ANTENNA ARRAY OPTIMIZATION CONTROL METHOD AND SYSTEM

Abstract

The present invention provides an optimization control method and system for an antenna array, including: obtaining the wave-transmitting rainproof coating parameters and conductive coating parameters of a single antenna in the antenna array; converting a feature set into a level parameter and sorting them to obtain a feature correlation parameter; selecting a feature correlation sub-parameter of the feature correlation parameter by a two-way search method; inputting the feature correlation sub-parameter into a screening model to obtain a screened feature correlation sub-parameter; inputting the screened feature correlation sub-parameter into an optimization model; inputting the wave-transmitting rainproof coating parameters and conductive coating parameters of multiple different single antennas into a trained optimization model to obtain output omnidirectional gain, directional beam gain, beam width and sidelobe suppression. A new simulation approach is provided for selecting the coating material and properties of antenna monomers, improving the optimization control capability of antenna arrays, improving the accuracy of antenna material optimization selection, saving test time and reducing costs.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to an optimization control method of an antenna array, an optimization control system of an antenna array, a computer device and a storage medium. Background Art

[0002] Antenna array optimization is a complex and multi-dimensional problem, which involves multiple considerations and the application of multiple technical means. In the existing technology, antenna array optimization can be performed through particle swarm optimization, quantum particle swarm optimization, genetic algorithm, machine learning assisted optimization, sparse linear array optimization, sparse linear array optimization, differential evolution algorithm and other methods. Specifically, the existing antenna optimization design can simulate the distribution and propagation of electromagnetic fields through electromagnetic simulation software. However, the existing antenna array optimization consumes a lot of computing resources. Summary of the invention

[0003] In view of the above problems, the present embodiment is proposed to provide an antenna array optimization control method, an antenna array optimization control system, an antenna array optimization control system, a computer device and a storage medium that overcome the above problems or at least partially solve the above problems.

[0004] In order to solve the above problem, this embodiment discloses an optimization control method of an antenna array, including:

[0005] Obtain the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the single antenna in the antenna array;

[0006] Obtaining the omnidirectional gain, directional beam gain, beam width and sidelobe suppression of the antenna array corresponding to the wave-transmitting and rain-proof coating parameters and the conductive coating parameters;

[0007] The wave-transmitting and rainproof coating parameters and the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression are combined into a feature set;

[0008] Converting the feature set into a level parameter and sorting the level parameters to obtain a feature correlation parameter;

[0009] Selecting feature correlation sub-parameters of feature correlation parameters by bidirectional search;

[0010] Inputting the feature correlation sub-parameter into the screening model to obtain the screened feature correlation sub-parameter;

[0011] Inputting the screened feature correlation sub-parameters into the optimization model to obtain a trained optimization model;

[0012] The wave-transmitting and rain-proof coating parameters and conductive coating parameters of multiple different single antennas are input into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

[0013] Preferably, converting the feature set into a level parameter and sorting the level parameters to obtain a feature correlation parameter includes:

[0014] Calculate the average value of each data in the feature set and sort them, and determine whether the average value meets the preset threshold range;

[0015] If the average value meets the preset threshold range, the level parameter corresponding to the preset threshold range is obtained;

[0016] The level parameter and the data in the corresponding feature set are converted into feature correlation parameters.

[0017] Preferably, the step of selecting the feature correlation sub-parameters of the feature correlation parameter by bidirectional search includes:

[0018] Starting from the empty set of feature correlation parameters through the sequence forward selection method, a certain feature correlation parameter is selected each time to be added to the feature subset of the feature correlation parameters so that the evaluation function reaches the optimal value;

[0019] Starting from the full set of feature correlation parameters through sequential backward selection, a feature correlation sub-parameter is eliminated from the full set of feature correlation parameters each time, so that the evaluation function reaches the optimal value after eliminating the feature correlation sub-parameter, and the feature correlation sub-parameter is selected.

[0020] Preferably, the inputting the feature correlation sub-parameters into the screening model to obtain the screened feature correlation sub-parameters comprises:

[0021] Establish the weak learner and loss function of the CART regression tree model;

[0022] The feature correlation sub-parameter is input into the weak learner, the weak learner is initialized, each regression tree model is iteratively optimized by gradient descent, and multiple base learners are iteratively generated. In each iteration, the gradient and second-order derivative of the current regression tree model are calculated to obtain the gradient and second-order derivative of the new regression tree model; a new regression tree model is constructed by the gradient and second-order derivative of the new regression tree model until the loss function converges to obtain the screened feature correlation sub-parameter.

