UUV cluster recovery task planning method

Abstract

The invention discloses a UUV cluster recovery task planning method, and provides a UUV cluster recovery task planning method innovatively for a UUV cluster recovery task scene, so that the shortest recovery process of recovering a UUV cluster by a carrying platform can be realized. The generated recovery scheme can be directly issued to the UUV by the carrying platform to drive the carrying platform and the UUV to complete the task, so that the task recovery time is shortened; aiming at the defects that the particle swarm algorithm is easy to fall into local optimum and premature phenomenon occurs, the invention provides an improved particle swarm algorithm, so that the probability that the particle swarm algorithm falls into local optimum is effectively reduced, and the rapid solution of the UUV cluster recovery task is realized.

Description

Technical Field

[0001] The present invention belongs to the field of UUV cluster mission planning, and more specifically, relates to a UUV cluster recovery mission planning method, which is suitable for planning of a UUV cluster recovery plan by a carrier platform. Background Art

[0002] In recent years, underwater unmanned vehicles (UUVs) have become the research focus of various countries in the unmanned field due to their advantages such as low cost, high flexibility and strong concealment. Compared with a single UUV, the collaborative operation of multiple UUVs is wider and more efficient. They can complete tasks that a single UUV cannot complete in complex environments. They are widely used in the fields of seabed topography exploration, marine resource surveys, underwater target search, etc. However, due to the current limitations of UUV battery capacity, UUVs are difficult to meet some large-scale and long-duration mission requirements. Using platforms such as surface unmanned boats as UUV carriers and support platforms, and using dynamic base recovery technology to recover, charge, download data, and then deploy again, can effectively expand the operating range and operation time of UUVs.

[0003] Therefore, it is crucial to plan the recovery plan of the UUV cluster reasonably for the continuous operation of the UUV cluster. In the recovery scenario, it is necessary to ensure the safety of the UUV, meet the constraints such as the UUV recovery time limit and energy consumption limit, and plan the plan for the carrier platform to complete the recovery of the UUV cluster as quickly as possible. Therefore, we designed a UUV cluster recovery task planning method based on the improved particle swarm algorithm, optimized the recovery plan, planned the recovery position of each UUV, and realized that the carrier platform starts from the initial position and recovers the UUVs in sequence in the shortest time. Summary of the invention

[0004] The present invention provides a UUV cluster recovery task planning method, which takes UUV cluster recovery position planning as the research object, takes the UUV recovery time window as a constraint condition, constructs a UUV cluster recovery task planning model, and adopts an improved particle swarm algorithm to find a solution that meets the constraint conditions and has the shortest recovery time.

[0005] To achieve the above purpose, the present invention adopts the following technical scheme.

[0006] A UUV cluster recovery mission planning method, comprising:

[0007] Step (1) The carrier platform sends a mission completion instruction and its own location information to the UUV cluster, and each UUV sends its location information to the carrier platform;

[0008] Step (2) the carrier platform plans the recovery position of each UUV and sends it to each UUV;

[0009] Step (3) After receiving the location information planned by the carrier platform, all UUVs sail to the predetermined recovery location, while the carrier platform sails to the first recovery point;

[0010] Step (4) the carrier platform and a UUV meet at a predetermined recovery point, the carrier platform guides the UUV for recovery and completes the UUV recovery task, and the carrier platform navigates along a predetermined trajectory to the next UUV recovery position;

[0011] Step (5) The transport platform arrives at the next recovery position and completes the UUV rendezvous and recovery task at that point; and so on, until all UUVs are recovered.

[0012] Preferably, the method for planning the recovery position of each UUV by the carrier platform in step (2) includes:

[0013] Step 1: Initialize population positions:

[0014] The recovery plan of the UUV cluster is used as particles, and the initial position of the particle population is initialized using Tent chaotic mapping:

[0015]

[0016] in, represents the position of the i-th particle in the t+1th iteration population, Represents the position of the i-th particle in the t-th iteration population, and the value range of α is (0,1);

[0017] Step 2: Update particle velocity using adaptive weight coefficients;

[0018]

