A multi-objective optimization design method for UUV fin-rudder layout profile

Abstract

The application discloses a UUV fin rudder layout shape multi-objective optimization design method, and particularly relates to the field of underwater vehicle design. The method comprises the following steps: taking the shape parameters of the fin rudder as to-be-optimized variables, and establishing a target function and constraint conditions; a real function calculation framework is built; the initial value of the design variable is taken as an initial sample, and a first set is constructed together with an initial target value; a second set is constructed by using non-dominated solutions in the initial sample; a RBF surrogate model is established by using the samples in the first set and the target function value; based on the RBF surrogate model, a constraint domain search strategy, a convergence exploration strategy and a diversity improvement strategy are sequentially executed in a multi-objective evolutionary algorithm to obtain a potential global optimal solution set; and when the potential global optimal solution set reaches a first preset condition, the potential global optimal solution set is output as a most promising solution set. The number of real function calls can be effectively reduced, thereby reducing the calculation cost of simulation optimization.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of underwater vehicle shape design, and particularly relates to a UUV fin rudder layout shape multi-objective optimization design method. BACKGROUND

[0002] An unmanned underwater vehicle (UUV) is widely used in seabed mapping, marine environment monitoring, marine resource development, underwater obstacle search positioning and other scenarios. The fin rudder layout form of the UUV is an important part of the entire design process and is an intuitive representation of the shape characteristics of the UUV. At the same time, the fin rudder layout design directly determines the fluid dynamic performance and navigation characteristics of the UUV. The traditional UUV shape optimization design is driven by numerical simulation methods such as CFD analysis, and directly outputs the results. However, due to the time-consuming nature of CFD simulation calculation, and the need for repeated simulation analysis for comprehensive evaluation during the optimization process, the calculation cost of the simulation-driven optimization method is very expensive, and the calculation efficiency is significantly reduced. SUMMARY

[0003] The main purpose of the present application is to provide a UUV fin rudder layout shape multi-objective optimization design method, which aims to solve the problem of excessive simulation analysis call times in the existing method, which reduces the calculation efficiency.

[0004] To achieve the above purpose, the present application provides a UUV fin rudder layout shape multi-objective optimization design method, comprising: step 1, taking the shape parameters of the fin rudder as the to-be-optimized variables, establishing a target function, the target function including the balance attack angle and the balance rudder angle, and taking the longitudinal dynamic stability of the UUV as the constraint condition; step 2, generating initial to-be-optimized variables, performing real function calculation on the initial to-be-optimized variables to obtain real target values, and updating the number of real function calculations; wherein the target values include the target function values and the constraint condition values; step 3, taking the initial to-be-optimized variables as initial samples and the corresponding real target values to construct a first set; using the non-dominated solutions in the initial samples to construct a second set; step 4, using the samples in the first set and the corresponding real target values to establish an RBF surrogate model; step 5, taking all non-dominated solutions in the second set as an initial population; step 6, performing a convergence exploration strategy on the initial population according to the real target values and the predicted target values to determine potential optimal individuals; wherein the predicted target values are obtained by the RBF surrogate model; step 7, performing a diversity improvement strategy on the potential optimal individuals to obtain a potential global optimal solution set; step 8, when the potential global optimal solution set meets a first preset condition, outputting the potential global optimal solution set as the most promising solution set.

[0005] Optionally, the step 8 comprises: performing real function calculation on the samples in the potential global optimal solution set which have not been calculated, obtaining real target values, and updating the number of real function calculations; and when it is determined that the number of real function calculations reaches a preset number of times, outputting the potential global optimal solution set as the most promising solution set.

[0006] Optionally, the step 5 further comprises: when it is determined that there is no feasible solution in the second set, performing a constraint domain search strategy on the space composed of the samples in the first set to obtain a local optimal solution; the step 7 further comprises: after obtaining the potential global optimal solution set, adding the local optimal solution to the potential global optimal solution set; and the step 8 comprises: performing real function calculation on the samples in the potential global optimal solution set which have not been calculated, obtaining real target values, and updating the number of real function calculations; and when it is determined that the number of real function calculations reaches a preset number of times, outputting the potential global optimal solution set as the most promising solution set.

