Optimization method of shock-resistant superstructure based on explicit topological description

Abstract

The present invention belongs to the technical field of impact load attenuation and topology optimization, and specifically relates to an optimization method for impact-resistant superstructures based on explicit topology description, including the following steps: S1: Construction of a two-dimensional cell library based on explicit topology description; the cells under the topology description method adopt C 4v symmetry, and the generation of two-dimensional cells is achieved by dividing the original square design domain. The coordinates of randomly generated control points and the width information of the initial components are used to obtain two-dimensional cells with various different configurations; S2: Description of the impact environment based on the shock response spectrum; S3: Modeling and analysis are carried out through ABAQUS with a Python interface; S4: Verify whether the cell configuration will attenuate the impact; S5: Optimize the process. The present invention can utilize the explicit topology description method, and due to its clear geometric boundary, it can describe the geometric characteristics of the structure in detail.

Description

Technical Field

[0001] The invention belongs to the technical field of impact load attenuation and topology optimization, and in particular relates to an impact-resistant superstructure optimization method based on explicit topology description. Background Art

[0002] The high-frequency, high-amplitude transient shocks generated by the aerospace separation device during the explosive separation process have a bad impact on the precision equipment inside the aerospace launch vehicle. In addition, when ships are attacked by enemy weapons in modern naval battles, the impact resistance of weak links such as important equipment and electronic components on the ship is a key factor restricting the ship's sustained combat capability, provided that the ship structure does not suffer major damage. Therefore, how to reduce the impact of shock loads on equipment has always been a hot research direction in this field.

[0003] As a forward design tool, topology optimization can creatively achieve a design from scratch within a given design domain, satisfy specific design constraints while ensuring the optimal performance, and help engineers design innovative structures and products. It has been successfully applied in various industrial fields. However, most of the current topology optimization methods are oriented towards the static performance of structures, which means that when it comes to dynamic problems, engineers and researchers often need to "equivalent" the problem to a static problem for optimization and solution. However, this so-called "equivalence" can only be guaranteed when the load is slow enough to ignore the inertia force. In other cases, the dynamic performance of the structure needs to be considered directly. Summary of the invention

[0004] The purpose of the present invention is to provide an optimization method for a shock-resistant superstructure based on explicit topological description, which can utilize the explicit topological description method to describe the geometric characteristics of the structure in detail thanks to its clear geometric boundaries.

[0005] The technical solution adopted by the present invention is as follows:

[0006] The impact-resistant superstructure optimization method based on explicit topological description includes the following steps:

[0007] S1: Construction of a two-dimensional unit cell library based on explicit topological description;

[0008] In S1, the unit cell under the topological description method adopts C 4v Symmetry is used to generate two-dimensional unit cells by dividing the original square design domain. The coordinates of randomly generated control points and the width information of the initial components are used to obtain two-dimensional unit cells of various configurations, including straight-edge components, curved-edge components and other configuration components.

[0009] For the construction of the unit cell library of straight-edge components, considering the nonlinearity of the impact load on the structural response, the parameters of the topological description function are fine-tuned and a sine term is added to ensure that the components are still connected at common nodes;

[0010] The shape perturbation of the component within half a period is considered, and the bending direction of the component is adjusted to satisfy the symmetry.

[0011] S2: Shock environment description based on shock response spectrum;

[0012] In S2, the shock response spectrum is a tool for evaluating the potential damage caused by shock, and is used to completely and effectively evaluate the destructive capacity of shock on the structure. The specific steps are as follows:

[0013] S21: A series of linear, single-degree-of-freedom spring oscillator systems with different natural frequencies are fixed on the same foundation, and arbitrary impact excitation is applied to the foundation to obtain the maximum response value of each oscillator in motion;

[0014] S22: Focus on the relative displacement, relative velocity and absolute acceleration of the oscillator. The abscissa of the obtained shock spectrum is the natural frequency based on the single-degree-of-freedom assumption, and the ordinate shows the peak value of the shock input borne by the single-degree-of-freedom system;

[0015] S23: For the separation process in the aerospace field, it is necessary to consider the impact of acceleration response on sensitive electronic components in the shock test platform and use the acceleration shock response spectrum to describe the shock reduction performance of the designed structure.

