A method for simulating and optimizing in-bore spin reduction and overload of unguided artillery shells

Abstract

The application discloses a method for optimizing design of in-bore spin-down and overload simulation of terminal guided cannonball, which comprises the following steps: establishing a finite element simulation model of cannonball-cannon coupling dynamics of the terminal guided cannonball; determining key characteristic parameter structure design variables of the terminal guided cannonball; adopting an optimal Latin hypercube experimental design method to sample; updating and modifying the finite element model of the terminal guided cannonball, then establishing a BP neural network proxy model of the terminal guided cannonball, and further verifying the accuracy of the BP neural network proxy model; establishing an optimization model of the terminal guided cannonball; and optimizing and solving the optimization model of the terminal guided cannonball. Compared with the prior art, the method not only optimizes the key characteristic parameter values of the terminal guided cannonball, but also obtains an optimal design scheme, the optimized scheme has better performance, the in-bore spin-down and overload of the terminal guided cannonball are reduced, and the key characteristic parameter factors of the terminal guided cannonball are fully considered, which conforms to the engineering practice and has obvious engineering significance.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of terminal guided cannonball simulation optimization, and particularly relates to a terminal guided cannonball bore spin reduction and overload simulation optimization method. BACKGROUND

[0002] As a new type of intelligent ammunition, the terminal guided cannonball is launched by a traditional cannon barrel. Due to the installation of a control system inside the cannonball, the bore overload control and the cannonball spin reduction are a difficult problem to be solved compared with conventional uncontrolled ammunition. During the launching process of the conventional uncontrolled ammunition in the cannon bore, the cannonball will obtain a very high muzzle velocity under the action of the cannon barrel rifling to ensure the stability of the external trajectory. However, the terminal guided cannonball adopts a tail wing stabilization, and the muzzle velocity should not be too high. The high muzzle velocity will cause the electronic components inside the cannonball to fall off or work unstably, and the high muzzle velocity is not conducive to the terminal trajectory correction.

[0003] In the paper "Terminal Guided Cannonball Bore Launching Mechanics Analysis and Test Device Protection Structure Optimization Research" by Yin Weihua, a cannonball system launching finite element model is established, and the terminal guided cannonball bore overload and muzzle velocity under a single factor are analyzed. In the paper "Dynamic Characteristics of Sliding Plastic Band Guided Cannonball Bore Spin Reduction Process" by Li Zhaohui, the influence of different parameters on the muzzle velocity of the terminal guided cannonball is also analyzed. The above papers only analyze the influence of a single factor on the muzzle velocity and bore overload of the terminal guided cannonball, and do not consider the influence of the key characteristic parameters of the terminal guided cannonball on each other. Therefore, the key characteristic parameters of the terminal guided cannonball are optimized, the target performance is controlled within a reasonable range, and a reasonable parameter design scheme is obtained. SUMMARY

[0004] In order to solve the key characteristic parameter optimization design problem in the process of cannon design, a terminal guided cannonball key characteristic parameter optimization design method is provided. By comprehensively considering the influence of the key characteristic parameters of the terminal guided cannonball on each other, and optimizing the simulation model and simulation results, the cannonball bore overload and the cannonball velocity are reduced.

[0005] The specific steps are as follows:

[0006] Step 1: Establish a terminal guided cannonball cannonball coupling dynamics finite element simulation model;

[0007] Step 2: Determine the terminal guided cannonball key characteristic parameter structure design variable;

[0008] Step 3: Sample the samples by using the optimal Latin hypercube design method;

[0009] Step 4: Update and modify the terminal guided cannonball finite element model, and then establish a terminal guided cannonball BP neural network proxy model, and further verify the accuracy of the BP neural network proxy model;

[0010] Step 5: Establishing an optimization model of the terminal guided shell;

[0011] With the structural design variables of the key characteristic parameters of the terminal guided shell as constraint conditions, the shell rotating speed and the chamber overload as optimization objective functions, and the initial speed (the initial speed is greater than the minimum muzzle velocity) and the maximum chamber pressure as constraint functions, an optimization model of the terminal guided shell is established as follows:

[0012]

[0013] Wherein, m is the mass of the shell, e1 is the front centering portion gap, e2 is the rear centering portion gap, d1 is the band width, d2 is the sliding ring width, m1 and m2 are the mass of the thin and thick gunpowder in the launching charge respectively, l is the distance from the band rear end face to the front centering portion, i=1 or 2, d min , e min , m min are the minimum values of the design variables of the key characteristic parameters of the terminal guided shell respectively, d max , e max , m max are the maximum values of the design variables of the key characteristic parameters of the terminal guided shell respectively;

[0014] Step 6: Optimization solving of the optimization model of the terminal guided shell.