[0007] Optionally, performing the constraint domain search strategy on the space composed of the samples in the first set comprises: determining constraint violation values of the samples in the first set according to constraint condition values, and selecting an individual with the smallest constraint violation value as a base point; calculating Euclidean distances between the remaining individuals in the first set and the base point, and sorting [constraint violation value, -Euclidean distance] in ascending order, and determining a potential space according to the constraint violation value or the Euclidean distance; and using a sequential quadratic programming algorithm to locate a feasible region in the potential space, and combining a multi-start search strategy to search for a local optimal solution.

[0008] Optionally, the step 8 further comprises: when it is determined that the number of real function calculations does not reach a preset number of times, adding the evaluated real target values and the corresponding samples to the first set, performing non-dominated sorting on the samples in the first set based on the real target values, updating the second set; and returning to the step 4 until the potential global optimal solution set reaches a first preset condition.

[0009] Optionally, in the step 8, when it is determined that the number of samples which have not been calculated is 0, a plurality of new samples are formed based on an optimized Latin hypercube sampling method, real function calculation is performed on the new samples to obtain real target values, and the number of real function calculations is updated.

[0010] Optionally, the fin-rudder shape parameters comprise a chord length of the fin plate, a distance from a fin leading edge to a UUV head, a chord length of the rudder plate, a distance from a rudder leading edge to the UUV head, a fin plate leading edge trailing edge angle, and a rudder plate leading edge trailing edge angle.

[0011] Optionally, the real function calculation is obtained by calling an automatic simulation framework, and a simulation method of the automatic simulation framework comprises: determining profile value points of upper and lower surfaces of the fin rudder according to the input optimization variable; establishing a UUV geometric model by using the profile value points of the upper and lower surfaces of the fin rudder, and generating a flow field model; determining fluid dynamic parameters of the UUV under different attack angles, fluid dynamic parameters of the UUV under different rudder angles, and fluid dynamic parameters of the UUV under different non-dimensional angular velocities on the flow field model respectively; wherein the fluid dynamic parameters comprise lift coefficients and pitching moment coefficients; determining position derivatives of the fluid dynamic parameters with respect to the attack angle and the rudder angle, and determining rotation derivatives of the fluid dynamic parameters with respect to the non-dimensional angular velocity, and determining a target function value according to the position derivatives of the fluid dynamic parameters with respect to the attack angle and the rudder angle; and determining a constraint condition value according to the rotation derivatives.

[0012] Optionally, step 6 comprises: step 6.1, taking the initial population as parent individuals; step 6.2, generating offspring individuals by using the parent individuals, determining target values of the offspring individuals by using the RBF surrogate model, and fusing the parent individuals and the offspring individuals to form a new population; step 6.3, performing population normalization by using the target values of the new population, and determining a convergence index of each individual; step 6.4, generating a set of uniformly distributed reference vectors; step 6.5, determining an included angle between each individual of the current population and each reference vector, and performing an association operation on the individual and the nearest reference vector; step 6.6, determining a constraint violation value and a convergence index according to the target value of the individual, performing non-dominated sorting on the individual associated with the reference vector by using the constraint violation value and the convergence index, and taking an individual with the minimum constraint violation value or convergence index as a parent of the next generation; and step 6.7, returning to step 6.2 until a second preset condition is reached, and obtaining a potentially better individual.

[0013] Optionally, step 7 comprises: determining the number of non-dominated solutions in the potentially better individual, taking all the non-dominated solutions as new samples when the number of the non-dominated solutions is less than a preset value, and performing secondary selection on all the non-dominated solutions by using a maximum penalty angle when the number of the non-dominated solutions is greater than the preset value, and taking a selection result as new samples.