[0016] S3: Modeling and analysis through ABAQUS with Python interface;

[0017] In S3, during the optimization process, the configuration of the unit cell keeps changing, and the model needs to be continuously updated. The modeling and analysis process is as follows:

[0018] S31: Finite element analysis is proposed to be carried out by batch parametric modeling;

[0019] S32: The pre- and post-processing modules of ABAQUS are compiled in Python;

[0020] Among them, the post-processing module includes the import of models, the assignment of materials, the array of unit cells, the division of units, and the data extraction that the job submission and post-processing module focuses on;

[0021] S33: Running the Python program completes the entire modeling and analysis process.

[0022] In the S3, a finite system is simulated by a unit cell array. The unit cells are arrayed in 5X5, and rigid plates are established at the bottom and top of the model respectively. The rigid plates and the design structure are connected by ties in the contact relationship. An axial acceleration time domain load is applied to the bottom rigid plate reference point, and a point section is set at the top reference point, and a large mass attribute is assigned to simulate a satellite connection on the top.

[0023] S4: Verify whether the unit cell configuration will attenuate the impact;

[0024] In S4, the specific steps of determining whether the impact will be attenuated are as follows:

[0025] S41: Randomly generate four sets of two-dimensional single-phase material structures to determine whether the topological structure affects the model's impact reduction ability;

[0026] Among them, the material parameters are aluminum alloy, and the density is 2785kg / m 3 , elastic modulus 71Gpa, Poisson's ratio 0.32;

[0027] S42: Extract the acceleration of the top and bottom reference points of the model and measure it using the shock response spectrum;

[0028] S43: Input the same load and compare the output of the four models under the same load to verify whether the impact attenuation capacity is related to the unit cell configuration.

[0029] S5: Optimize the process.

[0030] In S5, a genetic algorithm is introduced into the optimization process to directly use the fitness function of the target as the search basis, without the need for derivatives, without any requirements on the continuity and differentiability of the target function, and the definition domain can be arbitrarily assumed, without the need for complex mathematical transformation and constraint processing of the optimization target;

[0031] Among them, the genetic algorithm uses probabilistic search technology, that is, searching according to specific rules and methods to obtain the optimal value.

[0032] The technical effects achieved by the present invention are:

[0033] The impact-resistant superstructure optimization method based on explicit topological description of the present invention utilizes the mobile deformable component method to generate a unit cell library through explicit topological description. Unlike the traditional SIMP method, which has a large number of design variables and requires post-processing, this method can be seamlessly integrated with the CAD system.

[0034] The impact-resistant superstructure optimization method based on explicit topological description of the present invention is different from the traditional heuristic design, in which the design space is limited. The design space of this work is larger, and a genetic algorithm is used to find the global optimal topology.

[0035] Compared with other working design structures, the impact-resistant superstructure optimization method based on explicit topological description of the present invention often relies on a combination of multi-phase materials, which may lead to poor joints and bring manufacturing difficulties. This scheme proposes a superstructure of single-phase material, which exhibits better impact resistance in a wide frequency range. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 The embodiment of the present invention is based on the display topology description unit cell construction method and the corresponding unit cell library;

[0037] Figure 2 It is a curved edge unit cell library constructed based on the display topology description unit cell according to an embodiment of the present invention;

[0038] Figure 3 Schematic diagram of the impact response spectrum conversion principle of an embodiment of the present invention;

[0039] Figure 4 The embodiment of the present invention randomly generates a unit cell array finite system model;

[0040] Figure 5 are shock response spectrum curves corresponding to different unit cell configurations of the embodiments of the present invention;

[0041] Figure 6 is an optimization flow chart of an embodiment of the present invention;

[0042] Figure 7 1 is the optimal unit cell configuration and the schematic diagram of the finite system of the embodiment of the present invention;

[0043] Figure 8 It is a comparison of input and output acceleration time domain curves and corresponding impact response spectra of the embodiment of the present invention;

[0044] Fig. 9 It is a stress cloud diagram of a pseudo three-dimensional model according to an embodiment of the present invention;

[0045] Fig.10 It is a stress cloud diagram of an undesigned model in an embodiment of the present invention;

[0046] Fig.11 It is a schematic diagram comparing the output of the optimal design configuration of the embodiment of the present invention and the undesigned model;

[0047] Fig.12 4x4 is a schematic diagram of the optimization result of the initial unit cell number according to an embodiment of the present invention;

