A train roof insulator and its optimization design method
By designing alternately distributed umbrella skirt structure and wavy mandrel train roof insulators, the problems of high-speed train roof insulators to air flow disturbance and resistance are solved, the flow field performance is optimized and the resistance is reduced, and the train is improved acceleration performance and energy efficiency.
Patent Information
- Application Number
- CN202411274166.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-12
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-09-12
AI Technical Summary
The existing technology has failed to effectively solve the disturbance and resistance of roof insulators on air flow when high-speed trains are running at high speed. Traditional designs mainly focus on material selection and surface smoothness, ignoring the impact of shape on fluid dynamics, resulting in the inability to go from Fundamentally reduces the resistance of roof insulators to air flow.
A train roof insulator including mandrel and umbrella skirt structure was designed. The umbrella skirt structure was alternately distributed, and the outer surface of the mandrel showed periodic wave shape. The wavy diameter changes followed the definition of cosine function, and the Krikin agent model and Various group genetic algorithms optimize parameters to achieve fluid dynamics simulation verification.
By optimizing the insulator structure, the resistance of the roof insulator is significantly reduced, the flow field performance is improved, the calculation process is simplified, the analysis efficiency is greatly improved, and the acceleration performance and energy utilization efficiency of the train are improved.
Smart Images

Figure CN118824649B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of high-speed trains, and in particular to a train roof insulator and an optimization design method thereof. Background Art
[0002] With the continuous growth of my country's economic strength, my country has built the world's longest operating mileage and the largest passenger volume high-speed railway network. Faster and lower energy-consuming high-speed trains have become the main trend of my country's train development. Studies have shown that the air resistance of the train is proportional to the square of the train speed. Therefore, as the train speed increases, the impact of air resistance becomes more and more significant. This resistance not only leads to an increase in train energy consumption, but also limits the increase in train speed. The existing technology mainly relies on the streamlined design of the car body to reduce resistance. For example, the front resistance of the train is reduced by improving the shape of the front of the train, reducing the height of the car body, and optimizing the smoothness of the car body. Although these methods reduce the air resistance of the train to a certain extent, they fail to effectively solve the problem of flow field destruction caused by roof equipment such as insulators when the train is running at high speed. Especially with the continuous increase in the speed of high-speed railways, the roof insulators are causing more and more serious damage to the overall flow field environment of the train. Traditional technology, when designing roof insulators, mainly focuses on material selection and improvement of surface smoothness, and often ignores the influence of its shape on fluid dynamics. This method has limited effect and cannot fundamentally reduce the disturbance and resistance contribution of roof insulators to air flow.
[0003] In view of the above-mentioned shortcomings of the prior art, there is an urgent need for a train roof insulator and an optimized design method thereof. Summary of the invention
[0004] The purpose of the present invention is to provide a train roof insulator and an optimization design method thereof to improve the above problems. In order to achieve the above purpose, the technical solution adopted by the present invention is as follows:
[0005] In a first aspect, the present application provides a train roof insulator, comprising a core rod and at least two shed structures, all of the shed structures are fixedly arranged on the core rod, the shed structures are disc-shaped, the shed structures include two types, large sheds and small sheds, the diameter of the large shed is greater than the diameter of the small shed, the number of each type of shed is at least one, the large shed and the small shed are alternately distributed on the core rod, all of the shed structures are arranged along the axis direction of the core rod, adjacent shed structures are parallel to each other, the plane of each of the shed structures is perpendicular to the axis of the core rod, and the interval between the shed structures is a fixed value.
[0006] Furthermore, the core rod comprises at least one wave portion, the wave portion is arranged between adjacent shed structures, and the outer surface of the wave portion is periodically wavy along the length direction of the core rod.
[0007] Furthermore, the diameter change of the wave portion follows the cosine function definition, and the specific formula is:
[0008] ;
[0009] in, is the diameter of the corrugated mandrel, is the diameter of the original cylindrical mandrel, is the amplitude of the cosine function that constitutes the wavy core rod, is the wavelength of the cosine function, Indicates the lengthwise position of each wave portion.
[0010] Furthermore, the amplitude range of the cosine function is 0 mm to 2 mm.
[0011] Furthermore, the wavelength of the cosine function is 5.6 mm to 28 mm.
