontal and vertical directions i ,B i to the piecewise polynomial path optimization module;
[0034] S2, the piecewise polynomial path optimization module receives the decision variable Θ of the game optimization planner i and the path parameters A in both horizontal and vertical directions i ,B i , and convert it mathematically to generate the forward velocity U iTo the intelligent ship and control obstacle function safety optimization module, optimize the bow angular velocity r opt To the control barrier function safety optimization module;
[0035] S3, the obstacle control function safety optimization module receives the optimized bow angular velocity r from the piecewise polynomial path optimization module. opt , forward speed U i , the status information of the smart ship[x i ,y i ,ψ i ] T and the yaw angular velocity r i , external noise interference ω, and combined with the status information of the jth target ship to be overtaken [x j ,y j ,ψ j ] T Optimize the control obstacle function and finally generate the bow angular velocity r i to smart ships;
[0036] S4, the intelligent ship kinematics module receives the forward speed U of the piecewise polynomial path optimization module i , the yaw angular velocity r of the obstacle function safety optimization module i and external noise interference ω, and the state information of the smart ship [x i ,y i ,ψ i ] T Output to the sequential game optimization planner and the control obstacle function safety optimization module, the bow angular velocity r i Output to the control barrier function safety optimization module.
[0037] The present invention also provides a storage medium, which includes a stored program, wherein when the program is running, the multi-objective game overtaking path planning method considering safety optimization is executed.
[0038] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the multi-objective game overtaking path planning method considering safety optimization through the computer program.
[0039] Compared with the prior art, the present invention has the following advantages:
[0040] First, in the existing path planning methods for intelligent ships, overtaking dynamic target ships is rarely considered; the present invention combines sensitivity analysis with the best iterative response algorithm to design a sequential game optimization planner, in which the s(p) term is the best iterative response algorithm, and the ∑ part is the sensitivity analysis part. The game optimization algorithm uses a sequential calculation planning method to construct a two-party non-cooperative game between the intelligent ship and multiple target ships to be overtaken, which can reduce the total number of path planning iterations under multiple constraints and speed up the overall planning speed.
[0041] Second, in the existing path planning methods for intelligent ships, there are relatively few safety issues caused by noise interference; the present invention proposes to use noise interference to construct a control obstacle function that meets the conditions to achieve robust safety optimization, which not only meets the safety constraints, but also ensures the robust safety of the planned path when the ship status information is disturbed.
[0042] Third: In the existing path planning method of intelligent ships, the ship kinematic model is combined with the planned path; while the piecewise polynomial path method proposed in the present invention incorporates the radian of the path (the bow roll angle of the ship) as a constraint condition into the algorithm of the sequential game optimization planner, parameterizes the planned path, directly converts the state constraints of the ship into a polynomial function, generates segmented paths corresponding to different step indexes, and improves the accuracy of path planning. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0044] Figure 1 It is the overall flow chart of the present invention.
[0045] FIG. 2 is a moment state diagram of the present invention overtaking multiple target ships in restricted waters. Figure 2a This is the state diagram for the 10th second; Figure 2b This is the state diagram for the 50th second; Figure 2c This is the state diagram at the 100th second; Figure 2d This is the status diagram at 200 seconds.
[0046] FIG. 3 is a state diagram showing a comparison of the timing of sequential calculation and parallel calculation planning of the present invention. Figure 3a It is a schematic diagram of sequential computing planning; Figure 3b This is a schematic diagram of parallel computing planning.
[0047] FIG4 is a diagram showing the longitudinal and transverse position changes of the intelligent ship of the present invention. Figure 4a The vertical position represents the change along the x-axis direction, Figure 4b The horizontal position represents the change along the y-axis.
[0048] Figure 5 is the bow rolling angle ψ of the intelligent ship of the present invention i 's effect diagram.
[0049] Figure 6 is the bow angular velocity r of the intelligent ship of the present invention i 's effect diagram.
[0050] Figure 7 is the forward speed U of the intelligent ship of the present invention i With the maximum speed limit U max 's comparison.