[0048] Fig.13 1 is a schematic diagram of the optimization result of the number of 6x6 initial unit cells according to an embodiment of the present invention;

[0049] Fig.14 is a schematic diagram of a unit cell with different curvatures at a fixed node position according to an embodiment of the present invention;

[0050] Among them, (a) model-1 (0.003); (b) model-2 (0.005); (c) model-3 (0.010);

[0051] Fig.15 is the impulse response spectrum of the same output point of the unit cell with different curvatures in the embodiment of the present invention;

[0052] Fig.16 is a schematic diagram of optimization results of variable curvature of fixed nodes according to an embodiment of the present invention;

[0053] Fig.17 Schematic diagram of the optimized configuration of the curved edge system according to an embodiment of the present invention;

[0054] Fig.18 is a schematic diagram comparing the input and output accelerations of the curved edge system according to an embodiment of the present invention;

[0055] Fig.19 It is a schematic diagram comparing the same output acceleration of a straight edge system and a curved edge system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0056] In order to make the purpose and advantages of the present invention more clearly understood, the present invention is specifically described below in conjunction with embodiments. It should be understood that the following text is only used to describe one or several specific embodiments of the present invention, and does not strictly limit the scope of protection of the specific claims of the present invention.

[0057] Embodiment 1:

[0058] like Figure 1-Figure 6 As shown, the impact-resistant super structure optimization method based on explicit topological description includes the following steps:

[0059] S1: Construction of a two-dimensional unit cell library based on explicit topological description;

[0060] In S1, the unit cell under the topological description method adopts C 4v Symmetry is used to generate two-dimensional unit cells by dividing the original square design domain. The coordinates of randomly generated control points and the width information of the initial components are used to obtain two-dimensional unit cells of various configurations.

[0061] like Figure 1 and Figure 5 As shown in the figure, the square design domain represented by the black lines is divided into 8 subdomains by thin gray dividing lines. Three control points X1(x1,0), X2(0.5,x2), and X3(x3,x3) are randomly selected on the edges AB, BO, and AO, and connected to X1X2, X1X3, X2X3, X2O, and X3O respectively to form five initial components ( Figure 1 light gray component), and randomly select the initial component width ti ∈(0.001,0.25),i=1,…,5, the components on the other seven regions (thick gray components) can be generated according to the symmetry of the overall unit cell.

[0062] For the construction of the unit cell library of straight-edge components, considering the nonlinearity of the impact load on the structural response, the parameters of the topological description function are fine-tuned and a sine term is added to ensure that the components are still connected at common nodes;

[0063] The shape perturbation of the component within half a period is considered, and the bending direction of the component is adjusted to satisfy the symmetry.

[0064] S2: Shock environment description based on shock response spectrum;

[0065] In S2, the shock response spectrum is a tool for evaluating the potential damage caused by shock, and is used to completely and effectively evaluate the destructive capacity of shock on the structure. The specific steps are as follows:

[0066] S21: A series of linear, single-degree-of-freedom spring oscillator systems with different natural frequencies are fixed on the same foundation, and arbitrary impact excitation is applied to the foundation to obtain the maximum response value of each oscillator in motion;

[0067] S22: Focus on the relative displacement, relative velocity and absolute acceleration of the oscillator. The abscissa of the obtained shock spectrum is the natural frequency based on the single-degree-of-freedom assumption, and the ordinate shows the peak value of the shock input borne by the single-degree-of-freedom system;

[0068] S23: For the separation process in the aerospace field, it is necessary to consider the impact of acceleration response on sensitive electronic components in the shock test platform and use the acceleration shock response spectrum to describe the shock reduction performance of the designed structure.

[0069] S3: Modeling and analysis through ABAQUS with Python interface;

[0070] In S3, the simulation is completed by the commercial software ABAQUS. During the optimization process, the configuration of the unit cell is constantly changing, and the model needs to be continuously updated. However, in the finite element analysis process, blindly repeating manual modeling and analysis is inefficient, and the pre-processing will take up a lot of working time. In order to improve the efficiency of the entire analysis, the modeling and analysis process is as follows:

[0071] S31: Finite element analysis is proposed to be carried out by batch parametric modeling;

[0072] S32: The pre- and post-processing modules of ABAQUS are compiled in Python;

[0073] Among them, the post-processing module includes the import of models, the assignment of materials, the array of unit cells, the division of units, and the data extraction that the job submission and post-processing module focuses on;

[0074] S33: Running the Python program completes the entire modeling and analysis process.