[0012] In a second aspect, the present application provides a method for optimizing the design of a train roof insulator, comprising:
[0013] The original structural data of the train roof insulator is acquired for modeling to obtain an original three-dimensional model, wherein the structural data includes the size parameters of the original cylindrical core rod;
[0014] The structural data is processed based on a preset cosine function, by improving the cylindrical core rod to a wavy profile that satisfies the cosine function definition, and obtaining a key parameter range through parameterization, wherein the key parameter range includes a parameter range of an amplitude and a wavelength of the cosine function;
[0015] Perform sampling processing according to the key parameters, and extract sample points from the parameter space by using a Latin hypercube sampling method to obtain a sample point data set;
[0016] Performing model building processing according to the sample point data set to obtain a Kriging proxy model;
[0017] Performing prediction processing according to the Kriging proxy model, predicting the value of the unsampled point by a spatial interpolation method to obtain a predicted value, wherein the predicted value includes at least two sets of parameter values of amplitude and wavelength;
[0018] The predicted value is input into a preset multi-population genetic algorithm, and a final optimized parameter value is obtained through iterative optimization processing;
[0019] A simulation process is performed according to the optimized parameter values and the original three-dimensional model, and a simulation result is obtained by applying the optimized parameter values to the three-dimensional model for fluid dynamics simulation verification.
[0020] The beneficial effects of the present invention are:
[0021] The present invention improves the core rod into a wavy cylinder composed of a cosine function, obtains the optimal structural parameters through algorithm optimization, optimizes the flow field performance of the insulator, and achieves the drag reduction effect of the roof insulator. At the same time, a proxy model is used to construct the relationship between the structural parameters of the roof insulator and the resistance, which simplifies the calculation cycle and greatly improves the analysis efficiency.
[0022] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or be understood by implementing the embodiments of the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.
[0024] Figure 1 It is a schematic structural diagram of a train roof insulator according to an embodiment of the present invention;
[0025] Figure 2 It is a flow chart of the train roof insulator optimization design method described in an embodiment of the present invention;
[0026] Figure 3 It is a schematic diagram of pressure distribution of the improved roof insulator described in an embodiment of the present invention;
[0027] Figure 4 It is a schematic diagram of pressure distribution of the original roof insulator described in an embodiment of the present invention;
[0028] Figure 5 A comparison diagram of the surface pressure on the leeward side of the train roof insulator described in an embodiment of the present invention;
[0029] Figure 6 It is a schematic diagram of the velocity field of the improved roof insulator described in an embodiment of the present invention;
[0030] Figure 7 It is a schematic diagram of the velocity field of the original roof insulator described in an embodiment of the present invention;
[0031] Figure 8 It is a schematic diagram of flow field streamline distribution of the improved roof insulator described in an embodiment of the present invention;
[0032] Fig. 9 Schematic diagram of the flow field streamline distribution of the original roof insulator described in an embodiment of the present invention.
[0033] Markings in the figure: 1, core rod; 11, wave part; 2, umbrella skirt structure; 21, large umbrella skirt; 22, small umbrella skirt. DETAILED DESCRIPTION
[0034] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0035] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.
[0036] Embodiment 1:
[0037] This embodiment provides a train roof insulator, such as Figure 1As shown, it includes a core rod 1 and at least two shed structures 2, all shed structures 2 are fixedly arranged on the core rod 1, the shed structure 2 is disc-shaped, and the shed structure 2 includes two types: a large shed 21 and a small shed 22. The diameter of the large shed 21 is greater than the diameter of the small shed 22. The number of sheds of each type is at least one. The large shed 21 and the small shed 22 are alternately distributed on the core rod 1. All shed structures 2 are arranged along the axis of the core rod 1. Adjacent shed structures 2 are parallel to each other. The plane of each shed structure 2 is perpendicular to the axis of the core rod 1, forming a stable and symmetrical arrangement pattern. The interval between the shed structures 2 is a fixed value. This fixed interval can ensure the uniformity of air flow during high-speed operation, thereby reducing vibration and noise caused by irregular airflow. In the design of high-speed trains, resistance is a major limiting factor, and the resistance of the roof insulator has a significant impact on the overall performance of the train. By optimizing the insulator design, the acceleration performance and energy utilization efficiency of the train can be significantly improved.