[0023] Preferably, the step of inputting the screened feature correlation sub-parameters into an optimization model to obtain a trained optimization model comprises:

[0024] Constructing multiple decision tree models, and forming an optimization model through the decision tree models;

[0025] At each splitting node of each decision tree model, feature data is randomly selected for segmentation according to the selected feature correlation sub-parameters until the function converges to obtain the initial prediction result;

[0026] The initial prediction results are analyzed through the improved voting operation to complete the training of the optimization model.

[0027] Preferably, the step of inputting the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression comprises: inputting the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas into multiple decision tree models constructed by the optimization model;

[0028] At each splitting node of each decision tree model, feature data is randomly selected to segment the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas until the function converges to obtain an initial prediction result;

[0029] The initial prediction results are analyzed through the improved voting operation to obtain the final omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

[0030] This embodiment discloses an optimization control system of an antenna array, including:

[0031] The first acquisition module is used to acquire the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the single antenna in the antenna array;

[0032] A second acquisition module is used to obtain the omnidirectional gain, directional beam gain, beam width and sidelobe suppression of the antenna array corresponding to the wave-transmitting and rain-proof coating parameters and the conductive coating parameters;

[0033] A merging module, used to merge the wave-transmitting and rain-proof coating parameters and the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression into a feature set;

[0034] A sorting module, used for converting the feature set into a level parameter and sorting the level parameters to obtain a feature correlation parameter;

[0035] A selection module, used for selecting a feature correlation sub-parameter of a feature correlation parameter by a bidirectional search method;

[0036] A screening module, used for inputting the feature correlation sub-parameter into the screening model to obtain the screened feature correlation sub-parameter;

[0037] A training module, used for inputting the screened feature correlation sub-parameters into the optimization model to obtain a trained optimization model;

[0038] The output module is used to input the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of multiple different single antennas into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

[0039] Preferably, the sorting module comprises:

[0040] A sorting submodule is used to calculate the average value of each data in the feature set and sort it, and determine whether the average value meets a preset threshold range;

[0041] An acquisition submodule, used for acquiring a level parameter corresponding to a preset threshold range if the average value meets the preset threshold range;

[0042] The conversion submodule is used to convert the level parameter and the data in the corresponding feature set into feature correlation parameters.

[0043] This embodiment further discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the above-mentioned steps of optimizing the control of the antenna array when executing the computer program.

[0044] This embodiment further discloses a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned steps of optimizing the control of the antenna array are implemented.

[0045] This embodiment includes the following advantages:

[0046] In an embodiment of the present invention, the optimization control method of the antenna array includes: obtaining the wave-transmitting rainproof coating parameters and the conductive coating parameters of the single antenna in the antenna array; obtaining the omnidirectional gain, directional beam gain, beam width and sidelobe suppression of the antenna array corresponding to the wave-transmitting rainproof coating parameters and the conductive coating parameters; merging the wave-transmitting rainproof coating parameters and the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression into a feature set; converting the feature set into a level parameter and sorting them to obtain a feature correlation parameter; selecting a feature correlation sub-parameter of the feature correlation parameter by a two-way search method; inputting the feature correlation sub-parameter into the screening model to obtain the screened feature correlation sub-parameter; inputting the screened feature correlation sub-parameter into the optimization model to obtain a trained optimization model; inputting the wave-transmitting rainproof coating parameters and the conductive coating parameters of multiple different single antennas into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression. It provides a new simulation approach for selecting the coating material and properties of antenna monomers, improves the optimization control capability of antenna arrays, improves the accuracy of antenna material optimization selection, saves test time and reduces costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solution in this embodiment, the drawings required for use in the description of the embodiment will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 is a flowchart of steps of an embodiment of an antenna array optimization control method of this embodiment;

[0049] Figure 2 is a structural block diagram of an embodiment of an antenna array optimization control system of the present embodiment;

[0050] Figure 3 The diagram is a diagram of the internal structure of a computer device according to an embodiment. DETAILED DESCRIPTION

[0051] In order to make the technical problems, technical solutions and beneficial effects solved by this embodiment more clearly understood, this embodiment is further described in detail below in conjunction with 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 used to limit the present invention.

[0052] In a core concept in an embodiment of the present invention, various surface coatings of a single antenna are one of the important factors affecting the signal quality of an antenna array. In an embodiment of the present invention, the relationship between various properties of different types of surface coatings and signal quality strength is used as an initial training sample, and the relationship between various properties of the different types of surface coatings and signal quality strength is first feature processed and feature screened to obtain training samples with high training value, and then the training samples with high training value are used to train an optimization model, and the optimization model is used to predict corresponding information quality strength parameters, namely omnidirectional gain, directional beam gain, beam width and sidelobe suppression, etc., to provide a new simulation approach for selecting coating materials and properties of antenna single units, improve the optimization control capability of antenna arrays, improve the accuracy of optimized selection of antenna materials, save test time, and reduce costs.