[0051] Figure 8 This is a rendering of the distance between the smart ship and the target ship of the present invention.
[0052] Fig. 9 This is a rendering of the distance between the intelligent ship of the present invention and the boundary of the waterway. DETAILED DESCRIPTION
[0053] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.
[0054] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0055] The s(p) term is the optimal iterative response algorithm, and the ∑ part is the sensitivity analysis part
[0056] like Figure 1As shown, the present invention provides a multi-objective game overtaking path planning controller considering safety optimization, an intelligent ship kinematics module, a sequential game optimization planner module, a piecewise polynomial path optimization module and a control obstacle function safety optimization module.
[0057] The input end of the intelligent ship kinematics module is connected to the output end forward speed of the piecewise polynomial path optimization module and the bow angular velocity of the control obstacle function safety optimization module, and the output end state information of the intelligent ship kinematics module is connected to the input end of the sequential game optimization planner module and the input end of the control obstacle function safety optimization module. The input end of the sequential game optimization planner module is also connected to the state information of multiple target ships, and the output of the sequential game optimization planner module is the decision variable of the intelligent ship and the path parameters in the horizontal and vertical directions. The input end of the piecewise polynomial path optimization module is connected to the output of the sequential game optimization planner module, and the output end of the piecewise polynomial path optimization module is connected to the input end of the control obstacle function safety optimization module and the input end of the intelligent ship kinematics module. The input end of the control obstacle function safety optimization module is connected to the output end of the piecewise polynomial path optimization module to optimize the bow angular velocity, and is also connected to the bow angular velocity of the intelligent ship kinematics module.
[0058] A: Intelligent Ship Kinematics Module
[0059] The input of the module is the forward velocity U i , yaw angular velocity r i And the noise interference to smart ships; the modules are as follows:
[0060]
[0061] Among them, [x i ,y i ,ψ i ] T are the longitudinal position, transverse position and pitch angle of the smart ship respectively; the output of the module is the status information of the smart ship [x i ,y i ,ψ i ] T and the yaw angular velocity r i .
[0062] B: Sequential Game Optimization Planner Module
[0063] The input of the module is the state information of the intelligent ship [x i ,y i ,ψ i ] T and the status information of multiple target ships to be overtaken [x j ,y j,ψ j ] T ,[x j ,y j ,ψ j ] T are the longitudinal position, transverse position and pitch angle of the target ship respectively, the subscript j=1,...,k is the sequence index of k different target ships to be overtaken, and k is the total number index of target ships; the modules are as follows:
[0064]
[0065] sth eq (Θ i ,A i ,B i )=0,
[0066] g ineq (Θ i ,A i ,B i )≤0,
[0067] γ(Θ i ,Θ j )≤0. (2)
[0068] in, β ij is an adjustable parameter reflecting the degree of coupling, λ ij The Lagrange multiplier representing the degree of association in the constraint condition; n = 1, 2..., N is the step index of the global planning, and N is the total number of steps of the global planning; is the decision variable of the smart ship, is the position of the smart ship under different step indexes, is the speed of the smart ship at different indexes; is the decision variable of the target ship, is the position of the jth target ship to be overtaken under different step indexes, is the speed of the jth target ship to be overtaken at different indexes; represents the polynomial path fitting parameters in the longitudinal direction, represents the polynomial path fitting parameters in the horizontal direction, m represents the order of the polynomial; h eq (Θ i ,A i ,B i ) is an equality constraint, which is equivalent to Indicates the relationship between the position and speed corresponding to the step index of the current and previous step, Δt is the step length; g ineq (Θ i ,A i ,B i) are inequality constraints including: and τ is the central channel of the second-order parametric fitting, n = τ″ / ||τ″|| is the unit normal vector of the central channel, W represents the distance between the central channel and the channel boundary, is the shortest distance between the smart ship and the central channel under the current step index, U max represents the maximum speed limit of the smart ship, r max is the maximum bow angular velocity of the intelligent ship; γ(Θ i ,Θ j ) is the interaction inequality constraint, which is equivalent to the minimum safe collision avoidance constraint between the intelligent ship and the jth target ship to be overtaken d min represents the minimum safe collision avoidance distance between the smart ship and the jth target ship to be overtaken; the output of the module is the decision variable Θ of the smart ship i and the horizontal and vertical planning path parameters A i ,B i .