[0038] Furthermore, the core rod 1 includes at least one wave portion 11, which is arranged between adjacent shed structures 2, and the outer surface of the wave portion 11 is periodically wavy along the length direction of the core rod 1. This distribution can better optimize the airflow path, guide the airflow between the sheds, and make it bypass the surface of the insulator more smoothly. When the high-speed airflow flows through the wavy cylindrical core rod 1, in addition to flowing backward to form a wake vortex, a part of the airflow will flow to both ends of the core rod 1, forming an airflow flow in the direction of the core rod 1, and the flow of the airflow in the direction of the core rod 1 will inhibit the process of the airflow generating a wake vortex behind the core rod 1, thereby delaying the separation of the boundary layer, inhibiting the periodic shedding vortex, and improving the flow field on the leeward side of the insulator, so that the pressure on the leeward side of the insulator increases, thereby reducing the pressure difference resistance of the insulator, and realizing the drag reduction of the roof insulator. At the same time, the wavy core rod 1 can increase the flow velocity on the leeward side of the insulator, strengthen the arc blowing effect of the airflow on the arc, increase the flashover path, and increase the flashover voltage.
[0039] Furthermore, the diameter change of the wave part follows the definition of the cosine function. This mathematical model provides the core rod with unique geometric shape and fluid mechanics properties. The specific formula is:
[0040] ;
[0041] in, is the diameter of the corrugated mandrel, is the diameter of the original cylindrical mandrel, is the amplitude of the cosine function that constitutes the wavy core rod, is the wavelength of the cosine function, Indicates the lengthwise position of each wave section.
[0042] Furthermore, the amplitude range of the cosine function is 0 mm to 2 mm.
[0043] Furthermore, the wavelength of the cosine function is 5.6 mm to 28 mm.
[0044] Embodiment 2:
[0045] This embodiment provides a method for optimizing the design of train roof insulators.
[0046] See also Figure 2 , the figure shows that the method includes steps S100 to S700.
[0047] Step S100, obtaining the structural data of the original train roof insulator and performing modeling processing to obtain an original three-dimensional model, wherein the structural data includes the size parameters of the original cylindrical core rod;
[0048] It is understandable that the structural data includes the design size parameters of the original cylindrical core rod, such as diameter, length, diameter and thickness of the shed skirts, the number of shed skirts and their positions on the core rod, etc. These parameters are the basic geometric and physical characteristics initially determined during the engineering design process to ensure that the 3D model can accurately reflect the performance of the insulator in actual application. A comprehensive understanding and accurate modeling of these parameters can provide an accurate basis for subsequent optimization analysis.
[0049] Step S200, processing the structural data based on a preset cosine function, improving the cylindrical mandrel into a wavy profile that satisfies the cosine function, and obtaining a key parameter range through parameterization, wherein the key parameter range includes a parameter range of an amplitude and a wavelength of the cosine function;
[0050] It should be noted that step S200 includes steps S210 to S230.
[0051] Step S210, processing the structural data, calculating the diameter change of the core rod at each axial position by applying a preset cosine function formula, and obtaining a preliminary contour model of the wavy core rod;
[0052] It is understandable that this step is intended to make the surface of the core rod present a periodic wavy profile, changing the surface morphology of the core rod in order to produce different effects on the airflow. When the airflow encounters the wavy structure, it will produce a flow tendency to move laterally (spanwise) along the core rod. This spanwise flow suppresses the tendency of the airflow to form a large-scale wake vortex directly behind. Since part of the airflow flows to both ends, the formation of the wake vortex is suppressed and the separation of the boundary layer is delayed. This means that the fluid is not easy to separate quickly behind the core rod, reducing the chance of forming a large vortex, thereby reducing the resistance on the leeward side of the cylinder.
[0053] In addition to the aerodynamic improvements, the wavy core rod design can also improve its electrical performance. Specifically, the wavy structure increases the blowing effect of the airflow on the arc. The faster airflow speed not only helps to reduce aerodynamic drag, but also enhances the anti-flashover ability of the insulator. The blowing effect refers to the accelerated airflow passing around the insulator, which will increase the air flow's ability to suppress the arc and increase the complexity of the arc path. Since the airflow has a stronger sweeping effect on the arc, the flashover path becomes longer, thereby increasing the flashover voltage. This means that under the same conditions, the insulator can withstand higher voltages without flashover. Increasing the flashover voltage is crucial to the safety and stability of high-speed trains, especially in severe weather conditions (such as thunderstorms or high humidity environments). Such a design can effectively reduce the probability of electrical accidents.
[0054] Step S220, performing parameterization processing according to the preliminary contour model, adjusting the wavelength and amplitude based on preset design requirements to obtain a parameterized model;
[0055] It should be noted that the core of parametric processing is to analyze the differences in aerodynamic performance of core rods of different shapes by changing the wavelength and amplitude. During the parametric processing, the design requirements, such as the target speed range, material properties, and manufacturing process limitations, must be determined first. Next, parametric modeling is performed using professional engineering design tools. This modeling process can generate different versions of the core rod model by adjusting the parameters in the formula to ensure that it exhibits excellent fluid dynamic characteristics in all situations. Through parametric processing, the effects of different parameter combinations on aerodynamic performance can be effectively explored.