[0014] Compared with the prior art, the application has the following beneficial effects:

[0015] The UUV fin rudder layout shape multi-objective optimization design method provided by the application takes the fin and rudder airfoil and the relative position of the fin and rudder as design variables, models the UUV main body and the fin and rudder, takes the minimum balance attack angle and balance rudder angle of the UUV as a design target, and takes the longitudinal dynamic stability as a constraint condition, to build an automatic simulation calculation framework of the UUV shape fluid dynamics; the design variables are optimized by combining the constraint multi-objective evolutionary algorithm based on the agent assistance, the target function and the constraint condition are predicted by the agent assistance in the optimization process, the number of CFD simulations is reduced, and the UUV fin rudder shape with excellent comprehensive performance can be obtained under a limited number of real function calculations, thereby improving the efficiency of the UUV fin rudder shape design. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 It is a flowchart of the UUV fin rudder layout shape multi-objective optimization design method of the application;

[0017] Figure 2 It is a shape diagram of the fin and rudder of the underwater vehicle in the embodiment of the UUV fin rudder layout shape multi-objective optimization design method of the application.

[0018] The implementation, functional features and advantages of the application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0019] To make the purpose, technical scheme and advantages of the application more clear, the technical scheme of the application will be described clearly and completely in the following with reference to the drawings in the application. Obviously, the described embodiments are some embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the application.

[0020] The first embodiment of the application provides a UUV fin rudder layout shape multi-objective optimization design method, as shown in Figure 1 The specific steps include the following steps:

[0021] Step 1, taking the shape parameters of the fin and rudder as the to-be-optimized variables, respectively establishing the target functions, the target functions including the balance attack angle and the balance rudder angle, and taking the longitudinal dynamic stability of the UUV as the constraint condition;

[0022] The fin and rudder shape parameters include the chord length L F of the fin plate, the distance R1 from the fin leading edge to the UUV head, the chord length R F of the rudder plate, the distance R2 from the rudder leading edge to the UUV head, the fin plate leading edge trailing edge angle x0 and the rudder plate leading edge trailing edge angle x1, and the to-be-optimized variables X = [L F ,R1,R Ff(X), g(X) subject to g(X) < 0

[0023] Minimize f m (X)m=1,2;

[0024] subject to-g(X)<0

[0025] g(X)-1<0

[0026] In the formula, X is a variable to be optimized; f m (X) is a target function, m=1 represents a balanced attack angle, and m=2 represents a balanced rudder angle; g(X) is a constraint condition.

[0027] Step 2, generate initial variables to be optimized, perform real function calculation on the initial variables to be optimized, obtain real target values including target function values and constraint condition values, and update the number of real function calculations, the number of real function calculations NFE=NFE0. Wherein, the real function calculation is obtained by calling the automatic simulation framework, the automatic simulation framework is a UUV shape fluid dynamics real function calculation framework, which can realize input of the variables to be optimized X, automatic fluid dynamics simulation, extraction of fluid dynamics parameters, calculation of f m (X) and g(X), step 2 specifically includes the following steps:

[0028] Step 2.1, generate initial variables to be optimized, i.e. fin rudder shape parameters, by using optimization Latin hypercube sampling;

[0029] Step 2.2, input the initial variables to be optimized into the automatic simulation framework to obtain the real target values. The specific steps are as follows:

[0030] Step 2.2.1, determine the loft points of the upper and lower surfaces of the fin rudder according to the fin rudder shape parameters through MATLAB programming software;

[0031] Step 2.2.2, call CAD secondary development technology to establish a UUV geometric model and generate a flow field model by using the loft points of the upper and lower surfaces of the fin rudder, and perform mesh division on the flow field model and make appropriate mesh encryption in the fin rudder area;

[0032] Step 2.2.3, determine the fluid dynamics parameters of the UUV under different attack angles, different rudder angles and different dimensionless angular velocities on the grid cells of the flow field model respectively; wherein, the fluid dynamics parameters include lift coefficient and pitching moment coefficient;

[0033] Step 2.2.4, determine the position derivative of the lift coefficient with respect to the angle of attack, determine the position derivative of the lift coefficient with respect to the rudder angle; determine the position derivative of the pitching moment coefficient with respect to the angle of attack, determine the position derivative of the pitching moment coefficient with respect to the rudder angle; determine the rotational derivative of the lift coefficient with respect to the non-dimensional angular velocity, determine the rotational derivative of the pitching moment coefficient with respect to the non-dimensional angular velocity; according to the position derivatives and the rotational derivatives obtained above, determine the longitudinal dynamic stability (i.e. constraint condition value) and the target function value, and the specific formula is as follows:

[0034] The calculation formula of the equilibrium angle of attack f1(X) and the equilibrium rudder angle f2(X) is as follows:

[0035]

[0036] In the formula, ΔG is the negative buoyancy of the UUV, G is the gravity of the UUV; is the relative longitudinal distance between the center of gravity and the center of buoyancy, and if the center of gravity is in front of the center of buoyancy, it is positive; is the position derivative of the pitching moment coefficient with respect to the rudder angle, is the position derivative of the lift coefficient with respect to the rudder angle, is the position derivative of the pitching moment coefficient with respect to the angle of attack, is the position derivative of the lift coefficient with respect to the angle of attack;

[0037] The calculation formula of the longitudinal dynamic stability g(X) is as follows:

[0038]

[0039] In the formula, is the relative density; C x is the drag coefficient, taking the maximum cross-sectional area S of the UUV as the characteristic area; is the rotational derivative of the lift coefficient with respect to the non-dimensional angular velocity; is the rotational derivative of the pitching moment coefficient with respect to the non-dimensional angular velocity.

[0040] Step 3, take the initial to-be-optimized variables as initial samples, and construct a first set with the corresponding true target values; construct a second set using the non-dominated solutions in the initial samples;

[0041] Specifically, a first set is constructed using initial samples and initial target values; the initial samples are non-dominantly sorted according to a dominance relationship to obtain a Pareto non-dominant solution set, i.e., a second set. Specifically, if sample A is better than sample B in at least one target or constraint and is not worse than sample B in all targets and constraints, sample A dominates sample B. The dominance levels of sample points in the sample library are divided according to the dominance relationship, and a Pareto non-dominant solution set from the sample points to the dominance levels is established according to the sorting results of the sample points. The Pareto non-dominant solution set is composed of sample points with a dominance level of 1, and the solution set is referred to as a Pareto front.

[0042] Step 4, an RBF proxy model is established using samples in the first set and corresponding real target values;

[0043] Step 5, non-dominant solutions in the second set are used as an initial population; and when it is determined that there is no feasible solution in the second set, a constraint domain search strategy is performed on a space formed by all samples in the first set in combination with a constraint condition, and the constraint domain search strategy stops when a feasible solution is found, to obtain a local optimal solution Local.

[0044] The constraint domain search strategy uses a space reduction method to reduce the search range of the domain, and can improve the search efficiency of the feasible solution. The constraint domain search strategy specifically includes:

[0045] The constraint violation value of each sample in the first set is determined according to the constraint condition, and an individual with the smallest constraint violation value is selected as a base point. The constraint violation value (CV) is defined as:

[0046]

[0047] In the formula, g j (x) and h k (x) are j inequality constraints and k equality constraints, respectively.

[0048] In the fin-rudder layout optimization problem, only inequality constraint conditions exist, and the specific calculation method of the CV value is as follows:

[0049] CV(x)=max{0,-g(x)}+max{0,g(x)-1}

[0050] The Euclidean distances of the remaining individuals in the first set from the base point are calculated as I1; the CV values of the remaining individuals are marked as I2. [I2, -I1] is sorted in ascending order, i.e., the CV values are sorted in ascending order first, and when the CV values are equal, the individual with a larger Euclidean distance is selected, and finally the first Number2 individuals with smaller CV values or larger Euclidean distances (when the CV values are equal) are selected. The range of these selected individuals is marked as a potential space;

[0051] The feasible solution is determined in the potential space by combining a sequence quadratic programming (SQP) algorithm and a multi-start search strategy. The multi-start search strategy is used to determine the feasible solution, which can avoid the SQP algorithm from falling into a local optimum. The multi-start search strategy is specifically that, in the potential space, a plurality of starting points are sampled by using a Latin sampling method, and the feasible solution is searched simultaneously.