[0019] Where w represents the inertia weight coefficient; represents the velocity component of the dth dimension of the ith particle in the t+1th iteration, that is, the position increment of the dth UUV. c1 and c2 represent acceleration factors; r1 and r2 are random numbers in the range of (0,1). is the position component of the dth dimension during the tth iteration of the ith particle, which means the recovery position of the dth UUV. Represents the local optimal solution in the tth iteration process The local optimal solution of the dth dimension in , that is, the optimal solution of the dth UUV in this iteration, x gbest,d represents the global optimal solution x during the first t iterations gbest The global optimal solution of the d-th dimension in , that is, the optimal solution of the d-th UUV in all current iterations; N represents the number of UUVs;

[0020] Step 3: Update particle positions:

[0021]

[0022] Step 4: Calculate the fitness value of each particle:

[0023]

[0024] in, It represents the time taken by the carrier platform to reach the recovery position of the nth UUV from the previous position; represents the recovery constraint term, τ represents the recovery time of the carrier platform; (t n -BE n ) represents the power constraint term, BE n represents the remaining battery life of UUV n; t n represents the time from the initial position of UUV n to its recovery position; σ1 and σ2 are the penalty coefficients of the set recovery constraint and power constraint respectively;

[0025] Step 5: Update the local optimal and global optimal solutions; find the particle with the minimum fitness function in the particle swarm obtained in this iteration as the local optimal solution

[0026] For the global optimal solution, if the fitness value of the current global optimal solution is greater than the local optimal solution The fitness value of As the global optimal solution x gbest ; On the contrary, the global optimal solution keeps the current global optimal solution unchanged;

[0027] Step 6: Use the Gauss-Cauchy mutation strategy to update the global optimal solution of the particle:

[0028] According to the current global optimal solution x gbest Calculate the Gaussian mutation operator as follows:

[0029]

[0030] According to the current global optimal solution x gbest Calculate the Cauchy mutation operator as follows:

[0031]

[0032] Then the current global optimal solution is updated as:

[0033]

[0034] Among them, Gaussian (μ1, σ1) is Gaussian distribution, μ1 is the expectation of Gaussian distribution, σ1 is the variance of Gaussian distribution; Cauchy (μ2, σ2) is Cauchy distribution, μ2 is the location parameter of Cauchy distribution, σ2 is the scale parameter of Cauchy distribution;

[0035] Step 7: Add 1 to the number of iterations t to determine whether the number of iterations t exceeds the maximum number of iterations T. If t≤T, jump to step 2 to continue execution; if t>T, end the solution and output the current global optimal solution, which is the UUV cluster recovery mission planning solution.

[0036] Preferably, in step 2, an adaptive weight coefficient is used to update the particle velocity:

[0037]

[0038] w h and w l are the maximum and minimum values ​​of the weight coefficient respectively.

[0039] Preferably, take w h =0.9,w l =0.4.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] (1) The present invention innovatively proposes a UUV cluster recovery mission planning method for the UUV cluster recovery mission scenario, which can achieve the shortest recovery process for the carrier platform to recover the UUV cluster. The generated recovery plan can be directly sent from the carrier platform to the UUV, driving the carrier platform and UUV to complete the mission and shorten the recovery mission time.

[0042] (2) In view of the shortcomings of the particle swarm algorithm that it is easy to fall into local optimality and premature phenomenon, the present invention proposes an improved particle swarm algorithm, which effectively reduces the probability of the particle swarm algorithm falling into local optimality and realizes the rapid solution of the UUV cluster recovery task. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is the principle diagram of the method of the present invention;

[0044] Figure 2 is a flow chart of the method of the present invention;

[0045] Figure 3 This is an example test scenario of the method described in the present invention. DETAILED DESCRIPTION

[0046] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying 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 belong to the protection scope of the present invention.

[0047] Without loss of generality, the present invention first introduces the recovery process of the UUV by the carrier platform, such as Figure 1 As shown in the figure, the UUV cluster recovery process can be divided into the following stages:

[0048] (1) The carrier platform sends a mission completion command and its own location information to the UUV cluster, and each UUV sends its location information to the carrier platform.

[0049] (2) The carrier platform plans the recovery location of each UUV and sends it to each UUV.

[0050] (3) After receiving the location information planned by the carrier platform, all UUVs sail to the predetermined recovery location, while the carrier platform sails to the first recovery point;

[0051] (4) The carrier platform and a UUV meet at a predetermined recovery point, and the carrier platform guides the recovery of the UUV and completes the UUV recovery task. During this recovery period, the carrier platform cannot recover other UUVs, and the carrier platform continues to navigate along the predetermined trajectory to the next UUV recovery point;

[0052] (5) The transport platform arrives at the next recovery point and completes the UUV rendezvous and recovery mission at that point.