[0056] Step S230: performing data extraction processing according to the parameterized model, and obtaining the key parameter range by calculating the variation range of wavelength and amplitude.
[0057] Preferably, statistical analysis methods such as response surface methodology or experimental design are used to help identify the sensitivity of parameters to performance, so as to obtain the optimal parameter range. In the process of extracting the key parameter range, manufacturing and practical limitations must also be considered. For example, too large wavelength and amplitude will affect the structural integrity and manufacturing difficulty of the core rod, while too small may not achieve the expected drag reduction effect. Through a comprehensive analysis of these factors, a reasonable parameter range is determined to ensure that the core rod can achieve the optimization effect in both design and practical application.
[0058] Step S300, performing sampling processing according to key parameters, and obtaining a sample point data set by extracting sample points from the parameter space using a Latin hypercube sampling method;
[0059] It should be noted that step S300 includes steps S310 to S340.
[0060] Step S310, performing grid division processing according to the key parameter range, creating a parameter grid in the parameter space based on preset rules to obtain an initial parameter grid;
[0061] It is understandable that the parameter space needs to be meshed first according to the key parameter range defined above. The key parameters include important geometric parameters such as the amplitude of the cosine function and the wavelength of the wavy core rod. The parameter space represents all possible parameter combinations, and meshing is to discretize it into a multidimensional grid so that each grid cell represents a possible parameter configuration. It is carried out through preset rules to ensure that the range of variation of each parameter is reasonably covered. Through this meshing, an initial parameter grid can be created. In this process, the creation of the initial parameter grid is similar to building a framework, in which each grid point represents a potential design solution or experimental condition. The goal of this step is to ensure that the initial grid can completely cover the parameter space and provide a basis for the screening of subsequent sample points. This process also provides a visual structure for multi-parameter combinations, which is convenient for subsequent analysis and processing.
[0062] Step S320, performing screening processing according to the initial parameter grid, obtaining at least one set of initial sample points by calculating representative results of each grid point and screening representative reference points;
[0063] It is understood that each grid point is evaluated based on the design objectives (such as aerodynamic drag of the insulator), and representative results involve calculating the design performance indicators (such as hydrodynamic drag, insulation performance, etc.) for each grid point.
[0064] Step S330, performing optimization processing according to the initial sample points, dividing the space in each parameter dimension by applying the Latin hypercube sampling method and randomly selecting the positions of the sample points to obtain optimized sample points;
[0065] It can be understood that hypercube sampling is a random sampling technique used to generate a sample set covering the entire multidimensional parameter space. The basic principle is to divide the range of each parameter into uniform subintervals and randomly select a point in each subinterval. These random points are then combined to form a set of the entire sample point.
[0066] Step S340: construct a data set based on the optimized sample points, perform simulation calculations on the sample points and the corresponding models, obtain the resistance value of the insulator formed by each group of sample points, and form a sample point data set.
[0067] In this process, computational fluid dynamics or other related simulation tools are used to simulate the design of each sample point to obtain the resistance value of the insulator at each group of sample points. The simulation results of all sample points are summarized to form a sample point data set that reflects the characteristics of the design space. This data set contains the performance indicators at each sample point. This data set not only provides the necessary basic data for subsequent optimization, but also helps identify the trend of the impact of parameters on performance and the optimal design solution by analyzing the data set.
[0068] Step S400, performing model building processing according to the sample point data set to obtain a Kriging proxy model;
[0069] It should be noted that step S400 includes steps S410 to S430.
[0070] Step S410: Perform model parameter screening processing according to the sample point data set, determine the number of key parameters that affect the insulator resistance, and select second-order polynomial regression parameters and Gaussian correlation function to construct a preliminary proxy model;
[0071] Step S420: Perform model progress verification processing according to the preliminary proxy model, compare the predicted results of the test sample points with the actual results, and use the generalized least squares estimation and process variance estimation methods to evaluate the accuracy of the model, thereby obtaining the model accuracy evaluation result;
[0072] Step S430: Perform model optimization processing based on the model accuracy evaluation result, integrate the new test sample point data into the sample point data set, reconstruct and adjust the preliminary proxy model to obtain the Kriging proxy model.