[0052] Step 6, a convergence exploration strategy is performed on the initial population to determine a potentially better individual; wherein, in the execution process, a predicted target value of the individual is obtained through an RBF surrogate model; and the specific steps are as follows:

[0053] Step 6.1, the initial population is taken as a parent individual;

[0054] Step 6.2, a child individual is generated by using the parent individual, and a target value of the child individual is determined by using the RBF surrogate model, and a new population is formed by fusing the parent individual and the child individual;

[0055] Step 6.3, the target function value of the new population is used for population normalization, and a convergence index of each individual is determined; the calculation formula of the convergence index CI is as follows:

[0056]

[0057] In the formula, f i (x)-z i * is a target function value conversion value of each individual, z i * = min{f i (x)|x∈P}, which indicates the minimum value of each target function in the current population.

[0058] The detailed steps of population normalization are as follows:

[0059] (i) Calculation of ideal point: the ideal point refers to the minimum value of each target function value in the current population. For each target function f i , the ideal point is calculated as follows: Wherein, P is the current population.

[0060] (ii) Calculation of extreme point: the extreme point refers to the point farthest in the direction of each target function in the target space. These points are used to construct a hyperplane for normalization. For each target function f i , find the individual x that maximizes f i (x).

[0061] (iii) Constructing a hyperplane: a hyperplane is constructed using the extreme points, which will be used to normalize the objective function values. The equation of the hyperplane can be determined by the coordinates of the extreme points. The intercept a of the hyperplane on each objective can be obtained by using the equation of the hyperplane i .

[0062] (iv) Normalizing the objective function values: the objective function values f i (x) Normalized to the same scale. The normalization formula is usually:

[0063] Step 6.4, generating a set of reference vectors covering the entire objective function space; specifically, the number of vectors is where M is the number of objectives, and H is the number of divisions of the objectives. The size of each value in the vector coordinates can be j is the number of objectives. And the vector coordinates should satisfy For example, there are 2 objective functions in this problem. If each objective is divided into 5 parts, the number of vectors should be The size of each value in the vector coordinates can be S j ∈{0,0.2,0.4,0.6,0.8,1}, and should satisfy the sum of 1. For example, 6 reference vectors are: (0,1), (1,0), (0.2,0.8), (0.8,0.2), (0.4,0.6), (0.6,0.4).

[0064] Step 6.5, determine the angle between the individuals of the current population and each reference vector, and associate the individual with the nearest (smallest angle) reference vector;

[0065] Step 6.6, using the constraint violation value and the convergence index, non-dominant sorting of the individuals associated with the reference vectors, the individual with the smallest constraint violation value or convergence index as the next generation parent; the sorting process is as follows:

[0066]

[0067] Step 6.7, return to step 6.2 until the second preset condition is met, and get the potentially better individuals. The second preset condition can be a preset number of cycles.

[0068] Step 7, performing a diversity improvement strategy on the potentially better individuals to obtain a set of potentially global optimal solutions; in another embodiment, when there is no feasible solution in the second set in step 5, the local optimal solution is added to the set of potentially global optimal solutions as the final set of potentially global optimal solutions.

[0069] Specifically, step 7.1, the potential better individuals are non-dominantly sorted according to the objective function value, to obtain a non-dominant solution set O, the number of non-dominant solutions in the potential better individuals is determined, that is, the number of samples in the non-dominant solution set O, when it is determined that the number of non-dominant solutions is less than a preset value Number1, all non-dominant solutions are taken as new samples; if it is determined that the number of non-dominant solutions is greater than the preset value Number1, a maximum penalty angle is used to select twice in all non-dominant solutions, so that the number of non-dominant solutions is less than the preset value Number1, to obtain new samples.

[0070] The "maximum penalty angle" secondary selection method includes:

[0071] Step a, the objective function values of each individual in the non-dominant solution set O and the second set are respectively converted into f i (x)-z i * ;

[0072] Step b, the CV value of each individual in O is calculated and normalized to be used as a penalty factor λ(x); the calculation formula of the penalty factor λ(x) is:

[0073]

[0074] In the formula, CV max = max{CV(x) | x e O}, which is the maximum value of the CV value in the current solution set; CV min = min{CV(x) | x e O}, which is the minimum value of the CV value in the current solution set.

