L-shaped wave-breaking plate and design optimization method thereof

Abstract

The present invention discloses an L-shaped wave-breaking plate and a design optimization method thereof. The L-shaped wave-breaking plate comprises a transverse plate, a circular plate, a longitudinal plate, a vent, a liquid through hole, and a manhole. The transverse plate has a first edge section that matches the edge of a portion of a tank cross section, one side of the circular plate is connected to the second edge section, and the longitudinal plate is connected to the other side of the circular plate. The vent and liquid through hole are respectively provided at the upper and lower ends of the connection between the transverse plate and the circular plate, and the manhole is provided on the longitudinal plate. The design optimization method comprises constructing a response surface approximation model of the design variables of the L-shaped wave-breaking plate and the liquid sloshing response of a liquid tank truck using the finite element method and the response surface method, constructing a mathematical model for the multi-objective optimization problem of the L-shaped wave-breaking plate, solving the mathematical model using an improved non-dominated sorting genetic algorithm (NSGA-III), and analyzing and determining the optimal design parameter combination of the L-shaped wave-breaking plate. The present invention can effectively suppress the transverse and longitudinal sloshing impact of liquid in a tank while meeting the requirements for lightweighting the wave-breaking plate.

Description

Technical Field

[0001] The present invention relates to the technical field of wave-breaking plates for liquid tank trucks, and in particular to an L-shaped wave-breaking plate and a design optimization method thereof. Background Art

[0002] When a tanker truck is turning, emergency braking, or avoiding obstacles, the liquid inside the tank will slosh violently due to inertia, affecting driving safety. To effectively suppress sloshing of the liquid inside the tank and reduce the impact of the liquid on the head, bulkhead, and cylinder, the tank of the tank truck is generally equipped with a certain number of wave-breaking plates. Currently, the most widely used wave-breaking plates are flat-plate wave-breaking plates and butterfly-shaped wave-breaking plates. These wave-breaking plates can effectively reduce the sloshing of the liquid inside the tank under emergency braking conditions and reduce the longitudinal impact force of the liquid. However, when the tank is not fully loaded and is in a turning condition, especially for large tank trucks, studies have shown that when the filling ratio is greater than 0.5, its roll stability is poor. The lateral sloshing reduction ability of flat-plate wave-breaking plates and butterfly-shaped wave-breaking plates is weak, and they cannot effectively suppress the lateral sloshing impact of the liquid inside the tank, making the tank truck extremely prone to rollover accidents. Although there are currently related technologies that improve the comprehensive anti-sloshing ability of the wave-breaking plates by simultaneously installing both lateral and longitudinal wave-breaking plates inside the tank, this greatly increases the weight of the tank. Therefore, it is urgent to develop an L-shaped wave-breaking plate to effectively suppress the lateral and longitudinal shaking impact of the liquid in the tank and meet the lightweight design requirements of the entire vehicle.

[0003] On the other hand, with the rapid development of design methods, a variety of modern design methods, such as the finite element method, multi-objective optimization design, and computer-aided design, have been widely used. However, existing wave-breaking shroud design methods still rely primarily on engineers' design experience, lacking a rigorous analysis and research process and a standardized optimization design method. In addition, during the wave-breaking shroud optimization design process, different design evaluation indicators may be mutually constrained. For example, lightweight design of a wave-breaking shroud often comes at the expense of its strength. Therefore, there is an urgent need to design a comprehensive optimization design method for the multi-objective optimization design problem of wave-breaking shrouds. By establishing a reasonable multi-objective optimization mathematical model and solving the algorithm based on the characteristics of the model, the optimal combination of wave-breaking shroud design parameters can be analyzed and obtained. Summary of the Invention

[0004] The purpose of the present invention is to provide an L-shaped wave-breaking plate for a liquid tank truck and a design optimization method thereof, which can effectively suppress the lateral and longitudinal shaking impact of the liquid in the tank on the basis of meeting the lightweight requirements of the entire vehicle, and solve the technical problems that the existing wave-breaking plates are insufficient in suppressing the lateral shaking impact of the liquid in the tank, which makes it easy to roll over, and the existing wave-breaking plate design methods lack scientificity.

[0005] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:

[0006] In one aspect, the present invention provides an L-shaped wave-breaking plate, comprising:

[0007] a transverse plate, the transverse plate having a first edge section and a second edge section, the first edge section matching a portion of an edge of a cross section of the tank body;

[0008] a circular arc plate, one side of which is connected to the second edge section of the transverse plate;

[0009] a longitudinal plate connected to the other side of the arc plate;

[0010] The air vent, the liquid hole and the manhole are respectively provided at the upper end and the bottom end of the connection between the horizontal plate and the arc plate, and the manhole is provided on the longitudinal plate.

[0011] In some embodiments, the first edge segment of the transverse plate is formed by five arc segments connected end to end in sequence.

[0012] In some embodiments, the air vent and the liquid hole are circular holes, the manhole is an elliptical hole, and the center of the manhole is offset in the horizontal and vertical directions relative to the symmetry center of the longitudinal plate.

