Parafoil type underwater unmanned underwater vehicle cross-domain trajectory planning method and system

Abstract

The invention belongs to the field of path planning, and particularly relates to a parafoil type unmanned underwater vehicle cross-domain trajectory planning method and system. According to the method, the optimal water entry state problem is solved by combining the improved A * algorithm with the MOEA / D algorithm, then the optimal air-drop state is planned according to the optimal water entry state, the air and underwater tracks are generated, track planning can be conducted according to different media, water entry is conducted at a certain speed and angle, the water entry safety is guaranteed, energy consumption is low after water entry is conducted according to the planned path, and the method is suitable for large-scale popularization and application. According to the method, the effective voyage is large, the path is convenient to follow, integrated trajectory planning from an air-drop state to an underwater terminal point is achieved for the first time, the problem of independent design of air and underwater trajectories is solved, the water entry speed and angle are optimized through the MOEA / D algorithm, impact damage is avoided, and meanwhile underwater starting energy consumption is reduced; the effective voyage of the underwater vehicle is remarkably improved by combining the energy consumption model of the improved A * algorithm with wind field / ocean current compensation.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of path planning, and particularly relates to a wing-parachute type underwater unmanned underwater vehicle cross-domain trajectory planning method and system. BACKGROUND

[0002] In recent years, with the expansion of the range and depth of ocean observation, underwater unmanned underwater vehicles play an important role in deep water exploration, underwater cable laying, seabed topographic mapping, shipwreck salvage and other tasks.

[0003] However, it is difficult to cope with remote, harsh environments, inaccessible sea areas or sudden marine events such as marine disasters and marine ecological changes.

[0004] Air-dropped unmanned underwater vehicles can solve the problems of low speed, poor timeliness and limited endurance of traditional unmanned underwater vehicles, and can quickly reach the scene for real-time observation and data collection, providing timely data support for scientific research and emergency decision-making.

[0005] However, current trajectory planning is mostly for single medium. However, wing-parachute type unmanned underwater vehicles need to plan trajectories in air and underwater respectively, and need to ensure the safety of the unmanned underwater vehicle when entering the water. Existing planning methods cannot handle trajectory planning in different media. SUMMARY

[0006] The purpose of the present application is to provide a wing-parachute type underwater unmanned underwater vehicle cross-domain trajectory planning method and system.

[0007] A wing-parachute type underwater unmanned underwater vehicle cross-domain trajectory planning method, the specific steps are as follows:

[0008] Step 1: Obtain environmental information, plan task information, and set an initial water entry point;

[0009] Step 2: Calculate the underwater trajectory cost function using the improved A* algorithm to obtain the underwater trajectory energy consumption;

[0010] Step 3: Based on the underwater trajectory energy consumption, use the MOEA / D algorithm to calculate the water entry state search optimization objective function, and update the water entry point and water entry heading;

[0011] Step 4: Determine whether the water entry point search reaches the set termination condition and output the water entry point; otherwise, return to step 2 until the output condition is met;

[0012] Step 5: Set an initial air-drop point;

[0013] Step 6: Calculate the air trajectory cost function using the improved A* algorithm to obtain the air trajectory energy consumption;

[0014] Step 7: Based on the energy consumption of the aerial trajectory, use the MOEA / D algorithm to calculate the search optimization objective function of the aerial delivery state,

[0015] Update the aerial delivery point and the aerial delivery heading;

[0016] Step 8: Determine whether the aerial delivery point search reaches the set termination condition and output the aerial delivery point; otherwise, return to step 6 until the output condition is met.

[0017] Step 9: Re-plan the aerial and underwater trajectories according to the output aerial delivery point and the water entry point, and the task is completed.