[0075] Step c, the Euclidean distance between each individual in O and all individuals in the second set is calculated, the sample point in the second set closest to each sample point in O is found, and the angle size α(x) between the two sample points is calculated. α(x) is the selection index of the "maximum penalty angle" secondary selection. The penalty factor λ(x) is used to modify α(x) to reduce the possibility of selecting individuals with infeasible or large CV values, to obtain a penalty angle, and the calculation formula is:

[0076] fitness(x) = [1-λ(x)]α(x)

[0077] Step d, in order to enhance the diversity of the algorithm, the sample with the maximum penalty angle in O is selected, removed from O, and composed into a potential global optimal solution set, and added to the second set.

[0078] Step f, steps a to f are repeatedly executed until the number of samples in the potential global optimal solution set reaches a designed value, and the potential global optimal solution set is defined as Global.

[0079] Step 7.2, add the samples in Global which are not calculated by the real function to new; if there is no feasible solution in the second set in step 5, Global includes Local and Global; if the samples in new are not 0, that is, new samples are generated, calculate the real function of the samples in new which are not calculated by the real function, obtain the real target value, and update the number of real function calculations NFE, NFE = NFE + |new|;

[0080] If new is empty, the new sample is 0, form a number of new samples based on the optimized Latin hypercube sampling method, add them to new, calculate the real function of new, obtain the real target value, and update the number of real function calculations.

[0081] It is worth noting that after each loop is completed, new is emptied.

[0082] Step 8, when it is determined that the potential global optimal solution set reaches the first preset condition, output the potential global optimal solution set as the most promising solution set. Specifically, it includes:

[0083] When it is determined that the number of real function calculations does not reach the preset number of times, add the samples in set new and the corresponding real target value to the first set, sort the samples in the first set based on the real target value, update the second set; return to step 4 until the number of real function calculations reaches the preset number of times, output the potential global optimal solution set as the most promising solution set.

[0084] In order to select the optimal solution in the fin rudder layout optimization problem, the most promising solution set can be further processed, specifically: calculate the CV value of each individual in the most promising solution set, if the CV value is not 0, it means that the individual does not meet the constraint condition, and it is removed from the solution set; based on the objective function value, sort the remaining individuals to obtain the Pareto solution set. Further, according to the actual demand, select the optimal solution of the problem from the Pareto solution set, and output the aircraft fin rudder shape with excellent comprehensive performance.

[0085] In this embodiment, the constraint multi-objective evolutionary algorithm CE-SAEA based on agent assistance is used for optimization. The algorithm uses RBF proxy model for prediction, which can effectively reduce the number of real function calls, thereby reducing the calculation cost of simulation optimization; the constraint domain search strategy can quickly locate the feasible region through multi-start search and space contraction technology, while improving the continuity of the algorithm in the case of infeasible feasible region; the convergence exploration strategy selects the promising individuals related to each effective reference vector to promote the population to converge to the global optimal solution; the diversity improvement strategy reduces the possibility of selecting infeasible solutions or individuals with large CV values through the concept of "maximum penalty angle", enhancing the diversity of the algorithm.

[0086] Embodiment

[0087] Step 1: Establishing the parametric model of the fin-rudder shape

[0088] The UUV body of this embodiment selects MK46 vehicle, and the fin-rudder layout adopts "cross" fin "X" rudder, and the fin-rudder shape adopts flat plate shape. The total length of the UUV is 2790mm, and the maximum diameter is 323.85mm. The value range of the design variable is determined according to the main size of the UUV. The design variable of the fin-rudder shape is the chord length of the fin plate, the distance from the fin leading edge to the UUV head, the chord length of the rudder plate, the distance from the rudder leading edge to the UUV head, the fin plate leading edge trailing edge angle and the rudder plate leading edge trailing edge angle, so the design variable X = [L F ,R1,R F ,R2,x0,x1] parameterization result is shown in Figure 2 .