[0073] It can be understood that this process includes sample point selection, preliminary model construction, model verification and optimization. The number of initial sample points is based on the formula obtained, among which is the number of sample points, is the number of parameters that affect the resistance of the insulator. For this model, the key parameters are amplitude and wavelength, so 6 initial sample points are required. Then, the Latin hypercube sampling method is further used to uniformly sample from two sets of data sets with amplitudes ranging from 0mm to 2mm and wavelengths ranging from 5.6mm to 28mm, generating 6 sets of initial sample points with different amplitudes and wavelengths. Using these sample points, the corresponding resistance values are calculated through simulation software to provide initial data for the proxy model. In order to construct a high-precision proxy model, second-order polynomial regression and Gaussian correlation function are selected. The Gaussian function is selected because the phenomenon under study is continuously differentiable. This function exhibits the characteristics of an approximate parabola near the origin, which is suitable for capturing smooth physical phenomena. The model verification uses generalized least squares estimation and process variance estimation to evaluate the accuracy of the model by comparing the predicted values and actual values of the test sample points. As new test sample points are introduced into the initial sample data set, the model is continuously updated and optimized, and the error is gradually reduced. Finally, this iterative process formed a Kriging proxy model consisting of 24 groups of sample points, ensuring that the error between the proxy model prediction and the actual simulation value was less than 2%. This process significantly improved the prediction accuracy of the model, provided a reliable theoretical basis and data support for further optimizing the insulator design, and effectively achieved accurate evaluation of resistance and minimized the number of design iterations during the design process, thus improving overall efficiency.
[0074] Step S500, performing prediction processing according to the Kriging proxy model, predicting the value of the unsampled point by a spatial interpolation method to obtain a predicted value, wherein the predicted value includes at least two sets of parameter values of amplitude and wavelength;
[0075] It should be noted that step S500 includes steps S510 to S540.
[0076] Step S510, calculating the prediction parameter range according to the Kriging proxy model, determining the parameter range of the unsampled points by analyzing the amplitude and wavelength parameter ranges in the sample point data set, and obtaining the preliminary range of the prediction parameters;
[0077] It is understandable that due to the limited coverage of the sample points, there are some potential parameter areas that have not been explored. Through statistical analysis of the sample data, including statistical parameters such as mean and standard deviation, it is possible to identify parameter intervals that may affect the performance of the model. This step ultimately determines the preliminary prediction parameter range for the unsampled points.
[0078] Step S520, performing spatial interpolation processing according to the preliminary range, calculating the predicted value of each unsampled point by using the spatial interpolation algorithm of the Kriging proxy model, and obtaining a preliminary predicted value set;
[0079] It can be understood that through the global interpolation capability of the Kriging model, the generated set of preliminary prediction values can effectively capture the impact of parameter changes on insulator resistance.
[0080] Step S530, performing evaluation processing according to the preliminary predicted value set, by comparing the errors between the sample points and the predicted values in the Kriging proxy model, and using the weighted root mean square error to assign different weights to the amplitude and wavelength to calculate the evaluation result;
[0081] It should be noted that since the predicted values may have certain errors, these results need to be verified. The weighted root mean square error is used as the evaluation criterion. This error metric applies different weights to amplitude and wavelength to reflect their importance and influence in the physical model.
[0082] Step S540: Filter and obtain the final prediction value based on the evaluation results.
[0083] Step S600: input the predicted value into a preset multi-population genetic algorithm, and obtain the final optimized parameter value through iterative optimization processing;
[0084] It should be noted that step S600 includes steps S610 to S640.
[0085] Step S610: performing population initialization processing according to the predicted values, and obtaining an initial population by randomly assigning the predicted values to at least two different populations;
[0086] Step S620: performing collaborative search and crossover mutation processing according to the initial population, selecting different crossover mutation control parameters for each population based on a preset collaborative search strategy, and performing independent crossover and mutation to obtain an updated population;
[0087] Step S630: Perform immigration operator operation between populations according to the updated population, exchange the best individuals between different populations, and introduce the best individuals in each population into other populations to replace the worst individuals, so as to obtain an optimized population;
[0088] Step S640: Perform evaluation processing according to the optimized population, take the insulator resistance as the optimization objective function, and use the Kriging proxy model to calculate the resistance value of each individual, determine whether the optimal individual maintains the optimal state within the set algebra, and obtain the final optimization parameter value.