[0075] In the S3, a finite system is simulated by a unit cell array. The unit cells are arrayed in 5X5, and rigid plates are established at the bottom and top of the model respectively. The rigid plates and the design structure are connected by ties in the contact relationship. An axial acceleration time domain load is applied to the bottom rigid plate reference point, and a point section is set at the top reference point, and a large mass attribute is assigned to simulate a satellite connection on the top.

[0076] S4: Verify whether the unit cell configuration will attenuate the impact;

[0077] In S4, the specific steps of determining whether the impact will be attenuated are as follows:

[0078] S41: Randomly generate four sets of two-dimensional single-phase material structures to determine whether the topological structure affects the model's impact reduction ability;

[0079] Among them, the material parameters are aluminum alloy, and the density is 2785kg / m 3 , elastic modulus 71Gpa, Poisson's ratio 0.32;

[0080] S42: Extract the acceleration of the top and bottom reference points of the model and measure it using the shock response spectrum;

[0081] S43: Input the same load and compare the output of the four models under the same load to verify whether the impact attenuation capacity is related to the unit cell configuration.

[0082] The calculated shock response spectrum curves of different configurations are as follows: Figure 5 As shown in the figure, the horizontal axis is frequency, the vertical axis is the maximum amplitude corresponding to the response of a single free oscillator, curve b is the excitation load at the input end, each excitation load is the same, and curve a is the output response. It can be found that different configurations have a significant effect on the output response.

[0083] (a) The impulse spectrum peak corresponding to the unit cell has a significant amplification effect compared to the input, and in certain specific frequency ranges, the output response is significantly amplified compared to the input. Only in the range of 500 to 1800 Hz is the impulse attenuated.

[0084] (b) The peak value of the shock spectrum corresponding to the unit cell remains flat, and there is no attenuation of the shock.

[0085] The peak value of the shock spectrum of (c) is significantly attenuated, indicating that the shock is effectively suppressed.

[0086] (d) The configuration of the unit cell attenuates shocks in the entire frequency range, achieving a broadband shock-resistant design.

[0087] S5: Optimize the process.

[0088] In S5, due to the complexity of the load, it is impossible to obtain the derivative information of the response to the design variable, so this process can only be achieved by automatically searching for the optimal solution through the optimization algorithm. A genetic algorithm is introduced into the optimization process to directly use the fitness function of the target as the search basis, without the need for derivatives, and without any requirements for the continuity and differentiability of the target function. The definition domain can be arbitrarily assumed, and there is no need to perform complex mathematical transformations and constraint processing on the optimization target.

[0089] Among them, the genetic algorithm uses probabilistic search technology, that is, it searches according to specific rules and methods to obtain the optimal value. Later, we use the genetic algorithm to achieve the maximum peak value of the impact response spectrum at the output end compared to the input end. The genetic algorithm sets parameters, where the population size is 50, the crossover probability is 0.5, and the maximum iteration number is 30 generations. The optimization process uses Matlab commercial software and ABAQUS joint simulation, and the genetic algorithm calls Matlab's own toolbox.

[0090] Embodiment 2:

[0091] Numerical example:

[0092] In the embodiment, a straight edge component calculation example and a curved edge component calculation example will be shown to verify the efficiency and accuracy of the method. In addition, all calculation examples do not consider the volume, and all calculation examples are calculated by a computer equipped with a 12th Gen Intel(R) Core(TM) i7-12700KF 3.60GHz CPU and 32GB memory.

[0093] Embodiment 3:

[0094] Straight edge component study:

[0095] In this embodiment, Figure 7-Figure 13 As shown in the figure, the genetic algorithm is used to optimize and maximize the impact attenuation. The final optimized unit cell configuration is as follows Figure 7 and Figure 8 As shown in the figure, the unit cell is filled densely in the center to form a local resonance structure, and the surrounding parts are connected to each other through fine components. The time domain curves of the acceleration at the loading point and the output point are extracted. The axial acceleration at the output end is significantly lower than that at the input end, and the overall oscillation range is narrowed, indicating that less energy is transmitted to the output end, achieving the purpose of reducing the impact.