[0053] Reference Figure 2 , shows a flowchart of an embodiment of an antenna array optimization control method of this embodiment, which may specifically include the following steps:

[0054] Step 101, obtaining the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the single antenna in the antenna array;

[0055] In an embodiment of the present invention, the antenna array may include multiple antenna sub-arrays, and the antenna sub-array may include a single antenna; the main material of each single antenna is generally a coating on a metal or lightweight plastic main body, and the coating may include wave-transmitting and rain-proof coating parameters and a conductive coating. Specifically, the main material of the single antenna may be a metal material such as aluminum alloy, a fiber-reinforced composite material, a metamaterial, etc., and the embodiment of the present invention does not impose too many restrictions on this, and the coating may also include other types of coatings, which are not restricted by the embodiment of the present invention.

[0056] In the embodiment of the present invention, the method can be applied to various terminals such as smart computers. The embodiment of the present invention does not limit the specific type of the terminal. The operating system of the terminal may include Android, Harmony OS, IOS, Windows Phone, Windows, etc., and the present invention does not impose too many restrictions on this.

[0057] In the embodiment of the present invention, the wave-transmitting and rain-proof coating parameters refer to the thickness and surface resistance of the wave-transmitting and rain-proof coating when other conditions remain unchanged; and the conductive coating parameters also refer to the thickness and surface resistance of the conductive coating when other conditions remain unchanged.

[0058] Step 102, obtaining the omnidirectional gain, directional beam gain, beam width and sidelobe suppression of the antenna array corresponding to the wave-transmitting and rain-proof coating parameters and the conductive coating parameters;

[0059] Furthermore, in the embodiment of the present invention, various quality parameters of the antenna array can also be obtained, such as omnidirectional gain, directional beam gain, beam width and sidelobe suppression. It should be noted that the quality parameters refer to the omnidirectional gain, directional beam gain, beam width and sidelobe suppression when a wave-transmitting rainproof coating and / or a conductive coating is provided under a single condition (such as thickness and surface resistance);

[0060] For example, as shown in Table 1 below:

[0061]

[0062] Gain refers to the ratio of the radiated power flux density of an antenna in a certain specified direction to the maximum radiated power flux density of a reference antenna at the same input power. In the embodiment of the present invention, it mainly includes omnidirectional gain and directional beam gain.

[0063] The sidelobe suppression in the embodiment of the present invention mainly refers to the sidelobe suppression ratio. The sidelobe suppression ratio is the ratio of the radiation intensity of the first sidelobe in the antenna radiation pattern to the maximum radiation intensity of the main lobe.

[0064] The beam width in the embodiment of the present invention mainly refers to the half-power beam width. The half-power beam width refers to the angle between two points where the power flux density relative to the maximum radiation direction drops to half.

[0065] Step 103, combining the wave-transmitting rainproof coating parameters and the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression into a feature set;

[0066] Further applied to the embodiments of the present invention, the wave-transmitting and rain-proof coating parameters and the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression can be combined into a feature set, that is, a certain feature set includes the wave-transmitting and rain-proof coating parameters and / or the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression. Specifically, a certain feature set s=(material type 1, thickness 1, resistance 1, omnidirectional gain 1, directional beam gain 1, half-power beam width, sidelobe suppression ratio 1).

[0067] Step 104, converting the feature set into a level parameter and sorting the level parameters to obtain a feature correlation parameter;

[0068] Specifically applied to the embodiment of the present invention, the feature set can also be converted into a level parameter and sorted to obtain a feature correlation parameter;

[0069] Specifically, the feature set is converted into a level parameter and sorted to obtain a feature correlation parameter, including: calculating the average value of each data in the feature set and sorting it, and judging whether the average value meets the preset threshold range; if the average value meets the preset threshold range, obtaining the level parameter corresponding to the preset threshold range; and converting the level parameter and the data in the corresponding feature set into a feature correlation parameter.

[0070] In actual application to the embodiments of the present invention, first, the average value of each data can be calculated and sorted, and then the average value can be divided into different levels according to the preset threshold range, and the corresponding level parameters and the data in the corresponding feature set are converted into feature correlation parameters; for example, the feature set s=(material type 1, thickness 1, resistance 1, omnidirectional gain 1, directional beam gain 1, half-power beam width, sidelobe suppression ratio 1), the calculated mean value is x1, and the x1 meets the preset threshold range T1, and the level parameter corresponding to the preset threshold range T1 is G1. The level parameter G1 is combined with the feature set s to obtain the feature correlation parameter (level parameter G1, material type 1, thickness 1, resistance 1, omnidirectional gain 1, directional beam gain 1, half-power beam width, sidelobe suppression ratio 1).

[0071] That is, the feature correlation parameter refers to the parameter set including the level parameter and the sub-data in the feature set.