[0018] Further, the improved A* algorithm for calculating the trajectory cost function in steps 2 and 6 is as follows:

[0019] f(n) = k g ×g(n) + k h ×h(n)

[0020] g(n) = g move + g turm

[0021]

[0022] h(n) = k move h move +k turn h turn

[0023]

[0024] where f (n) is the trajectory cost function, g (n) is the actual energy cost function, g move is the motion energy consumption model, g turn is the turning energy consumption model, h (n) is the heuristic function, h move is the motion heuristic function, h turn is the turning heuristic function, k g , k h are the actual cost weight coefficient and the heuristic function weight coefficient, respectively, k move , k turn represent the heuristic motion energy coefficient and the heuristic turning energy coefficient, respectively, L i represents the Euclidean distance from the current node to the parent node, ψ n is the current node heading, k h0 , k hmin , λ, k h (n) are the initial weight, the minimum weight of the heuristic function, the weight decay ratio, the current weight, respectively, (x f , y f, Z f ) is the underwater end point, (x n , y n , Z n ) is the current node coordinate, a is the angle of the sea current in the geodetic coordinate system, (ψ i+1 - ψ i ) is the cumulative angle change value, ψ i is the heading angle of the current point, h (start) is the heuristic function of the initial point.

[0025] Further, the improved A* algorithm in steps 2 and 6 is obtained, and the energy consumption of the underwater trajectory and the energy consumption of the air trajectory are specifically:

[0026] Step 2.1: initialize the open list and the closed list of the improved A* algorithm;

[0027] Step 2.2: determine whether the current point is the end point;

[0028] If yes, backtrack the closed list; if no, execute step 2.3;

[0029] Step 2.3: calculate the actual energy cost function, the heuristic function, and the heuristic function weight;

[0030] Step 2.4: calculate the adjacent point cost function and determine whether the adjacent point meets the constraint;

[0031] If yes, update the open list, the closed list, the weight coefficient, the current point, and return to step 2.2 until the output condition is met;

[0032] If no, remove the adjacent point that does not meet the dynamic constraint, reupdate the open list, the closed list, the weight coefficient, the current point, and return to step 2.2 until the output condition is met;

[0033] Step 2.5: according to the output end point, calculate the energy consumption of the underwater trajectory and the energy consumption of the air trajectory.

[0034] Further, the MOEA / D algorithm is used to calculate the water entry state target function in step 3, and the water entry point and the water entry state are updated.

[0035] The water heading is specifically:

[0036] J enter = {J water , J save}

[0037] J water = f(F water )

[0038] J save = cos(ψ r - β) + 1

[0039] J enter is the optimization objective function of the water entry state search, J water is the energy objective function of the underwater trajectory planning, J save is the water entry safety objective function, F water is the terminal point state, and β is the angle of the wind field in the earth coordinate system, ψ r is the water entry heading angle.

[0040] Further, the step 7 of calculating the air drop state search optimization objective function by using the MOEA / D algorithm, and updating the air drop point and the air drop heading specifically comprises:

[0041] J drop = min f (F air )

[0042] J drop is the air drop state search optimization objective function, F air is the air drop point state.

[0043] Further, the step 9 of re-planning the air and underwater trajectories adopts the B-spline fitting air-water trajectory.

[0044] A wing-parachute type underwater unmanned submarine cross-domain trajectory planning system, which realizes the method of any one of claims 1-6, the system comprises an environment perception module, a water entry state search module, a cost calculation module, an air drop state search module and a trajectory point fitting module.

[0045] The environment perception module is used for perceiving the wind field information in the air drop airspace of the wing-parachute type unmanned submarine and the sea current information of the underwater navigation area.

[0046] The water entry state search module is used for initializing a water entry state search population, decomposing a multi-objective problem, cooperatively solving sub-problems and outputting Pareto optimal solutions.

[0047] The cost calculation module is used for calculating the underwater trajectory cost under the current water entry state, calculating the air trajectory cost function under the current air head state and storing trajectory points.

[0048] The air drop state search module is used for initializing an air drop state, iteratively optimizing the air drop state and outputting optimal solutions.

[0049] The trajectory point fitting module is used for reading the optimal trajectory stored by the cost calculation module and fitting the trajectory points.