[0013] In some embodiments, the short axis length of the manhole is greater than the diameter of the liquid hole, and the diameter of the liquid hole is greater than the diameter of the air hole.

[0014] In some embodiments, a wave-breaking plate reinforcement ring is further included, and the wave-breaking plate reinforcement ring is welded to the edges of the transverse plate, the arc plate and the longitudinal plate.

[0015] In some embodiments, a liquid hole reinforcement ring is further included, and the liquid hole reinforcement ring is welded to the circumferential edge of the liquid hole.

[0016] In some embodiments, a manhole reinforcement ring is further included, and the manhole reinforcement ring is welded to the circumferential edge of the manhole.

[0017] On the other hand, the present invention provides an L-shaped wave-breaking plate design optimization method, which is used to design the above-mentioned L-shaped wave-breaking plate. The method constructs a response surface approximation model of the design variables of the L-shaped wave-breaking plate and the liquid sloshing response of the liquid tank truck through the finite element method and the response surface method, and constructs a mathematical model of the multi-objective optimization problem of the L-shaped wave-breaking plate. The improved non-dominated sorting genetic algorithm NSGA-III with an elite strategy is used to solve the mathematical model to obtain the Pareto optimal solution set. Finally, the finite element method is applied to perform fluid-solid coupling experiments to determine the optimal design parameter combination of the L-shaped wave-breaking plate.

[0018] In some embodiments, the design optimization method comprises the following steps:

[0019] Step 1: Parametrically model the L-shaped wave-breaking plate and construct the geometric model of the tank truck superstructure in SolidWorks.

[0020] Step 2: Using the controlled variable method, single-factor experiments were conducted on the design parameters of the L-shaped wave-breaking plate. Numerical simulations were performed to determine the slosh response for each design scheme. The impact of the design parameters on the slosh response was analyzed, and the range of values ​​for each design parameter was initially narrowed.

[0021] Step 3: Use the Plackett-Burman experiment to identify the significant factors that affect the longitudinal impact force of the liquid during the braking process of the tank truck, the force on the first wave-breaking plate, and the rolling moment during the turning process;

[0022] Step 4: using the steepest climbing test to determine the high level, medium level, and low level of the design parameters of the L-shaped wave-breaking plate;

[0023] Step 5: Use the Box-Behnken design to determine the design scheme, and obtain the liquid slosh response of each design scheme through numerical simulation experiments. Using the horizontal axis a and vertical axis b of the manhole on the longitudinal plate of the L-shaped wave-breaking plate, the horizontal offset distance c of the manhole, and the vertical offset distance d of the manhole as independent variables, and the peak longitudinal impact force F1 of the liquid during the braking process of the liquid tank truck, the peak force F2 of the first wave-breaking plate, and the peak rolling moment M of the liquid tank truck during cornering as dependent variables, the least squares method is used to fit a quadratic polynomial to determine the regression equations for the three liquid slosh responses of the L-shaped wave-breaking plate. The basic form of the second-order polynomial regression equations for the three liquid slosh responses of the L-shaped wave-breaking plate is:

[0024]

[0025] Where: is the number of design parameters; is the intercept term; Indicates design parameters The linear effect of Indicates design parameters and The linear interaction effect of Design parameters The quadratic linear effect of represents the random error term;

[0026] Step 6, evaluate the regression equation;

[0027] Calculate the root mean square error (RMSE), coefficient of variation (CV), and coefficient of determination (R-Square) of the regression equations for the three slosh responses of the L-shaped wave-breaking plate. If the RMSE of the regression equation is less than 0.5, the CV is less than 10%, and the R-Square is greater than 95%, proceed to step 7. Otherwise, check the distribution diagram of the predicted and actual values, repeat the numerical simulation experiment for points with large deviations, and return to step 5 until the evaluation indicators of the regression equation meet the above requirements.

[0028] Step 7: The horizontal axis a and the vertical axis b of the manhole on the longitudinal plate of the L-shaped wave-breaking plate, the horizontal offset distance c of the manhole, and the vertical offset distance d of the manhole are used as design variables. The optimization objectives are the peak longitudinal impact force F1 of the liquid during the braking process of the tank truck, the peak force F2 of the first wave-breaking plate, the peak rolling moment M of the tank truck during the turning process, and the mass m of the L-shaped wave-breaking plate. The range of values ​​of the design parameters of the L-shaped wave-breaking plate determined by the steepest climbing test is used as the constraint condition. The mathematical model of the multi-objective optimization problem of the L-shaped wave-breaking plate is:

[0029]

[0030] Step 8: Based on the established mathematical model of the multi-objective optimization problem of the L-shaped wave-breaking plate, the improved non-dominated sorting genetic algorithm NSGA-III with elitist strategy is used to solve the mathematical model and obtain the Pareto optimal solution set;

[0031] Step 9: Conduct fluid-solid coupling experiments on the tank truck under braking and cornering conditions based on the Pareto optimal solution set. Select the design parameter combination in which the overall force of the tank truck body meets the design requirements and all evaluation indicators of the wave-breaking plate reach a relatively low level as the Pareto optimal solution. Output the Pareto optimal solution and use it as the optimal design parameter combination for the L-shaped wave-breaking plate.