[0096] like Fig. 9 and Fig.10As shown, the superstructure considered assumes that the length of the third direction is large enough and the elastic wave propagates in the plane, so it is simplified to a two-dimensional optimization problem. In order to verify that the simplified superstructure can achieve the purpose of impact resistance, the optimal result is stretched to a certain thickness (0.02) and compared with the undesigned control model.

[0097] In order to further compare the optimized model to effectively reduce the impact, we extracted the acceleration response curves of the optimal model and the control model at the same output point, and compared the impact spectra, such as Fig.11 As shown in the figure, the axial acceleration of the optimal array is significantly attenuated, and it is almost a straight line compared with the control group, indicating that the acceleration oscillation range is very small. It can also be observed from the shock spectrum that the optimal model achieves shock attenuation in a wide frequency range compared with the control group, and the peak value decreases significantly.

[0098] In order to verify whether the number of initial unit cells will lead to changes in the optimal configuration, we adjusted the size of the unit cells in the same design domain size, arrayed 4x4 and 6x6 initial unit cells respectively, and optimized based on the same objective function. The optimization results are shown in the figure below. Fig.12 and Fig.13 shown.

[0099] Under different initial conditions, the optimized unit cell configurations are different, but they all tend to be filled with density in the middle, connected by thin components on all sides, and the acceleration response extracted at the output end is attenuated compared to the input end, indicating that this type of configuration is beneficial to shock attenuation. This shows the effectiveness of the work.

[0100] Embodiment 4:

[0101] In order to explore whether the curved edge components can significantly improve the impact resistance, we first fixed the node position and only adjusted the curvature of the curved edge components to compare the impact resistance of components with different curvatures. Fig.14 As shown in the figure, the bending degree of the components corresponding to the three unit cell models gradually increases, and the bending degree of each component becomes larger, but because the perturbation adjustment is within half or one cycle, the components are still connected to the same node, and there is no obvious overlap of the components. We also array the unit cells into 5X5 and solve them based on explicit dynamics in ABAQUS to extract the response acceleration time domain curve and shock spectrum curve of the same output point.

[0102] like Fig.15 As shown in the figure, as the bending degree of the component increases, the acceleration oscillation range of the same node corresponding to the output end becomes smaller, indicating that the acceleration is attenuated. From the corresponding impact response spectrum, it can be found that the greater the curvature of the component, the lower the corresponding peak value. It can be concluded that the bending degree of the component has a significant impact on the acceleration of the output end.

[0103] Then, the nodes are still fixed and only the curvature is adjusted. The curvature of the component is optimized based on the above objective function using the genetic algorithm. The lattice model is optimized as follows: Fig.16 As shown in the figure, the overall bending degree of the unit cell is large. We also compared the optimal result with the straight component and found that the acceleration can be significantly attenuated by simply adjusting the bending degree of the component without changing the node position. At the connection of the curved edge component, the elastic wave propagation energy is reduced due to the larger angle formed by the bending of the component, protecting the sensitive components above from impact. In the future, we can consider changing the nodes, component width and adjusting the curvature to analyze the impact resistance of the component.

[0104] The final optimized curved edge optimal unit cell configuration is as follows Figure 17-19 As shown in the figure, it can be observed that the optimal curved edge unit cell is somewhat curved compared to the straight edge, and still has an intermediate filling density and is surrounded by curved components, indicating that this type of configuration has certain advantages in shock isolation. By comparing the input and output peak accelerations of the shock system, it can be found that the input peak acceleration of the curved edge component system is 1.023×104m.s -2 , the peak acceleration output is 277.813ms -2 , it is concluded that the curved edge component system can also achieve impact load attenuation. Next, by comparing the acceleration time domain curves of the same output point of the straight edge system (SC) and the curved edge system (SSC), it is found that the acceleration oscillation range of the curved edge system is significantly smaller, indicating that the curved edge system has more advantages than the straight edge system in terms of impact attenuation. This can also be seen from the impact response spectrum. The peak value of the response spectrum of the curved edge system is smaller than that of the straight edge system. The possible mechanism for the superiority of the curved edge system is that the angle of the connection between the curved edge and the straight edge is larger, resulting in energy loss here. It can also be considered from another perspective. According to the single-degree-of-freedom forced vibration, the acceleration of the forced object is related to the impulse, structural stiffness and mass of the forced object. Under the condition of consistent impact load intensity, the greater the system stiffness, the greater the acceleration of the object's movement. The lower the overall stiffness of the curved edge system, the stronger its impact attenuation energy.