[0069] C: Piecewise Polynomial Path Optimization Module
[0070] The input of the path optimization module is the decision variable Θ of the intelligent ship i and the horizontal and vertical planning path parameters A i ,B i , the modules are as follows:
[0071]
[0072] Among them, t∈[(n-1)Δt,nΔt] is the time domain interval corresponding to the step index; the output of the module is the forward speed U i and the optimized yaw angular velocity r opt .
[0073] D. Control barrier function safety optimization module
[0074] The input of the module is the optimized yaw angular velocity r opt , the status information of the smart ship[x i ,y i ,ψ i ] T and the yaw angular velocity r i , forward speed U i , and the status information of the jth target ship to be overtaken [x j ,y j ,ψ j ] T and speed U j , and the noise disturbance ω exists, the module is as follows:
[0075]
[0076] r i ≤|r max |. (7)
[0077] in, The control obstacle function constructed considering noise disturbance, η i =[x i ,y i ,ψ i ] T is the status information of the smart ship, p i =[x i ,y i ] T is the current position of the smart ship, p j =[x j ,y j ] T is the current position of the jth target ship to be overtaken, ω≤||ω max || is the noise disturbance, ω max is the maximum value of the noise disturbance, α(·) is Class equation, Respectively In (p i ,p j ) on f(p i ,p j ) and g(p i ,p j )’s Lie derivative, U j represents the forward combined velocity of the target ship, ψ j Represents the pitch angle of the target ship; the output of the module is the pitch angular velocity r i .
[0078] The present invention also provides a multi-objective game overtaking path planning method considering safety optimization, which is implemented based on a multi-objective game overtaking path planning controller considering safety optimization, and includes the following steps:
[0079] S1, the sequential game optimization planner module receives the state information of the intelligent ship [x i ,y i ,ψ i ] T and the status information of multiple target ships to be overtaken [x j ,y j ,ψ j ] T After performing mathematical processing of secondary optimization and game iteration, the decision variable Θ is generated. iand the path parameters A in both horizontal and vertical directions i ,B i to the piecewise polynomial path optimization module;
[0080] S2, the piecewise polynomial path optimization module receives the decision variable Θ of the game optimization planner i and the path parameters A in both horizontal and vertical directions i ,B i , and convert it mathematically to generate the forward velocity U i To the intelligent ship and control obstacle function safety optimization module, optimize the bow angular velocity r opt To the control barrier function safety optimization module;
[0081] S3, the obstacle control function safety optimization module receives the optimized bow angular velocity r from the piecewise polynomial path optimization module. opt , forward speed U i , the status information of the smart ship[x i ,y i ,ψ i ] T and the yaw angular velocity r i , external noise interference ω, and combined with the status information of the jth target ship to be overtaken [x j ,y j ,ψ j ] T Optimize the control obstacle function and finally generate the bow angular velocity r i to smart ships;
[0082] S4, the intelligent ship kinematics module receives the forward speed U of the piecewise polynomial path optimization module i , the yaw angular velocity r of the obstacle function safety optimization module i and external noise interference ω, and the state information of the smart ship [x i ,y i ,ψ i ] T Output to the sequential game optimization planner and the control obstacle function safety optimization module, the bow angular velocity r i Output to the control barrier function safety optimization module.
[0083] The present invention also provides a storage medium, which includes a stored program, wherein when the program is running, a multi-objective game overtaking path planning method considering safety optimization is executed.
[0084] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes a multi-objective game overtaking path planning method taking safety optimization into consideration through the computer program.
[0085] The kinematic design of the intelligent ship is as shown in formula (1), where the speed information of the intelligent ship is i The constraints include: forward speed U i ≤|U max |=1(m / s), yaw angular velocity r i ≤|r max |=0.5(rad / s).