[0050] The present application has the following beneficial effects:

[0051] The application solves the optimal water entry state problem by improving the A* algorithm combined with the MOEA / D algorithm, and then plans the optimal air drop state according to the optimal water entry state, generates the air and underwater trajectory, can plan the trajectory for different media, and enters the water at a certain speed and angle, ensures the safety of water entry, and the energy consumption of the planned path after water entry is low, the effective range is large, the path is easy to follow, and the following is realized:

[0052] 1. Cross-domain coordination: the integrated trajectory planning from the air drop state to the underwater terminal point is realized for the first time, and the problem of independent design of air and underwater trajectories is solved;

[0053] 2. Water entry safety: the water entry speed and angle are optimized by the MOEA / D algorithm to avoid impact damage and reduce underwater starting energy consumption;

[0054] 3. Energy efficiency: the energy consumption model combined with the improved A* algorithm and the wind field / current compensation significantly improves the effective range of the underwater vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0055] Figure 1 The application is a whole flowchart of water entry state and air drop state.

[0056] Figure 2 The application is a flowchart of trajectory point generation.

[0057] Figure 3 The application is a water entry state search flowchart.

[0058] Figure 4 The application is an air-water integrated trajectory diagram. DETAILED DESCRIPTION

[0059] To achieve the above functions, the application provides a wing-parachute underwater unmanned underwater vehicle cross-domain trajectory planning method, which specifically includes the following steps:

[0060] Step 1: Establishing a wing-parachute-vehicle four-degree-of-freedom kinematic model and a vehicle five-degree-of-freedom kinematic model in the geodetic coordinate system.

[0061] The geodetic coordinate system takes the sea surface directly above the system target point as the origin, the x-axis points to the east, the y-axis points to the north, and the z-axis is determined by the right-hand rule;

[0062] Establishing an air four-degree-of-freedom kinematic model of the wing-parachute-vehicle system in the geodetic coordinate system:

[0063]

[0064] In the formula: is the derivative of the position component of the vehicle in the geodetic coordinate system; V is the speed of the wing-parachute-vehicle system; θ, ψ are the glide angle and heading angle, respectively; is the heading angle rate; u is the parafoil control variable; V w , β are the wind speed and its angle with the earth coordinate system, respectively.

[0065] The five-degree-of-freedom underactuated kinematic model of the vehicle in the earth coordinate system is established:

[0066]

[0067] In the formula: is the derivative of the position component of the vehicle in the earth coordinate system; u, v, w, θ, ψ are the velocity and attitude angle of the vehicle in the moving coordinate system; are the pitch angle and the heading angle rate of change, respectively; V h , α are the current speed and its angle with the earth coordinate system, respectively.

[0068] Step two: Obtain environmental information and plan task information. The environmental information specifically includes: obtaining the wind speed and the angle V w , β; the current speed and the angle V h , α. The task information includes: the position information x f , y f , z f of the underwater termination point, and the minimum safe parachute drop height h min .

[0069] Step three: Set the water entry state search objective function and the parachute drop state search objective function.

[0070] J enter = {J water , J save}

[0071] J water = f (F water )

[0072] J save = cos (ψ r - β) + 1

[0073] J drop = min f (F air )

[0074] In the formula: J enter is the water entry state search optimization objective function, J water is the energy objective function of underwater trajectory planning, J save is the water entry safety objective function, F water is the termination point state, J drop is the parachute drop state search optimization objective function, u is the parafoil control variable, and F airFor the state of the drop point, β is the angle of the wind field in the geodetic coordinate system, ψ r is the water entry heading angle.

[0075] Step four: set the cost function and improved heuristic function for trajectory planning improved A* algorithm.