[0053] During the recovery mission, the UUV will be affected by external factors such as ocean currents, as well as its own dynamic characteristics such as acceleration and deceleration. Therefore, if all influencing factors are considered, the constructed model will be extremely complex and show obvious nonlinearity. Therefore, without affecting the solution of the problem, the following assumptions are made:

[0054] (1) All UUVs are of the same type, that is, they have the same physical properties and are considered as points in a two-dimensional plane, without considering the influence of factors such as the marine environment;

[0055] (2) The acceleration and deceleration process of the UUV is replaced by the average speed, and the motion of the UUV is regarded as uniform motion;

[0056] (3) There is no communication interference between UUVs and between UUVs and carrier platforms, that is, individuals can communicate smoothly;

[0057] (4) Due to the influence of the actual environment and the UUV's posture, the time from rendezvous to recovery is not fixed. In this paper, the average value τ is taken for simplification;

[0058] (5) The carrier platform and each UUV move in a two-dimensional plane at the same depth and the UUVs are randomly distributed in the two-dimensional plane.

[0059] Assume that a recovery mission includes N UUVs, which are located at different locations in the mission area. The UUV number index is N = {1, 2, ..., N}, and its initial position is (x i-init ,y i-init ), the speed is v i , the initial position of the carrier platform is (x0, y0), and the speed is v M The variable to be optimized is the expected recovery position of each UUV (x i ,y i ). UUV moves from the initial position (x i-init ,y i-init ) Navigate to the expected recovery position (x i ,y i ) is time t i , the time it takes for the carrier platform to reach the expected recovery position from the previous position is

[0060] According to the task scenario and task requirements, the present invention establishes the following objective function:

[0061]

[0062] in, It represents the time taken by the carrier platform to reach the recovery position of the i-th UUV from the previous position:

[0063]

[0064] The following constraints exist:

[0065] (1) Time constraint: The UUV must arrive at the expected recovery location before the carrier platform, that is:

[0066]

[0067] in, represents the time taken by the carrier platform to travel from the i-1 recovery position to the i recovery position, t m It represents the time for UUV m to travel from its initial position to the recovery point.

[0068]

[0069] (2) Power constraint: The UUV must reach the recovery position before the power is exhausted.

[0070]

[0071] Where E i Indicates the remaining power of UUV No. i, B is the proportional coefficient, BE i Indicates the remaining navigation time of UUV No. i.

[0072] (3) Recovery constraint: The carrier platform has a recovery occupancy time τ. From the time the carrier platform starts to recover a certain UUV, it cannot recover other UUVs within τ. Therefore, in order to ensure the smooth recovery of UUVs, the carrier platform must complete the recovery of UUVs before reaching the next recovery location, that is:

[0073]

[0074] The present invention designs a UUV cluster recovery task planning method based on an improved particle swarm algorithm, which can realize the rapid recovery of the UUV cluster.

[0075] The present invention adopts the following technical solutions:

[0076] A UUV cluster recovery mission planning method based on improved particle swarm optimization (PSO) algorithm includes the following improvement measures:

[0077] (1) Tent Chaos Mapping

[0078] Similar to most swarm intelligence algorithms, the position initialization of the particle population adopts a random generation strategy, which may lead to low diversity of the population and slow convergence. In order to enhance the global convergence ability of the algorithm, the Tent chaos map is added to initialize the particle population so that the initial individuals are distributed as much as possible in the solution space of the problem. The expression of the Tent chaos map is as follows:

[0079]

[0080] in, represents the position of the i-th particle in the t+1th iteration population, Represents the position of the i-th particle in the t-th iteration population; the value range of α is (0,1).

[0081] (2) Adaptive dynamic inertia weight coefficient

[0082] In the PSO algorithm, the inertia weight coefficient w is crucial to the effect of the algorithm. However, in the original PSO algorithm, w takes a fixed value, which leads to poor adaptability. The value of w is adaptively adjusted so that the improved algorithm focuses on global search in the early stage and local optimization in the later stage. The update formula of w is as follows:

[0083]

[0084] Among them, w h and w l are the maximum and minimum values ​​of the weight coefficients set, generally w h =0.9,w l =0.4, t represents the current number of iterations, and T represents the total number of iterations of the algorithm.