[0032] In some embodiments, the regression equations of the three liquid sloshing responses of the L-shaped wave-breaking plate in step 5 are:

[0033] 1) The regression equation for the peak longitudinal impact force F1 of the liquid during braking of a tank truck is:

[0034]

[0035] 2) The regression equation for the peak force F2 of the first wave-breaking plate when the tank truck brakes is:

[0036]

[0037] 3) The regression equation of the peak rolling moment M of the tank truck when turning is:

[0038] .

[0039] Compared with the prior art, the present invention has the following two advantages:

[0040] 1. The present invention provides an L-shaped wave-breaking plate installed inside the tank body of a liquid tank truck. When the liquid tank truck accelerates, decelerates, brakes suddenly, or turns, it can effectively reduce the longitudinal and lateral shaking of the liquid in the tank body, thereby improving the driving stability of the vehicle under various working conditions.

[0041] 2. This invention provides an L-shaped wave-breaking plate design optimization method. Using single-factor experiments, Plackett-Burman experiments, steepest-climb experiments, and Box-Behnken experimental designs, a second-order polynomial regression equation is established, linking the wave-breaking plate design parameters with the liquid sloshing response. The regression equation is evaluated using root mean square error (RMSE), coefficient of variation (CV), and coefficient of determination (R-Square), ensuring accuracy. A modified NSGA-III algorithm is proposed to solve the mathematical model of the multi-objective optimization problem for L-shaped wave-breaking plates. Latin hypercube sampling is combined with reference points for population initialization, ensuring a diverse and uniform distribution of individuals in the initial population. Crossover and mutation distribution indices are introduced to adaptively generate corresponding distribution indices based on the differences between individual populations and ideal points, enabling adaptive crossover and mutation of individual populations. This improves the algorithm's solution efficiency and prevents premature regression into local optimal solutions. Finally, based on the Pareto optimal solution set and fluid-structure interaction experiments, the optimal combination of wave-breaking plate design parameters is determined. This optimization method streamlines the wave-breaking plate design process and enhances design efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Schematic diagram of the structure of the L-shaped wave-breaking plate of the present invention;

[0043] Figure 2 Schematic diagram of the L-shaped wave-breaking plate of the present invention disposed inside the tank of a liquid tank truck;

[0044] Figure 3 yes Figure 2 The main view;

[0045] Figure 4 yes Figure 2 A top view of

[0046] Figure 5 This is a dimensional parameter diagram of the L-shaped wave-breaking board of the present invention (horizontal board);

[0047] Figure 6 This is a dimensional parameter diagram of the L-shaped wave-breaking board of the present invention (longitudinal board);

[0048] Figure 7 yes Figure 6 AA section view in;

[0049] Figure 8 is a flow chart of the design optimization method of the L-shaped wave-breaking plate of the present invention;

[0050] Figure 9 This is a flow chart of the NSGA-III multi-objective optimization algorithm described in the present invention.

[0051] The following are the descriptions of the reference numerals:

[0052] 100, L-shaped wave-breaking plate, 110, horizontal plate, 120, arc plate, 130, longitudinal plate, 140, vent, 150, liquid hole, 160, manhole, 170, wave-breaking plate reinforcement ring, 180, liquid hole reinforcement ring, 190, manhole reinforcement ring;

[0053] 200, tank body; DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0055] See also Figures 1 to 7 As shown, on the one hand, the present invention provides an L-shaped wave-breaking plate 100, which is arranged inside the tank body 200 of the liquid tank truck, and includes a transverse plate 110, an arc plate 120, a longitudinal plate 130, an air vent 140, a liquid hole 150 and a manhole 160. The transverse plate 110 matches the cross-sectional shape of the tank body 200, specifically, the transverse plate includes a first edge segment and a second edge segment, the first edge segment matches a partial edge of the cross section of the tank body 200, one side of the arc plate 120 is connected to the second edge segment of the transverse plate 110, and the longitudinal plate 130 is connected to the other side of the arc plate 120; the air vent 140 and the liquid hole 150 are respectively opened at the upper end and the bottom end of the connection between the transverse plate 110 and the arc plate 120, and the manhole 160 is opened on the longitudinal plate 130.

[0056] When the L-shaped wave-breaking plate involved in the technical solution of the present invention is in use, the horizontal plate 110 is arranged along the horizontal direction of the tank body 200, and the vertical plate 130 is arranged along the longitudinal direction of the tank body 200. The horizontal plate 110 matches the cross-sectional shape of the tank body 200, and the longitudinal length of the vertical plate 130 is the distance between two adjacent L-shaped wave-breaking plates 100. Then, multiple L-shaped wave-breaking plates 100 are arranged inside the tank body 200 of the liquid tank truck. When the liquid tank truck accelerates, decelerates, brakes suddenly or turns, it can effectively slow down the longitudinal and lateral shaking of the liquid in the tank body 200, thereby improving the driving stability of the vehicle under various working conditions.