[0015] Compared with the prior art, the present application has the following advantages:

[0016] (1) The optimization method established in the present application fully considers the launching charge parameters and the terminal guided shell structural parameters, and is more in line with the engineering practice.

[0017] (2) The simulation optimization method established in the present application not only optimizes the terminal guided shell structural parameters, but also obtains the launching charge parameters, thereby providing a theoretical basis for shell processing and manufacturing, and having practical engineering significance.

[0018] (3) The simulation optimization method established in the present application is based on the numerical simulation model and the BP neural network surrogate model of the terminal guided shell, and has both high calculation accuracy and high calculation efficiency.

[0019] (4) The simulation optimization method established in the present application fully considers the terminal guided shell structural parameters and the launching charge parameters, and the finally obtained optimization design scheme has good inhibitory effect on the shell chamber overload and rotation reduction and higher reliability. DETAILED DESCRIPTION

[0020] Figure 1 Flow chart of the terminal guided shell key characteristic parameter optimization design method;

[0021] Figure 2(a) the finite element mesh model diagram of the projectile body;

[0022] Figure 2 (b) the finite element mesh model diagram of the belt and sliding ring;

[0023] Figure 2 (c) the finite element mesh model diagram of the rifled barrel;

[0024] Figure 3 BP neural network agent model construction flow chart;

[0025] Figure 4 Optimized Pareto optimal solution set;

[0026] Figure 5 (a) is the comparison of the terminal guided projectile rotating speed before and after optimization;

[0027] Figure 5 (b) is the comparison of the terminal guided projectile axial overload in the bore before and after optimization;

[0028] 1-guide, 2-control cabin, 3-warhead, 4-fin cabin, 5-belt, 6-sliding ring, 7-barrel, 8-rifling. DETAILED DESCRIPTION

[0029] The design examples of the application will be described in detail below in combination with the drawings. The information involved in the drawings and examples is the actual application data of the application in the key characteristic parameters of the terminal guided projectile. The process of a terminal guided projectile bore spin reduction and overload simulation optimization design method is shown in Figure 1 The specific simulation optimization method steps are as follows:

[0030] Step 1: Establish a terminal guided projectile and gun coupling dynamics finite element simulation model.

[0031] First, a three-dimensional model of the terminal guided projectile and the rifled barrel is established in the three-dimensional software. The terminal guided projectile includes the projectile body, the belt 5 and the sliding ring 6, wherein the projectile body includes the guide 1, the control cabin 2, the warhead 3, and the fin cabin 4. The rifled barrel includes the barrel 7 and the rifling 8, wherein the belt 5 and the sliding ring 6 are fixedly connected, and the sliding ring 6 is sleeved on the outside of the projectile body and can rotate around the projectile body.

[0032] Secondly, the entity model is imported into the pre-processing software Hypermesh for meshing. As shown in Figure 2 The fin and the rudder of the terminal guided projectile are meshed with shell elements, and the rest are meshed with three-dimensional solid elements. The material properties of each part are assigned to the mesh.

[0033] Step 2: Determine the key characteristic parameter structure design variable of the terminal guided projectile.

[0034] Considering the rationality of the structure design of the terminal guided projectile, the selected key characteristic parameter value variable range of the terminal guided projectile is shown in Table 1, wherein m is the mass of the projectile, e1 is the front centering portion gap, e2 is the rear centering portion gap, d1 is the band width, d2 is the sliding ring width, m1 and m2 are the mass of the thin and thick gunpowder in the launch charge respectively, and l is the distance from the band rear end face to the front centering portion.

[0035] Table 1 Value range of design variables

[0036]

[0037] Step 3: The optimal Latin hypercube experimental design method is adopted to sample the samples. 100 groups of test samples and 10 groups of test samples are uniformly and randomly extracted in the above-mentioned value range of design variables, that is, the test samples and the test samples are represented as 100x8 and 10x8 order matrices respectively, and the sample values are shown in Table 2 and Table 3.