[0085] (3) Gauss-Cauchy mutation strategy

[0086] In order to reduce the probability of the PSO algorithm falling into the local optimum, the Gauss-Cauchy mutation strategy is introduced to perturb the global optimal solution of the particles in each round of iteration.

[0087] The Gaussian mutation operator is as follows:

[0088]

[0089] The Cauchy mutation operator is as follows:

[0090]

[0091] The Gauss-Cauchy mixed mutation operator is:

[0092]

[0093] in, is the global optimal solution of the particle, Gaussian(μ1,σ1) is the Gaussian distribution, μ1 is the expectation of the Gaussian distribution, σ1 is the variance of the Gaussian distribution; Cauchy(μ2,σ2) is the Cauchy distribution, μ2 is the location parameter of the Cauchy distribution, σ2 is the scale parameter of the Cauchy distribution, t is the current number of iterations, and T is the maximum number of iterations.

[0094] In summary, if Figure 2 As shown, the steps of the improved particle swarm algorithm of the present invention are as follows:

[0095] Step 1: Initialize the population position. The recovery plan of the UUV cluster is taken as a particle, the dimension of the particle is a 2*N matrix, and the initial position of the particle population is initialized using Tent chaotic mapping, so that the initial individuals are distributed as much as possible in the solution space of the problem;

[0096]

[0097] Step 2: Update particle velocity using adaptive weight coefficients;

[0098]

[0099]

[0100] in, represents the velocity component of the d-th dimension of the ith particle in the t+1th iteration, that is, The position increment of UUV, c1 and c2 represent acceleration factors, usually c1=c2=2, r1 and r2 are random numbers in the range of (0,1), is the position component of the dth dimension during the tth iteration of the ith particle, which means the recovery position of the dth UUV. Represents the local optimal solution in the tth iteration process The local optimal solution of the dth dimension in , that is, the optimal solution of the dth UUV in this iteration, x gbest,d represents the global optimal solution x during the first t iterations gbest The global optimal solution of the d-th dimension in is the optimal solution of the d-th UUV in all current iterations.

[0101] Step 3: Update particle positions:

[0102]

[0103] Step 4: Calculate the fitness value of each particle;

[0104]

[0105] Among them, σ1 and σ2 are the penalty coefficients of the penalty term.

[0106] Step 5: Update the local optimal and global optimal solutions; find the particle with the minimum fitness function in the particle swarm obtained in this iteration as the local optimal solution

[0107] For the global optimal solution, if the fitness value of the current global optimal solution is greater than the local optimal solution The fitness value of As the global optimal solution x gbest ; Otherwise, the global optimal solution remains the current global optimal solution unchanged.

[0108] Step 6: Use the Gauss-Cauchy mutation strategy to update the global optimal solution of the particle;

[0109] According to the current global optimal solution x gbest Calculate the Gaussian mutation operator as follows:

[0110]

[0111] According to the current global optimal solution x gbest Calculate the Cauchy mutation operator as follows:

[0112]

[0113] Then the current global optimal solution is updated as:

[0114]

[0115] Step 7: Add 1 to the number of iterations t to determine whether the number of iterations t exceeds the maximum number of iterations T. If t≤T, jump to step 2 to continue execution; if t>T, end the solution and output the current global optimal solution, which is the UUV cluster recovery mission planning solution.

[0116] Example:

[0117] The present invention provides a UUV cluster recovery mission planning method. The specific implementation of the method of the present invention is described in detail below in conjunction with the accompanying drawings.

[0118] Step 1: Input the position and speed of the carrier platform at the start of the mission; the position, speed, remaining power and other parameters of each UUV;

[0119] Step 2: Set the parameters of the improved particle swarm algorithm, such as the number of particles, learning factors c1, c2, and weight coefficient w h 、w l , Gaussian variation factors μ1, σ1, Cauchy variation factors μ2, σ2 and other parameters;

[0120] Step 3: Improve the particle swarm algorithm to calculate the UUV cluster recovery plan;

[0121] Step 4: The carrier platform sends the recovery location of each UUV to the corresponding UUV and starts executing the recovery plan.