[0057] In one embodiment, the first edge section of the transverse plate 110 is composed of five arc segments, which are arc segments formed by specific centers and radii. Both the air vents 140 and the liquid vents 150 are circular holes.

[0058] In this embodiment, the radii of the five arcs are R1=2794mm, R2 and R4 are 594mm, R3=1794mm, and R5=2494mm respectively; the center of the air vent 140 is point O1 and the diameter D1 is 120mm, the center of the liquid hole 150 is point O2 and the diameter D2 is 280mm, and the center O1 of the air vent 140, the center O2 of the liquid hole 150 and the longitudinal symmetry center O of the tank body 200 are on the same straight line.

[0059] In one embodiment, Figure 6 As shown, the manhole 160 is an elliptical hole, and the center E1 of the manhole 160 is offset by a certain distance in the horizontal and vertical directions relative to the symmetry center E of the longitudinal plate 130 .

[0060] In this embodiment, the short axis of the manhole 160 is the horizontal axis a, the long axis is the longitudinal axis b, the center E1 of the manhole 160 is offset by a distance c in the horizontal direction relative to the symmetry center E of the longitudinal plate 130, and the center E1 of the manhole 160 is offset by a distance d in the vertical direction relative to the symmetry center E of the longitudinal plate 130.

[0061] In one embodiment, the minor axis length of the manhole 160 is greater than the diameter of the liquid hole 150 , and the diameter of the liquid hole 150 is greater than the diameter of the air hole 140 .

[0062] In one embodiment, a wave-breaking plate reinforcement ring 170 is further included. The wave-breaking plate reinforcement ring 170 is welded to the edges of the transverse plate 110 , the arc plate 120 and the longitudinal plate 130 .

[0063] In one embodiment, a liquid hole reinforcement ring 180 is further included, and the liquid hole reinforcement ring 180 is welded to the circumferential edge of the liquid hole 150 .

[0064] In one embodiment, a manhole reinforcement ring 190 is further included, and the manhole reinforcement ring 190 is welded to the circumferential edge of the manhole 160 .

[0065] A certain type of gasoline tanker truck of Company A is taken as the research object. Figure 2-4 As shown, the L-shaped wave-breaking plates 100 are evenly arranged in the tank body 200. The size parameters of the L-shaped wave-breaking plates 100 are as follows: Figure 5-7As shown, the cross section of the tank body 200 of the tank truck is a circular rectangle, consisting of 8 circular arcs, the horizontal plate 110 is composed of five circular arcs matching it, the thickness of the L-shaped wave-breaking plate 100 is 6 mm, the width W1 of the wave-breaking plate reinforcement ring 170, the manhole reinforcement ring 190 and the liquid hole reinforcement ring 180 are all 140 mm and the thickness is 6 mm, the width W2 of the horizontal plate 110 is 1234 mm, and the length L of the vertical plate 120 is 1224 mm.

[0066] On the other hand, Figure 8 As shown, the present invention provides an L-shaped wave-breaking plate design optimization method, which is used to design the L-shaped wave-breaking plate 100 provided in the first aspect of the present invention. The finite element method and the response surface method are used to construct an approximate response surface model of the design variables of the L-shaped wave-breaking plate and the liquid sloshing response of the liquid tank truck, and a mathematical model of the multi-objective optimization problem of the L-shaped wave-breaking plate is constructed. The improved non-dominated sorting genetic algorithm NSGA-III with an elite strategy is used to solve the mathematical model to obtain the Pareto optimal solution set. Finally, the finite element method is used to perform a fluid-solid coupling experiment to determine the optimal design parameter combination of the L-shaped wave-breaking plate.

[0067] Specifically, the design optimization method of the L-shaped wave-breaking plate 100 includes the following steps:

[0068] Step 1: Parametrically model the L-shaped wave-breaking plate. Complete the parametric modeling of the L-shaped wave-breaking plate and construct the geometric model of the tank truck superstructure in SolidWorks.

[0069] Step 2: Using the controlled variable method, single-factor experiments were conducted on the design parameters of the L-shaped wave-breaking plate. Numerical simulations were performed to determine the slosh response for each design scheme. The impact of the design parameters on the slosh response was analyzed, and the range of values ​​for each design parameter was initially narrowed.

[0070] The specific process of the numerical simulation experiment is as follows: first, complete the mechanical modeling of the wave-breaking plate and the tank body; then import the model into the mesh processing module of Fluent to complete the meshing of the liquid tank truck; finally, complete the liquid sloshing experiment of the liquid tank truck during braking and cornering in Fluent, and output the peak longitudinal impact force F1 of the liquid during braking, the peak force F2 of the first wave-breaking plate, and the peak roll moment M of the liquid tank truck during cornering.