[0038] Table 2 Design table of 100 groups of test samples

[0039]

[0040] Table 3 Design table of 10 groups of test samples

[0041]

[0042]

[0043] Step 4: The finite element model of the terminal guided projectile is updated and modified, and then the BP neural network proxy model of the terminal guided projectile is established. The key characteristic parameters of the terminal guided projectile determined in step 2 and the 100 groups of test sample data extracted in step 3 are used to update and modify the finite element model. According to the modified finite element simulation model of the terminal guided projectile, the evaluation parameters of the design variables corresponding to the bore overload and the spin reduction are solved, and the evaluation parameters specifically include the maximum bore pressure P max , the initial velocity of the projectile V max , the bore overload G, and the projectile speed N.

[0044] By constructing the back propagation neural network (BP neural network) proxy model of the terminal guided projectile, the mapping relationship between the design variables and the bore overload and the spin reduction is established, and the specific process is shown in Figure 3 .

[0045] Step 4.1: Numerical simulation calculation is performed on the modified 100 groups of finite element models of the terminal guided projectile, and the maximum bore pressure P max , the initial velocity of the projectile V max , the bore overload G, and the projectile speed N are obtained, that is, the result sample is represented as a 100x4 order matrix.

[0046] Step 4.2: Establish the BP neural network agent model of the terminal guided shell; take 100 groups of test samples as input and the result samples obtained in step 4.1 as output, normalize the two, and train a five-layer BP neural network agent model in software;

[0047] Step 4.3: Calculate the finite element simulation result of the test sample; substitute 10 groups of test samples into the modified finite element simulation model of the terminal guided shell to calculate the result;

[0048] Step 4.4: Calculate the neural network model result of the test sample; substitute 10 groups of test samples into the neural network agent model to calculate the result;

[0049] Step 4.5: Verify the accuracy of the trained BP neural network agent model; compare the test results calculated by the modified finite element simulation model with the calculation results of the BP neural network model to calculate the determination coefficient R 2 , and verify the accuracy of the BP neural network agent model:

[0050]

[0051] In the formula: S R is the regression sum of squares, S T is the total sum of squares of the response, n is the number of test sample points, y i is the value of the simulation calculation result in the bore, is the average value of y i , is the predicted value of the BP neural network. R 2 is closer to 1, indicating that the accuracy of the agent model is higher.

[0052] Step 4.6: If the accuracy of the neural network agent model is qualified, save the BP neural network agent model, otherwise, perform a new round of loop calculation. The accuracy of the BP neural network agent model is shown in Table 4. It can be seen that the accuracy of the established agent model is high, and the determination coefficient is above 0.9.

[0053] Table 4: Accuracy verification of neural network

[0054]

[0055] Step 5: Establish the optimization model of the terminal guided shell. In order to ensure the normal work of the electronic components in the control cabin as much as possible, the rotational speed of the terminal guided shell when it exits the muzzle and the axial overload of the control cabin part are reduced, the key characteristic parameter structure design variables of the terminal guided shell are taken as constraint conditions, the projectile rotational speed and the bore overload are taken as optimization objective functions, and the initial speed of the projectile (the initial speed is greater than the minimum muzzle exit speed) and the maximum bore pressure are taken as constraint functions, and the optimization model of the terminal guided shell is established as follows:

[0056]

[0057] where i = 1 or 2, d min , e min , m min are the minimum values of the design variables of the key characteristic parameters of the terminal guided projectile, d max , e max , m max are the maximum values of the design variables of the key characteristic parameters of the terminal guided projectile.

[0058] Step 6: Optimize the terminal guided projectile optimization model.

[0059] Step 6.1: Use the fast non-dominated sorting genetic algorithm with elitist strategy (NSGA-II) to search and solve the optimization model in Step 5, obtaining the optimization results of the design variables, as shown in Table 5.

[0060] Step 6.2: Further obtain the pareto optimal solution set, and analyze the results of the pareto optimal solution set, selecting the solution with good interior ballistic performance as the optimal solution, as shown in Figure 4 .

[0061] Step 6.3: Bring the selected optimal solution into the aforementioned modified finite element model of the terminal guided projectile and perform numerical simulation calculation again, obtaining the optimized projectile rotational speed and bore overload results.

[0062] Step 6.4: Compare the obtained bore overload and projectile rotational speed before and after optimization, and the results are shown in Figure 5 Figure 5 . The optimization design scheme has good inhibitory effect on the bore overload and rotational speed reduction of the projectile and higher reliability.