[0086] The sequential game optimization planner is designed as shown in formula (2), where the coupling coefficient between the smart ship and each target ship is: β ij ∈[0,1], when β ij = 0, it means that the target equation is not affected by other target ships to be overtaken. Similarly, β ij The larger the value is, the more competitive the overtaking game process is. The minimum safe collision avoidance distance d between ships in the constraint condition is min ≥10(m), the distance W between the smart ship and the channel boundary is ≤30(m).
[0087] The piecewise polynomial path optimization module is designed as shown in formula (3), and the piecewise polynomial path corresponding to each step index can be obtained by Indicates that when the planned path is a straight line, then x n =A n,1 t+A n,0 ,y n =B n,1 t+B n,0 When performing the overtaking task and the planned path is expressed as a curve, the nonlinear relationship is as follows:
[0088] x n =A n,2 t 2 +A n,1 t+A n,0 ,y n =B n,2 t 2 +B n,1 t+B n,0 .
[0089] The design purpose of the present invention is to enable the intelligent ship path planning system to calculate the positions of the intelligent ship and multiple target ships by the sequential game optimization planner through the best iterative response algorithm when satisfying equations (1)-(7), and to use judgment statements in the algorithm to determine different target ships, output the optimization instruction position information for the intelligent ship, and then process the state information of the intelligent ship through the piecewise polynomial path optimization module to obtain the control input required by the intelligent ship, that is, to ensure that the intelligent ship can realize global overtaking path planning for multiple target ships in restricted waters and ensure collision avoidance.
[0090] Figure 2 is a rendering of an overtaking scenario in restricted waters, where (a), (b), (c), and (d) respectively represent the moment-by-moment status effects of tracking, overtaking, and completion of overtaking of the first and second target ships (the white ship is the smart ship, and the black ship is the target ship).
[0091] Figure 3 is a comparison of the planning effects of the sensitivity-enhanced sequential game optimization planner module using sequential calculation and parallel calculation (the white ship is an intelligent ship, and the black ship is a target ship). It can be seen that sequential calculation greatly reduces the number of iterative planning times.
[0092] FIG4 is a diagram showing the position change effect of an intelligent ship, where (a) the longitudinal position represents the change along the x-axis, and (b) the lateral position represents the change along the y-axis.
[0093] Figure 5 is the yaw angle ψ of the smart ship i The effect diagram of , whose changes match the moment-by-moment overtaking effect diagrams in Figure 2 (a)-(d).
[0094] Figure 6 is the forward speed U of the smart ship i With the maximum speed limit U max The effect diagram, forward speed U i With the maximum speed limit U max From the rendering, it can be seen that in the overtaking scenario, the intelligent ship moves at the maximum restricted speed.
[0095] Figure 7 is the angular velocity of the intelligent ship r i Effect diagram, r i The pitch angle ψ is described by i Rate of change.
[0096] Figure 8 It is a rendering of the distance between the smart ship and the target ship, describing the change in the distance between the two ships and ensuring that it is greater than the minimum safety distance.
[0097] Fig. 9It is a rendering of the distance between the smart ship and the channel boundary, which describes that during the overall overtaking process of the smart ship, the distance between the channel boundaries on both sides is within the safe distance.
[0098] It can be seen from the simulation results that the multi-objective game overtaking path planning design method considering safety optimization proposed in the present invention, by introducing the kinematic model of underactuated ships, using the game optimization algorithm combining the best iterative response algorithm with sensitivity analysis, and using the clear convergence conditions and convergence of the iterative algorithm, enables the intelligent ship to continuously approximate and calculate the path of the multi-objective ship online and achieve overtaking. At the same time, the piecewise polynomial path planning method is combined with the kinematic model of underactuated ships to optimize each time period and improve the accuracy of path planning. During planning, the planned path is planned using sequential calculation, which reduces the total number of iterative planning, greatly improves the planning speed, and effectively increases the accuracy of the simulation experiment effect. Finally, when the kinematic model of the ship has uncertainties such as perturbations, a new control obstacle function is constructed by using noise disturbances to eliminate the safety problems caused by noise interference and improve the robust safety of the design.