[0071] Step 3: Use the Plackett-Burman experiment to identify the significant factors that affect the longitudinal impact force of the liquid during the braking process of the tank truck, the force on the first wave-breaking plate, and the rolling moment during the turning process;

[0072] According to the Plackett-Burman experiment, the significant factors affecting the longitudinal impact force during the braking process of a tank truck are: the transverse axis a and the longitudinal axis b of the elliptical manhole on the longitudinal plate of the L-shaped wave-breaking plate, the horizontal offset distance c of the manhole, and the vertical offset distance d of the manhole; the significant factors affecting the force on the first wave-breaking plate during the braking process of a tank truck are: the transverse axis a and the longitudinal axis b of the elliptical manhole on the longitudinal plate of the L-shaped wave-breaking plate, and the vertical offset distance d of the manhole; the significant factors affecting the roll moment during the turning process of a tank truck are: the transverse axis a and the longitudinal axis b of the elliptical manhole on the longitudinal plate of the L-shaped wave-breaking plate, and the vertical offset distance d of the manhole.

[0073] Step 4: The steepest climbing test is used to determine the high, medium, and low levels of the design parameters of the L-shaped wave-breaking plate. According to the standard GB18564.1-2019 and the steepest climbing test analysis, the optimal value ranges of the optimization design variables of the L-shaped wave-breaking plate are: a∈[500,800], b∈[400,800], c∈[0,300], and d∈[0,390], all in mm.

[0074] Step 5: Use the Box-Behnken design to determine the design scheme, and obtain the liquid slosh response of each design scheme through numerical simulation experiments. Using the horizontal axis a and vertical axis b of the manhole on the longitudinal plate of the L-shaped wave-breaking plate, the horizontal offset distance c of the manhole, and the vertical offset distance d of the manhole as independent variables, and the peak longitudinal impact force F1 of the liquid during the braking process of the liquid tank truck, the peak force F2 of the first wave-breaking plate, and the peak rolling moment M of the liquid tank truck during cornering as dependent variables, the least squares method is used to fit a quadratic polynomial to determine the regression equations for the three liquid slosh responses of the L-shaped wave-breaking plate. The basic form of the second-order polynomial regression equations for the three liquid slosh responses of the L-shaped wave-breaking plate is:

[0075]

[0076] Where: is the number of design parameters; is the intercept term; Indicates design parameters The linear effect of Indicates design parameters and The linear interaction effect of Design parameters The quadratic linear effect of represents the random error term.

[0077] Taking a gasoline tanker truck of Company A as an example, according to the standards GB 7258-2017 and GB 28373-2012, the peak accelerations of the tanker truck during normal driving are 6m / s for braking and turning respectively. 2, 4.9m / s 2 The regression equations of the three liquid sloshing responses of the L-shaped wave-breaking plate and the mass equation of the L-shaped wave-breaking plate are as follows:

[0078] (1) The regression equation for the peak longitudinal impact force F1 of the liquid during braking of a tank truck is:

[0079]

[0080] (2) The regression equation of the peak force F2 of the first wave-breaking plate when the tank truck brakes is:

[0081]

[0082] (3) The regression equation of the peak value of the rolling moment M when the tank truck turns is:

[0083]

[0084] (4) The mass equation of the L-shaped wave-breaking plate is as follows:

[0085]

[0086] Step 6, evaluate the regression equation;

[0087] Calculate the root mean square error (RMSE), coefficient of variation (CV), and coefficient of determination (R-Square) of the regression equations for the three slosh responses of the L-shaped wave-breaking plate. If the RMSE of the regression equation is less than 0.5, the CV is less than 10%, and the R-Square is greater than 95%, proceed to step 7. Otherwise, check the distribution diagram of the predicted and actual values, repeat the numerical simulation experiment for points with large deviations, and return to step 5 until the evaluation indicators of the regression equation meet the above requirements.

[0088]

[0089]

[0090]

[0091] Where, is the predicted value of the regression model for the i-th sample point; is the actual value of the i-th sample point; is the mean of the sample data.

[0092] Step 7: The horizontal axis a and the vertical axis b of the manhole on the longitudinal plate of the L-shaped wave-breaking plate, the horizontal offset distance c of the manhole, and the vertical offset distance d of the manhole are used as design variables. The optimization objectives are the peak longitudinal impact force F1 of the liquid during the braking process of the tank truck, the peak force F2 of the first wave-breaking plate, the peak rolling moment M of the tank truck during the turning process, and the mass m of the L-shaped wave-breaking plate. The range of values ​​of the design parameters of the L-shaped wave-breaking plate determined by the steepest climbing test is used as the constraint condition. The mathematical model of the multi-objective optimization problem of the L-shaped wave-breaking plate is:

[0093]

[0094] Step 8: Based on the established mathematical model of the multi-objective optimization problem of the L-shaped wave-breaking plate, the improved non-dominated sorting genetic algorithm NSGA-III with elitist strategy is used to solve the mathematical model and obtain the Pareto optimal solution set;

[0095] Step 9: Conduct fluid-solid coupling experiments on the tank truck under braking and cornering conditions based on the Pareto optimal solution set. Select the design parameter combination in which the overall force of the tank truck body meets the design requirements and all evaluation indicators of the wave-breaking plate reach a relatively low level as the Pareto optimal solution. Output the Pareto optimal solution and use it as the optimal design parameter combination for the L-shaped wave-breaking plate.