[0063] Table 5 Optimization results of design variables

[0064]

Claims

1. A simulation optimization method for in-bore spin reduction and overload of terminally guided projectiles, characterized in that, The specific steps are as follows: Step 1: Establish a finite element simulation model of the projectile-gun coupling dynamics of the terminal-guided projectile; Step 2: Determine the key characteristic parameters and structural design variables of the terminally guided projectile; Step 3: Sample sampling is performed using the optimal Latin hypercube experimental design method; Step 4: Update and modify the finite element model of the terminal guided projectile, then establish a BP neural network surrogate model for the terminal guided projectile, and further verify the accuracy of the BP neural network surrogate model; Step 5: Establish an optimization model for terminally guided projectiles; Using the key characteristic parameters and structural design variables of the terminal-guided projectile as constraints, the projectile rotation speed and in-chamber overload as optimization objective functions, and the initial velocity and maximum chamber pressure of the projectile as constraint functions, the optimization model of the terminal-guided projectile is established as follows: Where m is the projectile mass, e1 is the gap between the front and rear centering sections, e2 is the gap between the rear and rear centering sections, d1 is the width of the projectile belt, d2 is the width of the sliding ring, m1 and m2 are the masses of the thin and thick propellant in the propellant charge, l is the distance from the rear end face of the projectile belt to the front centering section, i = 1 or 2, d min e min m min These are the minimum design variables for key characteristic parameters of terminally guided projectiles, d max e max m max These are the maximum values ​​of key characteristic parameters of the terminally guided projectile; barrel overload G; and projectile rotation speed N. Step 6: Optimize and solve the terminally guided projectile optimization model.

2. The simulation optimization method for in-bore spin reduction and overload of terminally guided projectiles according to claim 1, characterized in that, The terminally guided projectile gun-missile coupled dynamic model includes the projectile body, the projectile belt, the sliding ring, and the rifled barrel. The projectile body includes the guidance section, the control compartment, the warhead, and the tail fin compartment. The rifled barrel includes the barrel and the rifling. The projectile belt and the sliding ring are fixedly connected, with the sliding ring sleeved on the projectile body and rotating around the projectile body.

3. The simulation optimization method for in-bore spin reduction and overload of terminally guided projectiles according to claim 2, characterized in that, The tail fin and rudder sections of the tail fin compartment are divided using shell element meshes, while the remaining sections are divided using three-dimensional solid element meshes.

4. The simulation optimization method for in-bore spin reduction and overload of terminally guided projectiles according to claim 1, characterized in that, The sample sampling involves uniformly and randomly selecting 100 sets of experimental samples and 10 sets of test samples within the range of design variable values. That is, the experimental samples and test samples are represented as 100×8 and 10×8 matrices, respectively.

5. The simulation optimization method for in-bore spin reduction and overload of terminally guided projectiles according to claim 1, characterized in that, The specific steps in step 4 are as follows: Step 4.1: Perform numerical simulation calculations on the modified terminally guided projectile finite element model to obtain the maximum chamber pressure Pmax, the initial velocity of the projectile v, the chamber overload G, and the projectile rotational speed N. Step 4.2: Establish a BP neural network surrogate model for terminally guided projectiles; Step 4.3: Calculate the finite element simulation results of the test sample; Step 4.4: Calculate the neural network proxy model results for the test samples; Step 4.5: Verify the accuracy of the trained BP neural network surrogate model; compare the test results calculated by the modified finite element model with the calculation results of the BP neural network surrogate model to calculate the coefficient of determination R. 2 Verify the accuracy of the BP neural network surrogate model: In the formula: S R For the regression sum of squares, S T The sum of squares of the responses, n is the number of test sample points, and y i These are the values ​​calculated using in-bore simulation. For y i The average value, R represents the predicted value from the BP neural network. 2 The closer a value is to 1, the higher the accuracy of the surrogate model. Step 4.6: If the accuracy of the neural network surrogate model is satisfactory, save the BP neural network surrogate model; otherwise, perform a new round of iterative calculation.

6. The simulation optimization method for in-bore spin reduction and overload of terminally guided projectiles according to claim 1, characterized in that, The specific steps in step 6 are as follows: Step 6.1: Use a fast non-dominated sorting genetic algorithm with an elitist strategy to search and solve the optimization model in Step 5 to obtain the optimization results of the design variables; Step 6.2: Further obtain the Pareto optimal solution set, and at the same time analyze the results of the Pareto optimal solution set to select the optimal solution; Step 6.3: Substitute the selected optimal solution into the modified terminally guided projectile finite element model and perform numerical simulation calculations again to obtain the optimized projectile rotation speed and in-bore overload results; Step 6.4: Compare the in-bore overload and projectile rotation speed obtained before and after optimization.

Images