[0076] f(n) = k g × g(n) + k h × h(n)

[0077] g(n) = g move + g turm

[0078]

[0079] h(n) = k move h move + k turn h turn

[0080]

[0081] wherein: f (n) is the trajectory cost function, g (n) is the actual energy cost function, g move is the motion energy consumption model, g turn is the turning energy consumption model, h (n) is the heuristic function, h move is the motion heuristic function, h turn is the turning heuristic function, k g , k h are the actual cost weight coefficient and the heuristic function weight coefficient respectively, k move , k turn respectively represent the heuristic motion energy coefficient and the heuristic turning energy coefficient, L i represents the Euclidean distance from the current node to the parent node, ψ n is the current node heading, k h0 , k hmin , λ, k h (n) are the initial weight, the minimum weight of the heuristic function, the weight decay ratio, the current weight respectively, (x f , y f , Z f ) is the underwater termination point, (x n , y n , Z n ) is the current node coordinate, α is the angle of the sea current in the geodetic coordinate system, (ψ i+1 - ψ i ) is the cumulative angle change value, ψ i is the water entry heading angle, h(start) The heuristic function for the initial point.

[0082] The constraints are defined, specifically including: parachute kinematic constraints and unmanned underwater vehicle kinematic constraints.

[0083] The aforementioned parachute kinematic constraints: Since the air taxiing segment is an unpowered parachute, its simplified constraint condition is only the double-sided pull-down control quantity u constraint.

[0084] Kinematic constraints of unmanned underwater vehicles: Since the underwater vehicle is an underactuated unmanned underwater vehicle, its simplified constraints include: velocity constraints, acceleration constraints, attitude angle constraints, and attitude angular velocity constraints.

[0085] Step 5: Subproblem decomposition of MOEA / D in the water entry state search, weight vector initialization, and initial population generation.

[0086] Step 6: Initialize the open and closed lists for the improved A* algorithm used to calculate the underwater trajectory cost function.

[0087] Set the current water entry state as Start:(x r ,y r ,0,ψ r As the starting state for underwater trajectory search, the improved A* algorithm is used to begin planning the energy-optimal underwater trajectory under the current entry state. After the optimal trajectory search is completed, the energy consumption f(F) of the optimal trajectory is passed to step seven of the optimal entry state search.

[0088] Step 7: Determine if the water entry state search has reached the set termination condition.

[0089] If not: Continue to use the MOEA / D loop to iteratively search for the optimal entry state. The objective function for each iteration needs to be obtained through step six.

[0090] If: Output the best individual in the population as the optimal water entry state Start:(x r ,y r ,0,ψ r ).

[0091] Step 8: Utilize the optimal water entry state (x) planned in Step 7 r ,y r ,0,ψ r This serves as the termination state for air trajectory planning. An initial population is generated to search for airdrop states that satisfy the feasible airdrop domain, along with open and closed lists for the A* algorithm used to initialize the air trajectory.

[0092] Step Nine: Set the current airdrop status to Start:(x s ,y s ,hmin ,ψ s ), as the starting state of the air trajectory search, the improved A* algorithm is used to start planning the optimal air trajectory under the current air-drop state. After the optimal trajectory search is completed, the length f air (F) is passed to step ten.

[0093] Step ten: iteratively calculate the air-drop state with the minimum air value using the MOEA / D algorithm

[0094] Start: (x s ,y s ,h min ,ψ s ), the fitness function of each individual is obtained from step nine. When the termination condition of the search is reached, the optimal air-drop state and its corresponding air trajectory points are output.

[0095] Step eleven: according to the optimal entry state and the optimal air-drop state calculated in steps seven and ten, the corresponding generated trajectory is obtained, and the B-spline is used to fit the air-water trajectory.

[0096] The present application will be further described below in conjunction with the accompanying drawings. Figures 1 to 4

[0097] To achieve the above functions, the present application provides a specific embodiment of a wing-parachute underwater unmanned submarine cross-domain trajectory planning method and system, which uses a one-ton unmanned submarine as the entity for trajectory planning.

[0098] Specifically includes:

[0099] S1: establish a wing-parachute-vehicle four-degree-of-freedom kinematic model and a vehicle five-degree-of-freedom kinematic model under the earth coordinate system.