[0096] Specifically, the present invention also proposes an improved NSGA-III multi-objective optimization algorithm.

[0097] To improve the efficiency of the NSGA-III algorithm and avoid falling into local optimality, the present invention proposes a population initialization and adaptive crossover and mutation strategy based on reference points and Latin hypercube sampling. Latin hypercube sampling generates individuals associated with each reference point, ensuring the randomness and diversity of the initial population while improving the uniformity of the spatial distribution of the initial population individuals. By using a simulated binary crossover operator and polynomial mutation, the corresponding distribution index is introduced. Different distribution indices are adaptively generated based on the distribution of individuals in the population, and crossover and mutation are performed. The closer an individual is to the ideal point, the closer its offspring, produced by crossover and mutation, are to the parent generation. When the population is too dense or the individuals are further away from the ideal point, a smaller distribution index is generated to expand the search range of the population.

[0098] The improved NSGA-Ⅲ algorithm flow chart is as follows: Figure 9As shown in the figure, its content is mainly divided into five parts: the first part (step 1-step 2) is to initialize the population based on the reference point and Latin hypercube sampling; the second part (step 3-step 4) is the adaptive normalization of the population individuals, non-dominated sorting and calculation of the number of niches of the reference point; the third part (step 5) is the tournament selection of the population individuals and adaptive crossover and mutation; the fourth part (step 6-step 7) is to adopt the elite strategy to select excellent individuals to enter the next generation; the fifth part (step 8) is to output the Pareto optimal solution set based on the Pareto frontier. The specific steps of the algorithm are as follows:

[0099] Step 1: Calculate the hyperplane reference point between the design variables and the objective function

[0100] The hyperplane reference point is calculated as follows:

[0101] (1) Construct a (M-1)×1 two-dimensional array X.

[0102]

[0103] Where M is the number of research subjects; H is the number of segments for each research subject.

[0104] (2) Generated by array X The two-dimensional array Y, each element in Y (j=1,2,…,M-1), and must satisfy .

[0105] (3) Update the elements in array Y.

[0106]

[0107] Where, .

[0108] (4) Obtain reference points on each research object to form hyperplane reference points.

[0109]

[0110] Step 2: Population initialization

[0111] (1) Generate the initial optimal solution of the multi-objective optimization problem: Generate a set of experimental populations based on the orthogonal Latin method, and calculate the factors in the population m Level The sum of experimental results K i m ,according to K i m Determining factors mThe optimal level of each factor is determined, and the optimal level of each factor is used to form the initial optimal individual and put it into the initial population.

[0112] (2) Population initialization based on reference points and Latin hypercube sampling

[0113] Associate the individuals generated by Latin hypercube sampling with the hyperplane reference points of the design variables. For each reference point, only the individual closest to it is selected and added to the initial population. Repeat the above steps until all reference points have corresponding individuals, completing the population initialization operation.

[0114] Step 3: Calculate the individual objective function values ​​of the initial population, adaptively normalize the individuals, and perform non-dominated sorting on the individuals.

[0115] The population individual adaptive normalization process is as follows:

[0116] (1) Calculate the minimum value of each objective function in the population individual and use it as the ideal point .

[0117]

[0118] (2) The objective function value of individuals in the mobile population makes the ideal point Become a zero vector, and the objective function value after moving is expressed as .

[0119]

[0120] (3) Calculate the extreme points on each target axis

[0121]

[0122] In the formula, the weight vector The axis direction.

[0123] (4) Normalization of the objective function of individuals in the population.

[0124]

[0125] Where, a is the hyperplane intercept.

[0126] Step 4: Update the number of niches of the reference point, calculate the distance between the population individual and the reference point of the objective function hyperplane, and associate the individual with the nearest reference point.

[0127] Step 5: Randomly select two individuals through the tournament selection method, compare their objective function values, the number of niches associated with the reference point, and the distance from the reference point, and select the better individual for adaptive crossover and mutation to generate a new generation of subpopulations.

[0128] The calculation method for the adaptive crossover and mutation of parent individuals to generate offspring individuals is as follows:

[0129] Simulate a binary crossover:

[0130]

[0131]

[0132]

[0133] Where, and is the parent individual; and For offspring individuals; is a random number, ; is the cross-distribution index; and is a constant term; is the mean point of the population objective function value and ideal point distance; The individual whose objective function value is farther from the ideal point than the other two individuals in the intersection Distance from the ideal point.