[0105] The above is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications should also be considered as the protection scope of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention shall be implemented according to the conventional means in the art unless otherwise specified and limited.

Claims

1. Optimal design of shock-resistant superstructure based on explicit topological description, characterized by: The following steps are involved: S1: Construction of a two-dimensional unit cell library based on explicit topological description; In S1, the unit cell under the topological description method adopts C 4v Symmetry, by dividing the original square design domain for the generation of two-dimensional unit cells, using the coordinates of randomly generated control points and the width information of the initial components to obtain a variety of two-dimensional unit cells with different configurations; For the construction of the unit cell library of straight-edge components, considering the nonlinearity of the impact load on the structural response, the parameters of the topological description function are fine-tuned and a sine term is added to ensure that the components are still connected at common nodes; Consider the shape perturbation of the component within half a period and adjust the bending direction of the component to meet the symmetry; S2: Shock environment description based on shock response spectrum; S3: Modeling and analysis through ABAQUS with Python interface; In S3, a finite system is simulated by a unit cell array, a 5X5 array is made for the unit cells, and rigid plates are respectively established at the bottom and top of the model. The rigid plates and the design structure are connected by ties in the contact relationship. An axial acceleration time domain load is applied to the bottom rigid plate reference point, and a point section is set at the top reference point, and a large mass attribute is assigned to simulate a satellite connection on the top. S4: Verify whether the unit cell configuration will attenuate the impact; In S4, the specific steps of determining whether the impact will be attenuated are as follows: S41: Randomly generate four sets of two-dimensional single-phase material structures to determine whether the topological structure affects the model's impact reduction ability; Among them, the material parameters are aluminum alloy, and the density is 2785kg / m 3 , elastic modulus 71Gpa, Poisson's ratio 0.32; S42: Extract the acceleration of the top and bottom reference points of the model and measure it using the shock response spectrum; S43: Input the same load and compare the output of the four models under the same load to verify whether the impact attenuation capacity is related to the unit cell configuration; S5: Optimize the process.

2. The impact-resistant superstructure optimization design based on explicit topological description according to claim 1, characterized in that: In S2, the shock response spectrum is a tool for evaluating the potential damage caused by shock, and is used to completely and effectively evaluate the destructive capacity of shock on the structure. The specific steps are as follows: S21: A series of linear, single-degree-of-freedom spring oscillator systems with different natural frequencies are fixed on the same foundation, and arbitrary impact excitation is applied to the foundation to obtain the maximum response value of each oscillator in motion; S22: Focus on the relative displacement, relative velocity and absolute acceleration of the oscillator. The abscissa of the obtained shock spectrum is the natural frequency based on the single-degree-of-freedom assumption, and the ordinate shows the peak value of the shock input borne by the single-degree-of-freedom system; S23: For the separation process in the aerospace field, it is necessary to consider the impact of acceleration response on sensitive electronic components in the shock test platform and use the acceleration shock response spectrum to describe the shock reduction performance of the designed structure.

3. The impact-resistant superstructure optimization design based on explicit topological description according to claim 1, characterized in that: In S3, during the optimization process, the configuration of the unit cell keeps changing, and the model needs to be continuously updated. The modeling and analysis process is as follows: S31: Finite element analysis is proposed to be carried out by batch parametric modeling; S32: The pre- and post-processing modules of ABAQUS are compiled in Python; Among them, the post-processing module includes the import of models, the assignment of materials, the array of unit cells, the division of units, and the data extraction that the job submission and post-processing module focuses on; S33: Running the Python program completes the entire modeling and analysis process.

4. The impact-resistant superstructure optimization design based on explicit topological description according to claim 1, characterized in that: In S5, a genetic algorithm is introduced into the optimization process to directly use the fitness function of the target as the search basis, without the need for derivatives, without any requirements on the continuity and differentiability of the target function, and the definition domain can be arbitrarily assumed, without the need for complex mathematical transformation and constraint processing of the optimization target; Among them, the genetic algorithm uses probabilistic search technology, that is, searching according to specific rules and methods to obtain the optimal value.

Images