[0100] The earth coordinate system takes the sea surface directly above the system target point as the origin, the x-axis points to the east direction, the y-axis points to the north direction, and the z-axis is determined by the right-hand rule;

[0101] An air four-degree-of-freedom kinematic model of the wing-parachute-vehicle system under the earth coordinate system is established:

[0102]

[0103] In the formula: is the derivative of the position component of the vehicle under the earth coordinate system; V is the speed of the wing-parachute-vehicle system; θ, ψ are the glide angle and the heading angle, respectively; is the heading rate; U is the wing-parachute control quantity; V w , β are the wind speed and its angle with the earth coordinate system, respectively.

[0104] ​A five-degree-of-freedom underactuated kinematic model of the vehicle in the geodetic coordinate system is established:

[0105]

[0106] In the formula: is the derivative of the position component of the vehicle in the geodetic coordinate system; u, v, w, θ, ψ are the velocity and attitude angle of the vehicle in the moving coordinate system; are the change rates of the pitch angle and the heading angle, respectively; V h , α are the sea current velocity and the angle in the geodetic coordinate system, respectively.

[0107] S2: Obtain environmental information and task information. The environmental information specifically includes: obtaining the wind field velocity and the included angle V w , β; the sea current velocity and the included angle V h , α. The task information includes: the underwater termination point position information x f , y f , z f , the minimum safe parachute release height h min . The embodiment generates the wind field size and the included angle as V w = 5 m / s, β = 102°, and generates the sea current size and the included angle as V h = 0.6 m / s, α = -15°

[0108] S3: Set the water entry state search objective function and the parachute release state search objective function.

[0109] J enter = {J water , J save}

[0110] J water = f(F water )

[0111] J save = cos(ψ r - β) + 1

[0112] J drop = min f(F air )

[0113] In the formula: J enter is the water entry state search optimization objective function, J water is the energy objective function of underwater trajectory planning, J save is the water entry safety objective function, F water is the termination point state, J drop is the parachute release state search optimization objective function, and u is the wing parachute control variable, F airFor the state of the drop point, β is the angle of the wind field in the geodetic coordinate system, ψ r is the water entry heading angle.

[0114] S4: Set the cost function and heuristic function for trajectory planning improved A* algorithm.

[0115] f(n)=k g ×g(n)+k h ×h(n)

[0116] g(n)=g move +g turm

[0117]

[0118] h(n)=k move h move +k turn h turn

[0119]

[0120] Wherein: f (n) is the trajectory cost function, g (n) is the actual energy cost function, g move is the motion energy consumption model, g turn is the turning energy consumption model, h (n) is the heuristic function, h move is the motion heuristic function, h turn is the turning heuristic function, k g ,k h are the actual cost weight coefficient and the heuristic function weight coefficient respectively, k move ,k turn respectively represent the heuristic motion energy coefficient and the heuristic turning energy coefficient, L i represents the Euclidean distance from the current node to the parent node, ψ n is the current node heading, k h0 ,k hmin ,λ,k h (n) are the initial weight, the minimum weight of the heuristic function, the weight decay ratio, the current weight, (x f , y f , Z f ) is the underwater termination point, (x n , y n , Z n ) is the current node coordinate, α is the angle of the sea current in the geodetic coordinate system, (ψ i+1 -ψ i ) is the cumulative angle change value, ψ i is the water entry heading angle of the current point, h(start) The heuristic function is the initial point.

[0121] The actual cost weight k in the embodiment g = 4; the minimum weight k of the heuristic function hmin = 4; the weight decay ratio λ = 1.8; the motion heuristic function weight k hmove = 0.8; the turning heuristic function weight k hturn = 0.2.

[0122] The constraint conditions include:

[0123] The constraint conditions are determined, which specifically include: wing parachute kinematics constraints and unmanned underwater vehicle kinematics constraints.

[0124] The wing parachute kinematics constraints: since the air sliding section is a powerless wing parachute, the simplified constraint condition thereof is only the bilateral pull-down control amount u constraint.

[0125] The unmanned underwater vehicle kinematics constraints: since the underwater section of the vehicle is an underactuated unmanned underwater vehicle, the simplified constraint condition thereof includes: the speed constraint u min ≤ u ≤ u max , the acceleration constraint The attitude angular velocity constraint -q max ≤ q ≤ q max , the attitude angle constraint -r max ≤ r ≤ r max , and the attitude angle constraint -θ max ≤ θ ≤ θ max .