[0134] Polynomial mutation:

[0135]

[0136]

[0137]

[0138]

[0139] Where, and are the parent generation and the mutated offspring individuals respectively; and are the upper and lower limits of the X value range, is the variation distribution index. and is a constant term; is the objective function value of the mutant individual.

[0140] Step 6: Combine the parent population with the offspring population resulting from crossover and mutation to form a new population. Perform steps 3 and 4 on the individuals in the population. Select high-quality individuals with high non-dominated ranks to join the parent population. If the number of individuals in the parent population exceeds the set number of individuals in the population, screen the last group of individuals in the parent population and select the individual closest to the reference point with the fewest niches to join the parent population.

[0141] Step 7: Repeat steps 5 and 6 until the number of iterations reaches the maximum value.

[0142] Step 8: Based on the Pareto frontier, the output is the Pareto optimal solution set.

[0143] In summary, in the embodiment of the present invention, taking a gasoline tank truck of Company A as an example, the optimal design parameter combination of the L-shaped wave-breaking plate is that the lengths of the horizontal axis a and the vertical axis b of the manhole 160 are 600 mm and 700 mm respectively, the horizontal offset distance c of the manhole 160 is 300 mm, and the vertical offset distance d of the manhole 160 is 160 mm; after installing the optimized L-shaped wave-breaking plate, when the filling ratio of the tank truck is 0.6, the peak accelerations of braking and turning are 6 m / s respectively. 2 , 4.9m / s 2 The peak longitudinal impact force of the liquid under braking conditions and the peak roll moment during cornering decreased by 40.08% and 15.56% respectively compared to the case without wave-breaking plates. The roll moment reduction rate increased by 15.14% compared to the original vehicle's corrugated wave-breaking plates. The weight of the optimized L-shaped wave-breaking plates is 81.39kg. The total mass of the wave-breaking plates of the tank truck equipped with the L-shaped wave-breaking plates is reduced by 36.93% compared to the case when both transverse and longitudinal wave-breaking plates are installed. Therefore, the optimized L-shaped wave-breaking plates effectively reduce the transverse and longitudinal swaying of the liquid in the tank truck under emergency conditions, improve the vehicle's driving stability under various working conditions, and make the entire vehicle lighter.

[0144] Furthermore, the proposed L-shaped wave-breaking plate optimization design method establishes a second-order polynomial regression equation linking wave-breaking plate design parameters with liquid sloshing response through single-factor experiments, Plackett-Burman experiments, steepest-climb experiments, and Box-Behnken experimental designs. The regression equation is evaluated using root mean square error (RMSE), coefficient of variation (CV), and coefficient of determination (R-Square) to ensure accuracy. A modified NSGA-III algorithm is proposed to solve the multi-objective optimization problem for L-shaped wave-breaking plates. Latin hypercube sampling is combined with reference points for population initialization, ensuring a diverse and uniform distribution of individuals in the initial population. Crossover and mutation distribution indices are introduced to adaptively generate corresponding distribution indices based on the differences between individuals in the population and the ideal point, enabling adaptive crossover and mutation of individuals in the population. This improves the algorithm's solution efficiency and prevents premature regression into local optimal solutions. Finally, based on the Pareto optimal solution set and fluid-structure interaction experiments, the optimal combination of design parameters for the wave-breaking plate is determined. This optimization design method streamlines the wave-breaking plate design process and enhances design efficiency.