[0089] Specifically, in this embodiment, a multi-population genetic algorithm is used to optimize the parameters. The genetic algorithm is a highly parallel, random, and adaptive global optimization probability search algorithm developed by drawing on the natural selection and evolution mechanism of the biological world. Since it does not rely on gradients during optimization, it has strong robustness. However, in the process of using the genetic algorithm, it is found that the original genetic algorithm will fall into the problem of immature convergence. When individual supernormal individuals appear in the population, the population will fall into the control of the supernormal individual, causing the population evolution to stagnate. In order to change the problem of immature convergence, the improved multi-population genetic algorithm introduces the concept of multi-population. First, instead of optimizing the search for a single population, a collaborative search is performed on multiple populations. To ensure that the offspring of different populations are different, different control parameters such as crossover and mutation are selected for each population to ensure the balance of the global and local search capabilities of the population. Then, to ensure the interconnection between populations, different populations are connected using immigration operators. The best individuals of different populations in the evolution process are introduced into other populations through immigration operators, and the best individuals replace the worst individuals to form a new population, achieving the collaborative evolution between different populations. Finally, the best individuals obtained in the evolution process of different populations are retained through artificial selection operators to obtain the elite population. The elite population does not perform genetic operations such as selection, crossover, and mutation to ensure that the best individuals produced by various populations in the evolution process are not destroyed or lost. The algorithm is terminated by judging whether the best individual can always maintain the best in the set minimum retention generation. In this example, the number of individuals in the population is set to 200, the population size is set to 10, the mutation probability of each offspring is randomly selected between 0.001 and 0.05, and the crossover probability is randomly selected between 0.7 and 0.9, which ensures the randomness of the crossover mutation probability of each offspring. The optimal individual is set to be maintained for at least 10 generations, and the optimization objective function is the resistance of the insulator to ensure that the resistance of the insulator is as small as possible. The resistance objective function is composed of the prediction function constructed by the agent model. By setting the relationship between the objective function and the parameters, a single-objective optimization algorithm is established with amplitude and wavelength as optimization parameters and resistance value as optimization objective. Finally, the optimal parameter combination is obtained.
[0090] Step S700: Perform simulation processing according to the optimized parameter values and the original three-dimensional model, and obtain simulation results by applying the optimized parameter values to the three-dimensional model for fluid dynamics simulation verification.
[0091] Specifically, in order to obtain the flow field of the roof insulator, COMSOL software is used for simulation calculation. The boundary conditions required for the simulation calculation are shown in Table 1 below.
[0092] Table 1 Boundary conditions table
[0093]
[0094] Furthermore, the physical fields in the finite element simulation were set as “fluid mechanics”, “RANS-based Spalart-Allmaras turbulence model”, and “global ordinary differential equations”.
[0095] Furthermore, the calculation domain in the finite element simulation is set to a cylinder, the diameter of the cylinder is three times the height of the roof insulator, the length of the cylinder is four times the height of the roof insulator, the distance between the air inlet and the roof insulator is 1.5 times the height of the roof insulator, and the distance between the air outlet and the roof insulator is 2.5 times the height of the roof insulator.
[0096] Furthermore, in the finite element simulation, the mesh of the core rod of the roof insulator is extremely refined, and the mesh of the rest is ultra-refined; a boundary layer is set for the roof insulator, the number of boundary layers is 2, the stretching factor is 1.2, and the thickness adjustment factor is 1; the complete mesh of the roof insulator includes 620521 units.
[0097] The resistance calculation formula in finite element simulation is:
[0098] ;
[0099] in, is the pressure difference resistance caused by the pressure difference between the front and rear of the roof insulator; is the viscous resistance caused by the friction between the roof insulator and the air.
[0100] The purpose of improving the flow field environment of the insulator is to reduce the resistance experienced by the insulator. According to research, the resistance experienced by the insulator is mainly divided into pressure difference resistance and viscous resistance. The pressure difference resistance is formed by the pressure difference before and after the insulator, and the viscous resistance is formed by the friction on the surface of the insulator. Through calculation, it is found that the pressure difference resistance accounts for a higher proportion of the resistance experienced by the insulator. Reducing the pressure difference resistance of the insulator by controlling the pressure on the leeward side of the insulator can effectively reduce the resistance experienced by the insulator.