[0122] A test case of the present invention is shown in the following table. The final recovery route of the carrier platform is as follows: Figure 3 shown.

[0123]

Claims

1. A UUV cluster recovery mission planning method, characterized in that: include: Step (1) The carrier platform sends a mission completion instruction and its own location information to the UUV cluster, and each UUV sends its location information to the carrier platform; Step (2) the carrier platform plans the recovery position of each UUV and sends it to each UUV; Step (3) After receiving the location information planned by the carrier platform, all UUVs sail to the predetermined recovery location, while the carrier platform sails to the first recovery point; Step (4) the carrier platform and a UUV meet at a predetermined recovery point, the carrier platform guides the UUV for recovery and completes the UUV recovery task, and the carrier platform navigates along a predetermined trajectory to the next UUV recovery position; Step (5) The transport platform arrives at the next recovery position and completes the UUV rendezvous and recovery task at that point; and so on, until all UUVs are recovered.

2. A UUV cluster recovery mission planning method as claimed in claim 1, characterized in that: in, The method for planning the recovery position of each UUV by the carrier platform in step (2) includes: Step 1: Initialize population positions: The recovery plan of the UUV cluster is used as particles, and the initial position of the particle population is initialized using Tent chaotic mapping: in, represents the position of the i-th particle in the t+1th iteration population, Represents the position of the i-th particle in the t-th iteration population, and the value range of α is (0,1); Step 2: Update particle velocity using adaptive weight coefficients; Where w represents the inertia weight coefficient; represents the velocity component of the dth dimension of the ith particle in the t+1th iteration, that is, the position increment of the dth UUV. c1 and c2 represent acceleration factors; r1 and r2 are random numbers in the range of (0,1). is the position component of the dth dimension during the tth iteration of the ith particle, which means the recovery position of the dth UUV. Represents the local optimal solution in the tth iteration process The local optimal solution of the dth dimension in , that is, the optimal solution of the dth UUV in this iteration, x gbest,d represents the global optimal solution x during the first t iterations gbest The global optimal solution of the d-th dimension in , that is, the optimal solution of the d-th UUV in all current iterations; N represents the number of UUVs; Step 3: Update particle positions: Step 4: Calculate the fitness value of each particle: in, It represents the time taken by the carrier platform to reach the recovery position of the nth UUV from the previous position; represents the recovery constraint term, τ represents the recovery time of the carrier platform; (t n -BE n ) represents the power constraint term, BE n represents the remaining battery life of UUV n; t n represents the time from the initial position of UUV n to its recovery position; σ1 and σ2 are the penalty coefficients of the set recovery constraint and power constraint respectively; Step 5: Update the local optimal and global optimal solutions; find the particle with the minimum fitness function in the particle swarm obtained in this iteration as the local optimal solution For the global optimal solution, if the fitness value of the current global optimal solution is greater than the local optimal solution The fitness value of As the global optimal solution x gbest ; On the contrary, the global optimal solution keeps the current global optimal solution unchanged; Step 6: Use the Gauss-Cauchy mutation strategy to update the global optimal solution of the particle: According to the current global optimal solution x gbest Calculate the Gaussian mutation operator as follows: According to the current global optimal solution x gbest Calculate the Cauchy mutation operator as follows: Then the current global optimal solution is updated as: Among them, Gaussian (μ1, σ1) is Gaussian distribution, μ1 is the expectation of Gaussian distribution, σ1 is the variance of Gaussian distribution; Cauchy(μ2,σ2) is the Cauchy distribution, μ2 is the location parameter of the Cauchy distribution, and σ2 is the scale parameter of the Cauchy distribution; Step 7: Add 1 to the number of iterations t to determine whether the number of iterations t exceeds the maximum number of iterations T. If t≤T, jump to step 2 to continue execution; if t>T, end the solution and output the current global optimal solution, which is the UUV cluster recovery mission planning solution.

3. A UUV cluster recovery mission planning method as claimed in claim 2, characterized in that: In step 2, the particle velocity is updated using an adaptive weight coefficient: w h and w l are the maximum and minimum values ​​of the weight coefficient respectively.

4. A UUV cluster recovery mission planning method as claimed in claim 3, characterized in that: Take w h =0.9,w l =0.4.

Images