[0145] The specific embodiments of the present invention described above do not limit the scope of protection of the present invention. Any other corresponding changes and modifications made based on the technical concept of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing the design of an L-shaped wave-breaking plate, characterized in that: The method is used to design an L-shaped wave-breaking plate, which is arranged inside the tank body of a liquid tank truck, and includes: a transverse plate, the transverse plate having a first edge section and a second edge section, the first edge section matching a portion of an edge of a cross section of the tank body; a circular arc plate, one side of which is connected to the second edge section of the transverse plate; a longitudinal plate connected to the other side of the arc plate; A vent hole, a liquid hole and a manhole, wherein the vent hole and the liquid hole are respectively provided at the upper end and the bottom end of the connection between the horizontal plate and the arc plate, and the manhole is provided on the vertical plate; The air vent and the liquid hole are both circular holes, the manhole is an elliptical hole, and the center of the manhole is offset in the horizontal and vertical directions relative to the symmetry center of the longitudinal plate; The design optimization method uses the finite element method and response surface methodology to construct a response surface approximation model of the design variables of the L-shaped wave-breaking plate and the liquid sloshing response of the tank truck, and then constructs a mathematical model for the multi-objective optimization problem of the L-shaped wave-breaking plate. The improved non-dominated sorting genetic algorithm (NSGA-III) with an elitist strategy is used to solve the mathematical model to obtain the Pareto optimal solution set. Finally, the finite element method is used to conduct fluid-structure coupling experiments to determine the optimal design parameter combination of the L-shaped wave-breaking plate. The method includes the following steps: Step 1: Parametrically model the L-shaped wave-breaking plate and construct the geometric model of the tank truck superstructure in SolidWorks. Step 2: Using the controlled variable method, single-factor experiments were conducted on the design parameters of the L-shaped wave-breaking plate. Numerical simulations were performed to determine the slosh response for each design scheme. The impact of the design parameters on the slosh response was analyzed, and the range of values ​​for each design parameter was initially narrowed. Step 3: Use the Plackett-Burman experiment to identify the significant factors that affect the longitudinal impact force of the liquid during the braking process of the tank truck, the force on the first wave-breaking plate, and the rolling moment during the turning process; Step 4: using the steepest climbing test to determine the high level, medium level, and low level of the design parameters of the L-shaped wave-breaking plate; Step 5: Use the Box-Behnken design to determine the design scheme, and obtain the liquid slosh response of each design scheme through numerical simulation experiments. Using the horizontal axis a and vertical axis b of the manhole on the longitudinal plate of the L-shaped wave-breaking plate, the horizontal offset distance c of the manhole, and the vertical offset distance d of the manhole as independent variables, and the peak longitudinal impact force F1 of the liquid during the braking process of the liquid tank truck, the peak force F2 of the first wave-breaking plate, and the peak rolling moment M of the liquid tank truck during cornering as dependent variables, the least squares method is used to fit a quadratic polynomial to determine the regression equations for the three liquid slosh responses of the L-shaped wave-breaking plate. The basic form of the second-order polynomial regression equations for the three liquid slosh responses of the L-shaped wave-breaking plate is: Where: is the number of design parameters; is the intercept term; Indicates design parameters The linear effect of Indicates design parameters and The linear interaction effect of Design parameters The quadratic linear effect of represents the random error term; Step 6, evaluate the regression equation; Calculate the root mean square error (RMSE), coefficient of variation (CV), and coefficient of determination (R-Square) of the regression equations for the three slosh responses of the L-shaped wave-breaking plate. If the RMSE of the regression equation is less than 0.5, the CV is less than 10%, and the R-Square is greater than 95%, proceed to step 7. Otherwise, check the distribution diagram of the predicted and actual values, repeat the numerical simulation experiment for points with large deviations, and return to step 5 until the evaluation indicators of the regression equation meet the above requirements. Step 7: The horizontal axis a and the vertical axis b of the manhole on the longitudinal plate of the L-shaped wave-breaking plate, the horizontal offset distance c of the manhole, and the vertical offset distance d of the manhole are used as design variables. The optimization objectives are the peak longitudinal impact force F1 of the liquid during the braking process of the tank truck, the peak force F2 of the first wave-breaking plate, the peak rolling moment M of the tank truck during the turning process, and the mass m of the L-shaped wave-breaking plate. The range of values ​​of the design parameters of the L-shaped wave-breaking plate determined by the steepest climbing test is used as the constraint condition. The mathematical model of the multi-objective optimization problem of the L-shaped wave-breaking plate is: Step 8: Based on the established mathematical model of the multi-objective optimization problem of the L-shaped wave-breaking plate, the improved non-dominated sorting genetic algorithm NSGA-III with elitist strategy is used to solve the mathematical model and obtain the Pareto optimal solution set; Step 9: Conduct fluid-solid coupling experiments on the tank truck under braking and cornering conditions based on the Pareto optimal solution set. Select the design parameter combination in which the overall force of the tank truck body meets the design requirements and all evaluation indicators of the wave-breaking plate reach a relatively low level as the Pareto optimal solution. Output the Pareto optimal solution and use it as the optimal design parameter combination for the L-shaped wave-breaking plate.

2. The L-shaped wave-breaking plate design optimization method according to claim 1, characterized in that: The regression equations for the three liquid sloshing responses of the L-shaped wave-breaking plate in step 5 are: 1) The regression equation for the peak longitudinal impact force F1 of the liquid during braking of a tank truck is: 2) The regression equation for the peak force F2 of the first wave-breaking plate when the tank truck brakes is: 3) The regression equation of the peak rolling moment M of the tank truck when turning is: 。 3. The L-shaped wave-breaking plate design optimization method according to claim 1, characterized in that: The first edge section of the transverse plate is formed by connecting five circular arcs end to end in sequence.

4. The L-shaped wave-breaking plate design optimization method according to claim 3, characterized in that: The short axis length of the manhole is greater than the diameter of the liquid through hole, and the diameter of the liquid through hole is greater than the diameter of the air vent.

5. The L-shaped wave-breaking plate design optimization method according to claim 1, characterized in that: It also includes a wave-breaking plate reinforcement ring, which is fixed to the edges of the transverse plate, the arc plate and the longitudinal plate.

6. The L-shaped wave-breaking plate design optimization method according to claim 1, characterized in that: It also includes a liquid hole reinforcement ring, which is fixed to the circumferential edge of the liquid hole.

7. The L-shaped wave-breaking plate design optimization method according to claim 1, characterized in that: It also includes a manhole reinforcement ring, which is fixed to the circumferential edge of the manhole.

Images