[0101] refer to Figure 3 , Figure 4 The two figures show the pressure conditions of the original roof insulator and the improved roof insulator. It can be clearly observed from the figures that the windward side area of the improved roof insulator is lighter in color than the windward side area of the original roof insulator, which means that the windward side pressure of the original roof insulator is greater than the windward side pressure of the improved roof insulator. At the same time, there is an area with obvious higher pressure behind the improved roof insulator, which means that the leeward side pressure of the improved roof insulator is greater than the leeward side pressure of the original roof insulator. Under the joint effect of these two, the front and rear pressure difference of the improved roof insulator is significantly smaller than that of the original roof insulator, and the pressure difference resistance is reduced accordingly.
[0102] refer to Figure 5, the figure collects the pressure values on the leeward side of the roof insulator, and uses the drawing software to draw a comparison of the leeward side surface pressures of the two roof insulators. It can be seen from the figure that the leeward side surface pressure of the improved roof insulator is significantly greater than the leeward side surface pressure of the original roof insulator. The pressure is similar only at the core rods at both ends of the insulator where no improved structure is added, which shows the comparison of the leeward side pressure of the insulator from a numerical point of view.
[0103] refer to Figure 6 , Figure 7 The two figures show the velocity field of the original roof insulator and the improved roof insulator. The velocity field result diagram can be used to analyze the principle of the pressure change of the insulator. When high-speed airflow flows through the insulator, the insulator will have a blocking effect on the airflow. This blocking effect will form a vacuum area on the leeward side of the insulator. Under the action of the vacuum area, the high-speed airflows at both ends will converge to the inside of the insulator to form a low-velocity negative pressure area. Figure 5 , Figure 6 It can be seen that the low velocity area on the leeward side of the improved roof insulator is significantly smaller than the low velocity area on the leeward side of the original roof insulator, which indicates that the negative pressure area on the leeward side of the improved roof insulator is relatively small, and thus the pressure on the leeward side of the roof insulator is greater.
[0104] refer to Figure 8 , Fig. 9 The two figures explain the flow field of the two roof insulators from the perspective of the cylinder flow theory, and then explain the reasons for the different velocity fields of the roof insulators. It can be clearly observed from the figure that the streamlines of the improved roof insulator are significantly better than those of the original roof insulator. From the streamline diagram of the original roof insulator, it can be seen that there are a large number of eddies in the area close to the insulator, which means that the vortex shedding of the original roof insulator is serious. The shedding of eddies will interfere with the normal flow of the airflow, resulting in a decrease in the airflow velocity. The streamlines of the improved roof insulator are relatively smooth, and no obvious eddies are generated. This difference causes the flow velocity of the improved roof insulator to be significantly greater than that of the original roof insulator.
[0105] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
[0106] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.
Claims
1. A train roof insulator, characterized in that: include: Mandrel (1); as well as At least two shed structures (2), all of the shed structures (2) are fixedly arranged on the core rod (1), the shed structures (2) are disc-shaped, the shed structures (2) include two types, a large shed (21) and a small shed (22), the diameter of the large shed (21) is greater than the diameter of the small shed (22), the number of sheds of each type is at least one, the large shed (21) and the small shed (22) are alternately distributed on the core rod (1), all of the shed structures (2) are arranged along the axis direction of the core rod (1), adjacent shed structures (2) are parallel to each other, the plane of each of the shed structures (2) is perpendicular to the axis of the core rod (1), and the intervals between the shed structures (2) are fixed; The core rod (1) comprises at least one wave portion (11), the wave portion (11) is arranged between adjacent shed structures (2), and the outer surface of the wave portion (11) is periodically wave-shaped along the length direction of the core rod (1); The diameter change of the wave portion (11) follows the definition of the cosine function, and the specific formula is: D=D a +2acos(2πx / λ); Where D is the diameter of the corrugated mandrel, D a is the diameter of the original cylindrical core rod, a is the amplitude of the cosine function constituting the wavy core rod, λ is the wavelength of the cosine function, and x represents the lengthwise position of each section of the wavy portion (11).
2. The train roof insulator according to claim 1, characterized in that: The amplitude range of the cosine function is 0 mm to 2 mm.
3. The train roof insulator according to claim 1, characterized in that: The wavelength of the cosine function is 5.6 mm to 28 mm.