[0089] Step 2: Establishing the multi-objective optimization problem of fin-rudder shape

[0090] Under the condition of 405RPM rotating speed, taking the minimum balance attack angle and balance rudder angle as the optimization objective, taking the longitudinal dynamic stability as the constraint condition, and taking the design variable of the fin-rudder shape as the optimization variable, the multi-objective optimization problem of the UUV fin-rudder shape is constructed. The objective function and the constraint condition are as follows:

[0091] Minimize f m (X)m=1,2;

[0092] subject to-g(X)<0

[0093] g(X)-1<0

[0094] In the formula, X is the optimization variable; f m (X) is the objective function, m = 1 represents the balance attack angle, and m = 2 represents the balance rudder angle; g(X) is the constraint condition.

[0095] Step 3: build a UUV shape fluid dynamics real function calculation framework

[0096] The initial value of the UUV fin rudder layout design variable is obtained by using Latin hypercube sampling, the initial value of the design variable corresponding to the geometric model is updated by calling a three-dimensional modeling software in a CFD simulation management platform through CAD secondary development technology, and geometric modeling of different fin rudder shapes is automatically realized; the grid division of the flow field model is performed by calling Fluent Meshing, and the grid is encrypted in the fin rudder area; the CFD simulation of the UUV is performed by calling Fluent, and the corresponding fluid dynamic parameters, including the lift coefficient and the pitch moment coefficient, are obtained; and the extraction of the fluid dynamic parameters, the processing of the fluid dynamic parameters, the output of the objective function value and the constraint condition value are automatically completed in the programming software.

[0097] Step 4: optimization model initialization

[0098] The initial value of the design variable is used as an initial sample, and the first set is constructed with the initial target value; the non-dominated solution in the initial sample is used to construct the second set;

[0099] Step 5: optimization algorithm optimization

[0100] The sample in the first set and the objective function value are used to establish an RBF surrogate model; based on the RBF surrogate model, the constraint domain search strategy, the convergence exploration strategy and the diversity improvement strategy are executed in the multi-objective evolutionary algorithm in turn, and a potential global optimal solution set is obtained; and when the potential global optimal solution set reaches the first preset condition, the potential global optimal solution set is output as the most promising solution set.

[0101] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application, and any equivalent structure or equivalent process transformation using the content of the specification and the drawings, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A multi-objective optimization design method for the fin-rudder layout of a UUV, characterized in that, include: Step 1: Take the shape parameters of the fin and rudder as variables to be optimized, establish an objective function, which includes the balanced angle of attack and the balanced rudder angle, and use the longitudinal dynamic stability of the UUV as a constraint condition. Step 2: Generate initial variables to be optimized, perform true function calculations on the initial variables to be optimized to obtain true target values, and update the number of true function calculations; wherein, the target value includes the target function value and the constraint condition value; Step 3: Use the initial variable to be optimized as the initial sample and construct a first set with the corresponding true target value; construct a second set using the non-dominated solutions in the initial sample; Step 4: Build an RBF proxy model using the samples in the first set and the corresponding real target values; Step 5: Use all non-dominated solutions in the second set as the initial population; Step 6: Based on the actual target value and the predicted target value, execute a convergence exploration strategy on the initial population to determine potential superior individuals; The target value is obtained by predicting using the RBF proxy model. Step 7: Implement a diversity improvement strategy on the potential better individuals to obtain a set of potential global optimal solutions; Step 8: When the set of potential global optimal solutions meets the first preset condition, output the set of potential global optimal solutions as the most promising solution set; Step 5 also includes: When it is determined that there is no feasible solution in the second set, a constraint domain search strategy is performed on the space composed of samples in the first set to obtain a local optimum solution. Step 7, after obtaining the set of potential global optimal solutions, further includes: adding the local optimal solutions to the set of potential global optimal solutions; Step 8 includes: performing real function calculations on samples in the potential global optimal solution set that have not undergone real function calculations to obtain the true target value, and updating the number of real function calculations; When the number of computations of the true function reaches a preset value, the set of potential globally optimal solutions is output as the most promising solution set. The execution of a constraint domain search strategy on the space composed of samples in the first set includes: The constraint violation value of each sample in the first set is determined based on the constraint condition value, and the individual with the smallest constraint violation value is selected as the base point. Calculate the Euclidean distance between the remaining individuals in the first set and the base point, sort the [constraint violation value, -Euclidean distance] in ascending order, and determine the potential space based on the constraint violation value or Euclidean distance; The feasible region is located in the potential space using a sequential quadratic programming algorithm, and a local optimum is obtained by combining a multi-starting point search strategy.