[0099] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, 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 multi-objective game overtaking path planning controller considering safety optimization, characterized in that: include: An intelligent ship kinematics module, wherein the intelligent ship kinematics module receives the forward velocity from the piecewise polynomial path optimization module and the bow angular velocity from the obstacle control function safety optimization module, and the intelligent ship kinematics module outputs state information to the sequential game optimization planner module and the obstacle control function safety optimization module; The intelligent ship kinematics module outputs the bow angular velocity information to the control obstacle function safety optimization module; A sequential game optimization planner module, wherein the sequential game optimization planner module receives status information from the intelligent ship kinematics module and status information from multiple target ships; the sequential game optimization planner module outputs decision variables and path parameters in the horizontal and vertical directions to the piecewise polynomial path optimization module; A piecewise polynomial path optimization module, wherein the piecewise polynomial path optimization module receives the output decision variables and path parameters in the horizontal and vertical directions from the sequential game optimization planner module; the piecewise polynomial path optimization module outputs the optimized forward speed to the intelligent ship kinematics module; the piecewise polynomial path optimization module outputs the optimized forward speed and the optimized bow angular velocity to the obstacle control function safety optimization module; A control obstacle function safety optimization module receives the optimized bow angular velocity and forward velocity from the piecewise polynomial path optimization module and the position information and bow angular velocity of the intelligent ship kinematics module; the control obstacle function safety optimization module outputs the bow angular velocity to the intelligent ship kinematics module.
2. The multi-objective game overtaking path planning controller considering safety optimization according to claim 1 is characterized in that: In the intelligent ship kinematics module: The input of the intelligent ship kinematics module is the forward velocity U i , yaw angular velocity r i And the noise interference ω to which the smart ship is subjected; the kinematic module of the smart ship is as follows: Among them, [x i ,y i ,ψ i ] T are the longitudinal position, transverse position and bow angle of the intelligent ship respectively; the output of the intelligent ship kinematics module is the state information of the intelligent ship [x i ,y i ,ψ i ] T and the yaw angular velocity r i .
3. The multi-objective game overtaking path planning controller considering safety optimization according to claim 1 is characterized in that: In the sequential game optimization planner module: The input of the sequential game optimization planner module is the state information of the intelligent ship [x i ,y i ,ψ i ] T and the status information of multiple target ships to be overtaken [x j ,y j ,ψ j ] T ;[x j ,y j ,ψ j ] T are the longitudinal position, transverse position and pitch angle of the target ship respectively, the subscript j=1,...,k is the sequence index of k different target ships to be overtaken, and k is the total number index of target ships; the sequential game optimization planner module is as follows: s.t.h eq (Θ i ,A i ,B i )=0, g ineq (I i ,A i ,B i )≤0, c(Θ i ,I j )≤0. (2) in, β ij is an adjustable parameter reflecting the degree of coupling, λ ij The Lagrange multiplier representing the degree of association in the constraint condition; n = 1, 2..., N is the step index of the global planning, and N is the total number of steps of the global planning; is the decision variable of the smart ship, is the position of the smart ship under different step indexes, is the speed of the smart ship at different indexes; is the decision variable of the target ship, is the position of the jth target ship to be overtaken under different step indexes, is the speed of the jth target ship to be overtaken at different indexes; represents the polynomial path fitting parameters in the longitudinal direction, represents the polynomial path fitting parameters in the horizontal direction, m represents the order of the polynomial; h eq (Θ i ,A i ,B i ) is an equality constraint, which is equivalent to Indicates the relationship between the position and speed corresponding to the step index of the current and previous step, Δt is the step length; g ineq (Θ i ,A i ,B i ) are inequality constraints including: and τ is the central channel of the second-order parametric fitting, n = τ″ / ||τ″|| is the unit normal vector of the central channel, W represents the distance between the central channel and the channel boundary, is the shortest distance between the smart ship and the central channel under the current step index, U max represents the maximum speed limit of the smart ship, r max is the maximum bow angular velocity of the intelligent ship; γ(Θ i ,Θ j ) is the interaction inequality constraint, which is equivalent to the minimum safe collision avoidance constraint between the intelligent ship and the jth target ship to be overtaken d min represents the minimum safe collision avoidance distance between the smart ship and the jth target ship to be overtaken; the output of the sequential game optimization planner module is the decision variable Θ of the smart ship i and the horizontal and vertical planning path parameters A i ,B i .