4. A method for optimizing the design of train roof insulators, characterized in that: include: The original structural data of the train roof insulator is acquired for modeling to obtain an original three-dimensional model, wherein the structural data includes the size parameters of the original cylindrical core rod; The structural data is processed based on a preset cosine function, by improving the cylindrical core rod to a wavy profile that satisfies the cosine function definition, and obtaining a key parameter range through parameterization, wherein the key parameter range includes a parameter range of an amplitude and a wavelength of the cosine function; Perform sampling processing according to the key parameters, and extract sample points from the parameter space by using a Latin hypercube sampling method to obtain a sample point data set; Performing model building processing according to the sample point data set to obtain a Kriging proxy model; Performing prediction processing according to the Kriging proxy model, predicting the value of the unsampled point by a spatial interpolation method to obtain a predicted value, wherein the predicted value includes at least two sets of parameter values of amplitude and wavelength; The predicted value is input into a preset multi-population genetic algorithm, and a final optimized parameter value is obtained through iterative optimization processing; A simulation process is performed according to the optimized parameter values and the original three-dimensional model, and a simulation result is obtained by applying the optimized parameter values to the three-dimensional model for fluid dynamics simulation verification.
5. The optimization design method for train roof insulators according to claim 4, characterized in that: Sampling is performed according to the key parameter range, and a sample point data set is obtained by extracting sample points from the parameter space using the Latin hypercube sampling method, including: Performing meshing processing according to the key parameter range, creating a parameter mesh in the parameter space based on a preset rule to obtain an initial parameter mesh; Performing screening processing according to the initial parameter grid, obtaining at least one set of initial sample points by calculating representative results of each grid point and screening representative reference points; Performing optimization processing according to the initial sample points, dividing the space in each parameter dimension and randomly selecting the positions of the sample points to obtain optimized sample points by applying the Latin hypercube sampling method; The data set is constructed and processed according to the optimized sample points, and simulation calculations are performed on the sample points and the corresponding models respectively to obtain the resistance value of the insulator formed by each group of sample points, thereby forming a sample point data set.
6. The optimization design method for train roof insulators according to claim 4, characterized in that: The Kriging proxy model is obtained by model building processing based on the sample point data set, including: Performing model parameter screening processing according to the sample point data set, determining the number of key parameters affecting insulator resistance, and selecting second-order polynomial regression parameters and Gaussian correlation functions to construct a preliminary proxy model; Performing model progress verification processing according to the preliminary proxy model, by comparing the predicted results of the test sample points with the actual results, and using generalized least squares estimation and process variance estimation methods to evaluate the accuracy value of the model, thereby obtaining a model accuracy evaluation result; Model optimization processing is performed based on the model accuracy evaluation result, and the preliminary proxy model is rebuilt and adjusted to obtain the Kriging proxy model by integrating new test sample point data into the sample point data set.
7. The optimization design method for train roof insulators according to claim 4, characterized in that: Prediction processing is performed according to the Kriging proxy model, and the value of the unsampled point is predicted by a spatial interpolation method to obtain a predicted value, wherein the predicted value includes at least two sets of parameter values of amplitude and wavelength, including: Calculating the prediction parameter range according to the Kriging proxy model, determining the parameter range of the unsampled points by analyzing the amplitude and wavelength parameter ranges in the sample point data set, and obtaining a preliminary range of the prediction parameter; Performing spatial interpolation processing according to the preliminary range, and calculating the predicted value of each unsampled point by using the spatial interpolation algorithm of the Kriging proxy model to obtain a preliminary predicted value set; Performing evaluation processing according to the preliminary predicted value set, by comparing the errors between the sample points and the predicted values in the Kriging proxy model, and using a weighted root mean square error to assign different weights to the amplitude and wavelength to calculate the evaluation result; The final prediction value is obtained based on the evaluation results.
8. The method for optimizing the design of train roof insulators according to claim 4, characterized in that: The predicted values are input into a preset multi-population genetic algorithm, and the final optimized parameter values are obtained through iterative optimization processing, including: Performing population initialization processing according to the predicted values, and obtaining an initial population by randomly allocating the predicted values to at least two different populations; Performing collaborative search and crossover mutation processing according to the initial population, selecting different crossover mutation control parameters for each population based on a preset collaborative search strategy, and performing independent crossover and mutation to obtain an updated population; Performing immigration operator operations between populations according to the updated population, exchanging the best individuals between different populations, and introducing the best individuals in each population into other populations to replace the worst individuals, thereby obtaining an optimized population; An evaluation process is performed based on the optimized population, by taking the insulator resistance as the optimization objective function and using the Kriging proxy model to calculate the resistance value of each individual, and judging whether the optimal individual maintains the optimal state within the set algebra, thereby obtaining the final optimization parameter value.
Citation Information
Patent Citations
Ultrasonic vibration assisted hole extrusion strengthening device as well as design method and operation process thereof
CN113977190A
Bullet train roof insulator resistance reduction structure
CN114783702A