[0126] In the embodiment, the upper and lower limits of the speed are set as u max = 5.14 m / s, and u min = -0.5 m / s; the upper and lower limits of the acceleration are set as The attitude angular velocity constraint condition q max = 5 deg / s, r max = 5 deg / s; the attitude angle constraint condition is θ max = 45°; and the bilateral pull-down amount constraint condition of the wing parachute in the air section is

[0127] S5: Sub-problem decomposition, weight vector initialization and initial population generation of the water entry state search MOEA / D.

[0128] The sub-problem decomposition: in the embodiment, the total problem decomposition adopts the weighted decomposition method to decompose the sub-problems, each weight vector corresponds to a single-objective sub-problem, and the objective of each sub-problem is to minimize the weighted sum: ming(s| λ i ) = λ i1 × J water + λ i2 × J save .

[0129] where λ i1 and λ i2 are two components of the weight vector λ i , and the weight vector satisfies λ i1 + λ i2 = 1.

[0130] The optimal entry state search problem is decomposed into 9 sub-problems, the weight vector is initialized as λ1 = [0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9], λ2 = [0.9, 0.8, 0.7, 0.6, 0.5, 0.4, 0.3, 0.2, 0.1], and a solution is randomly generated for each sub-problem.

[0131] The neighborhood of each sub-problem is defined by the Euclidean distance, and the neighborhood of each sub-problem contains the three sub-problems closest to it.

[0132] S6: Initialize the open list and the closed list of the improved A* algorithm used for the calculation of the underwater trajectory cost function. The current entry state Start: (x r , y r , 0, ψ r ) is taken as the starting state of the underwater trajectory search, and the improved A* algorithm is used to start the calculation of the cost function of the energy-optimal underwater trajectory under the current entry state.

[0133] S6.1: Select the point with the minimum f(n) from the open list as the current point. Initially, the open list only contains Start: (x r , y r , 0, ψ r ), and the f(n) of Finish: (x f , y f , z f , ψ f ) is calculated as g(S) = 0. Move S from the open list to the closed list and record its parent node.

[0134] S6.2: Generate the 9 adjacent points of the current point that are reachable by the AUV with ψ = (-5°, 0°, 5°), θ = (-5°, 0°, 5°), and exclude the points that are kinematically unreachable and already exist in the closed list.

[0135] S6.3: Calculate the cost of the adjacent points and update the list. For each adjacent point: calculate g(n) and f(n). If the adjacent point is not in the open list: directly add it and record its parent node. If A is in the open list: compare the newly calculated g(n) with the original g(n), and if it is smaller, update g(n), f(n), and its parent node.

[0136] S6.4: Determine if the current point is the end point of the trajectory Finish: (x f ,y f ,z f ,ψ f )

[0137] No: Repeat S6.2-S6.3, expand the search range step by step.

[0138] Yes: When the end point Finish: (x f ,y f ,z f ,ψ f ) is added to the closed list, the search is completed, and the optimal path is obtained by backtracking from Finish: (x f ,y f ,z f ,ψ f ) to Start: (x r ,y r ,0,ψ r ).

[0139] S6.5: Backtrack the parent nodes one by one to get the total cost f(Finish) that is the smallest. f(Finish) is passed to S7 as one of the objective functions of the optimal entry state search.

[0140] S7: Iteratively search the sub-problems obtained by MOEA / D decomposition in S5.

[0141] S7.1 Parent selection: For each sub-problem, randomly select 2 solutions from its neighborhood as parents.

[0142] S7.2 Crossover and mutation: Simulated binary crossover is used for crossover operation, and Gaussian perturbation with mean 0 and standard deviation 1 is used for mutation operation, represented as X new =X old +σ·N(0,1). The perturbation parameter intensity σ=1.5.

[0143] Update neighborhood solutions: Compare the child solution with the current solution of the sub-problem in the neighborhood. If the child solution has a smaller objective function, replace the current solution.