2. The multi-objective optimization design method for the UUV fin rudder layout according to claim 1, characterized in that, Step 8 also includes, when it is determined that the number of times the true function is calculated has not reached the preset number of times, adding the evaluated true target value and the corresponding sample to the first set, performing non-dominated sorting on the samples in the first set based on the true target value, and updating the second set; Return to step 4 until the set of potential global optimal solutions reaches the first preset condition.

3. The multi-objective optimization design method for the UUV fin rudder layout shape according to claim 1, characterized in that, In step 8, when the number of samples for which no true function calculation has been performed is determined to be 0, several new samples are generated based on the optimized Latin hypercube sampling method. The true function is then calculated on the new samples to obtain the true target value, and the number of true function calculations is updated.

4. The multi-objective optimization design method for the UUV fin rudder layout according to claim 1, characterized in that, The fin-rudder shape parameters include the chord length of the fin, the distance from the leading edge of the fin to the head of the UUV, the chord length of the rudder, the distance from the leading edge of the rudder to the head of the UUV, the leading and trailing edge angles of the fin and the rudder.

5. The multi-objective optimization design method for the UUV fin rudder layout according to claim 1, characterized in that, The actual function is calculated by calling an automated simulation framework, and the simulation method of the automated simulation framework includes: Based on the input variables to be optimized, determine the shape value points of the upper and lower surfaces of the fins and rudders; Using the shape points on the upper and lower surfaces of the fins, a UUV geometric model is established, and a watershed model is generated; On the aforementioned watershed model, the hydrodynamic parameters of the UUV at different angles of attack, different rudder angles, and different dimensionless angular velocities were determined respectively. The hydrodynamic parameters include the lift coefficient and the pitching moment coefficient. The positional derivatives of the hydrodynamic parameters as a function of angle of attack and rudder angle are determined, and the rotational derivatives of the hydrodynamic parameters as a function of dimensionless angular velocity are determined. Based on the positional derivatives of the hydrodynamic parameters as a function of angle of attack and rudder angle, the objective function value is determined. The constraint values ​​are determined based on the rotational derivative.

6. The multi-objective optimization design method for the UUV fin rudder layout according to claim 1, characterized in that, Step 6 includes: Step 6.1: Use the initial population as the parent individuals; Step 6.2: Generate offspring individuals using the parent individuals, determine the target value of the offspring individuals using the RBF surrogate model, and merge the parent individuals and offspring individuals to form a new population; Step 6.3: Perform population normalization using the target value of the new population, and determine the convergence index for each individual; Step 6.4: Generate a set of uniformly distributed reference vectors; Step 6.5: Determine the angle between each individual in the current population and each reference vector, and associate the individual with the nearest neighboring reference vector; Step 6.6: Determine the constraint violation value and convergence index based on the individual's target value. Using the constraint violation value and convergence index, perform non-dominated sorting on the individuals associated with the reference vector, and take the individual with the smallest constraint violation value or convergence index as the next generation parent. Step 6.7: Return to step 6.2 until the second preset condition is met, and a potentially better individual is obtained.

7. The multi-objective optimization design method for the UUV fin rudder layout according to claim 1, characterized in that, Step 7 includes: Determine the number of non-dominated solutions among potential better individuals. If the number of non-dominated solutions is less than a preset value, treat all non-dominated solutions as new samples. When the number of non-dominated solutions is greater than a preset value, a secondary selection is performed on all non-dominated solutions using the maximum penalty angle, and the selection result is used as a new sample.

Images