4. The multi-objective game overtaking path planning controller considering safety optimization according to claim 1 is characterized in that: In the piecewise polynomial path optimization module: The input of the piecewise polynomial path optimization module is the decision variable Θ of the intelligent ship i and the horizontal and vertical planning path parameters A i ,B i , the piecewise polynomial path optimization module is as follows: Among them, t∈[(n-1)Δt,nΔt] is the time domain interval corresponding to the step index; the output of the piecewise polynomial path optimization module is the forward speed U i and the optimized yaw angular velocity r opt .
5. The multi-objective game overtaking path planning controller considering safety optimization according to claim 1, characterized in that: In the control barrier function safety optimization module: The input of the control obstacle function safety optimization module is the optimized bow angular velocity r opt , the status information of the smart ship[x i ,y i ,ψ i ] T and the yaw angular velocity r i , forward speed U i , and the status information of the jth target ship to be overtaken [x j ,y j ,ψ j ] T and speed U j , and the existing noise disturbance ω, the control barrier function is safely optimized as follows: in, The control obstacle function constructed considering noise disturbance, η i =[x i ,y i ,ψ i ] T is the status information of the smart ship, p i =[x i ,y i ] T is the current position of the smart ship, p j =[x j ,y j ] T is the current position of the jth target ship to be overtaken, ω≤||ω max || is the noise disturbance, ω max is the maximum value of the noise disturbance, α(·) is Class equation, Respectively In (p i ,p j ) on f(p i ,p j ) and g(p i ,p j )’s Lie derivative, U j represents the forward combined velocity of the target ship, ψ j represents the pitch angle of the target ship; the output of the safety optimization of the control obstacle function is the pitch angular velocity r i .
6. A multi-objective game overtaking path planning method considering safety optimization, which is implemented based on the multi-objective game overtaking path planning controller considering safety optimization as described in any one of claims 1 to 5, characterized in that: The steps include: S1, the sequential game optimization planner module receives the state information of the intelligent ship [x i ,y i ,ψ i ] T and the status information of multiple target ships to be overtaken [x j ,y j ,ψ j ] T After performing mathematical processing of secondary optimization and game iteration, the decision variable Θ is generated. i and the path parameters A in both horizontal and vertical directions i ,B i to the piecewise polynomial path optimization module; S2, the piecewise polynomial path optimization module receives the decision variable Θ of the game optimization planner i and the path parameters A in both horizontal and vertical directions i ,B i , and convert it mathematically to generate the forward velocity U i To the intelligent ship and control obstacle function safety optimization module, optimize the bow angular velocity r opt To the control barrier function safety optimization module; S3, the obstacle control function safety optimization module receives the optimized bow angular velocity r from the piecewise polynomial path optimization module. opt , forward speed U i , the status information of the smart ship[x i ,y i ,ψ i ] T and the yaw angular velocity r i , external noise interference ω, and combined with the status information of the jth target ship to be overtaken [x j ,y j ,ψ j ] T Optimize the control obstacle function and finally generate the bow angular velocity r i to smart ships; S4, the intelligent ship kinematics module receives the forward speed U of the piecewise polynomial path optimization module i , the bow angular velocity r of the obstacle function safety optimization module i and external noise interference ω, and the state information of the smart ship [x i ,y i ,ψ i ] T Output to the sequential game optimization planner and the control obstacle function safety optimization module, the bow angular velocity r i Output to the control barrier function safety optimization module.
7. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is run, the multi-objective game overtaking path planning method considering safety optimization described in claim 6 is executed.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor executes the multi-objective game overtaking path planning method considering safety optimization as described in claim 6 through the computer program.
Citation Information
Cited By
Multi-agent boundary game method based on priority path planning
CN119740660A