[0144] S7.3 Iterative optimization: Repeat S7.1-S7.2 to gradually approach the Pareto optimal solution. After multiple iterations, the solution of each sub-problem will be gradually optimized, and finally form a solution set covering different Pareto optimal directions.

[0145] S7.4 Optimal sub-problem screening, output the optimal entry state (x r ,y r, 0, ψ r ), output the optimal underwater trajectory point in the current water entry state.

[0146] S8: Use the optimal water entry state (x r ,y r , 0, ψ r ) planned by S7 as the termination state of the air trajectory planning. Generate the initialization population of the air drop state search that satisfies the feasible air drop domain, and initialize the open list and closed list of the A* algorithm for the air trajectory.

[0147] The feasible air drop domain is specifically expressed as: where λ air is the glide ratio of the parafoil. In this embodiment, λ air = 3, h min = 1000.

[0148] S9: Use the current air drop state Start: (x s ,y s ,h min , ψ s ) as the starting state of the air trajectory search, and start planning the optimal air trajectory in the current air drop state using the improved A* algorithm. After the optimal trajectory search is completed, the cost function f air (F) of the optimal trajectory is passed to S10 of the optimal air drop state search.

[0149] S10: Use the MOEA / D algorithm to iteratively calculate the air drop state Start: (x s ,y s ,h min , ψ s ) with the minimum air cost, and determine whether the termination condition

[0150] If yes: output the optimal air drop state and the corresponding optimal air trajectory point.

[0151] If no: determine whether there is an individual outside the feasible air drop domain.

[0152] If yes: set the coordinates of the individual to the intersection point of the line connecting the point and the center of the feasible air drop domain, and continue iteration until the termination condition is met.

[0153] If no: continue to call S9 until the termination condition is met.

[0154] S11: According to the generated trajectory, use B-spline to fit the air-water trajectory, and the result is shown in Figure Four .

[0155] The above merely provides the preferred embodiments of the present application, and is not used to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the principles and technical scope of the present application shall fall into the scope of the present application.

Claims

1. A cross-domain trajectory planning method for a paraglider-type underwater unmanned vehicle, characterized in that: The specific steps are as follows: Step 1: Obtain environmental information, plan task information, and set the initial water entry point; Step 2: Calculate the underwater trajectory cost function using the improved A* algorithm to obtain the underwater trajectory energy consumption; Step 3: Based on the underwater trajectory energy consumption, use the MOEA / D algorithm to calculate the objective function for water entry state search optimization, and update the water entry point and water entry heading; Step 4: Determine if the water inlet point search has reached the set termination condition and output the water inlet point; otherwise, return to Step 2 until the output condition is met. Step 5: Set the initial airdrop point; Step 6: Use the improved A* algorithm to calculate the cost function of the air trajectory and obtain the energy consumption of the air trajectory; Step 7: Based on the aforementioned air trajectory energy consumption, calculate the airdrop state search optimization objective function using the MOEA / D algorithm. Update airdrop location and airdrop course; Step 8: Determine if the airdrop point search has reached the set termination condition and output the airdrop point; otherwise, return to step 6 until the output condition is met. Step 9: Based on the output drop point and water entry point, replan the aerial and underwater trajectories. Mission complete.

2. The cross-domain trajectory planning method for a paraglider-type underwater unmanned vehicle according to claim 1, characterized in that: The improved A* algorithm for calculating the trajectory cost function in steps 2 and 6 is specifically as follows: f(n)=k g ×g(n)+k h ×h(n) g(n)=g move +g turm h(n)=k move h move +k turn h turn Where: f (n) Let g be the trajectory cost function. (n) Let g be the actual energy cost function. move For the motion energy consumption model, g turn For the turning energy consumption model, h (n) h is a heuristic function move h is the motion heuristic function. turn For the turning heuristic function, k g ,k h These are the actual cost weighting coefficient and the heuristic function weighting coefficient, respectively, k move ,k turn L represents the inspired kinetic energy coefficient and the inspired turning energy coefficient, respectively. i ψ represents the Euclidean distance from the current node to its parent node. n For the current node's heading, k h0 ,k hmin ,λ,k h (n) represent the initial weight, the minimum weight of the heuristic function, the weight decay ratio, and the current weight, respectively. f y f Z f (x) is the underwater termination point. n y n Z n ) represents the coordinates of the current node, α represents the angle between the ocean currents in the geodetic coordinate system, and (ψ) represents the angle between the currents and the currents in the geodetic coordinate system. i+1 -ψ i ) represents the cumulative angle change value, ψ i h is the heading angle of the current point. (start) The heuristic function for the initial point.

3. The cross-domain trajectory planning method for a paraglider-type underwater unmanned vehicle according to claim 2, characterized in that: The improved A* algorithm in steps 2 and 6, specifically for obtaining the energy consumption of the underwater trajectory and the aerial trajectory, is as follows: Step 2.1: Initialize the open and closed lists for the improved A* algorithm; Step 2.2: Determine if the current point is the destination; If yes, backtrack the closed list; otherwise, proceed to step 2.

3. Step 2.3: Calculate the actual energy cost function, heuristic function, and heuristic function weights; Step 2.4: Calculate the cost function of adjacent nodes and determine whether adjacent nodes satisfy the constraints; If so, update the open list, closed list, weight coefficients, and current point, and return to step 2.2 until the output condition is met; Otherwise, remove adjacent nodes that do not meet the dynamic constraints, and update the open list, closed list, and weight coefficients. From the current point, return to step 2.2 until the output condition is met; Step 2.5: Calculate the energy consumption of the underwater trajectory and the energy consumption of the aerial trajectory based on the output endpoint.

4. The cross-domain trajectory planning method for a paraglider-type underwater unmanned vehicle according to claim 1, characterized in that: In step 3, the MOEA / D algorithm is used to calculate the objective function for the water entry state, and the water entry point and water entry heading are updated as follows: J enter ={J water ,J save } J water =f(F water ) J save =cos(ψ r -b)+1 Among them: J enter To optimize the objective function for the water entry state search, J water The energy objective function for underwater trajectory planning, J save Let F be the objective function for water entry safety. water The endpoint state is given, where β is the angle between the wind fields in the geodetic coordinate system, and ψ is the angle between the wind fields and the ground coordinate system. r This refers to the heading angle upon entering the water.

5. The cross-domain trajectory planning method for a paraglider-type underwater unmanned vehicle according to claim 1, characterized in that: In step 7, the MOEA / D algorithm is used to calculate the objective function for optimizing the airdrop state search, and the airdrop point and airdrop heading are updated as follows: J drop =minf(F air ) Among them: J drop To optimize the objective function for the airdrop state search, F air This indicates the airdrop point is in status.

6. The cross-domain trajectory planning method for a paraglider-type underwater unmanned vehicle according to claim 1, characterized in that: In step 9, the aerial and underwater trajectories are replanned, and B-spline fitting is used to fit the aerial and underwater trajectories.

7. A cross-domain trajectory planning system for a paraglider-type underwater unmanned vehicle, implementing the method described in any one of claims 1-6, characterized in that: The system includes an environmental perception module, a water entry status search module, a cost calculation module, an airdrop status search module, and a trajectory point fitting module. The environmental perception module is used to perceive wind field information in the airspace where the paraglider unmanned underwater vehicle is airdropped and ocean current information in the underwater navigation area. The water entry state search module is used to initialize the water entry state search population, decompose the multi-objective problem, collaboratively solve the sub-problems, and output the Pareto optimal solution. The cost calculation module is used to calculate the underwater trajectory cost under the current water entry state, calculate the aerial trajectory cost function under the current airborne state, and store the trajectory points. The airdrop status search module is used to initialize the airdrop status, iteratively optimize the airdrop status, and output the optimal solution. The trajectory point fitting module is used to read the optimal trajectory stored in the cost calculation module and fit the trajectory points.

Images