[0072] Step 105, selecting a feature correlation sub-parameter of the feature correlation parameter by a bidirectional search method;

[0073] Further applied to the embodiments of the present invention, feature correlation sub-parameters of the feature correlation parameter may be selected by bidirectional search;

[0074] In a specific application, the feature correlation sub-parameters of the feature correlation parameter are selected by bidirectional search, including:

[0075] Starting from the empty set of feature correlation parameters by sequential forward selection, a certain feature correlation parameter is selected each time to be added to the feature subset of feature correlation parameters so that the evaluation function reaches the optimal value; the empty set refers to the feature set with missing parameters;

[0076] Starting from the full set of feature correlation parameters by sequential backward selection, a feature correlation sub-parameter is removed from the full set of feature correlation parameters each time, so that the evaluation function value reaches the optimal value after the feature correlation sub-parameter is removed, and the evaluation function value reaches the optimal value at the same time, and the feature correlation sub-parameter is selected; the full set refers to the feature set in which the parameters are not missing. In the embodiment of the present invention, by reducing the computational requirements, reducing the time consumption in the feature selection process and improving the prediction accuracy, data can be effectively screened and the model training cost can be reduced.

[0077] Step 106, inputting the feature correlation sub-parameter into the screening model to obtain the screened feature correlation sub-parameter;

[0078] In an embodiment of the present invention, after obtaining the feature correlation sub-parameter, the feature correlation sub-parameter is input into the screening model to obtain the screened feature correlation sub-parameter. The screening model may include a GBDT model. The GBDT (Gradient Boosting Decision Tree) model may include multiple learners.

[0079] In actual application to the embodiment of the present invention, the step of inputting the feature correlation sub-parameter into the screening model to obtain the screened feature correlation sub-parameter includes:

[0080] Establish weak learners and loss functions for regression tree models;

[0081] The feature correlation sub-parameter is input into the weak learner, the weak learner is initialized, each regression tree model is iteratively optimized by gradient descent, and multiple base learners are iteratively generated. In each iteration, the gradient and second-order derivative of the current regression tree model are calculated to obtain the gradient and second-order derivative of the new regression tree model; the negative gradient is calculated, and a new regression tree model is constructed by the gradient and second-order derivative of the new regression tree model. The negative gradient is used as the input of the new regression tree model until the loss function converges to obtain the screened feature correlation sub-parameter.

[0082] In the embodiment of the present invention, the loss function L may include the following expression:

[0083] ;

[0084] Among them, the represents the sample label; represents the input sample value, n represents the number of samples, represents the constructed weak learner, Thickness parameter adjustment coefficient, Indicates the conductive parameter adjustment coefficient; In the embodiment of the present invention, various types of data can be flexibly processed, including continuous values ​​and discrete values. In the case of relatively short parameter adjustment time, the prediction accuracy is relatively high, the loss function used is adjusted for sample data, has strong robustness to abnormal values, and improves the effectiveness of output sample data.

[0085] Step 107, inputting the screened feature correlation sub-parameters into the optimization model to obtain a trained optimization model;

[0086] After obtaining the screened feature correlation sub-parameters with high effectiveness, the screened feature correlation sub-parameters are input into the optimization model to obtain the trained optimization model; specifically, the screened feature correlation sub-parameters are input into the optimization model to obtain the trained optimization model, including: constructing multiple decision tree models, and forming an optimization model through the decision tree models; at each splitting node of each decision tree model, randomly selecting feature data for segmentation according to the screened feature correlation sub-parameters until the function converges to obtain an initial prediction result; analyzing the initial prediction result through the improved voting operation to complete the training of the optimization model.

[0087] In an embodiment of the present invention, the optimization model may include a random forest algorithm model. Of course, it may also include other supervised machine learning algorithm models, such as a K-nearest neighbor algorithm model, a naive Bayes algorithm model, and a logistic regression algorithm model. The embodiment of the present invention does not impose too many restrictions on this.

[0088] Specifically applied to the embodiment of the present invention, sample features are selected according to feature weights to improve the generalization ability of the model. The feature weights of the random forest algorithm model The expressions can include the following:

[0089] ;

[0090] in, represents the scoring vector of each sample feature of the i-th tree, represents the number of decision tree models, Indicates the antenna gain adjustment factor.

[0091] Step 108, input the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of a plurality of different single antennas into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

[0092] The wave-transmitting and rainproof coating parameters and conductive coating parameters of multiple different single antennas are input into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression, including:

[0093] Inputting the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas into multiple decision tree models constructed by the optimization model;

[0094] At each splitting node of each decision tree model, feature data is randomly selected to segment the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas until the function converges to obtain an initial prediction result;

[0095] The initial prediction results are analyzed through the improved voting operation to obtain the final omnidirectional gain, directional beam gain, beam width and sidelobe suppression; the probability of model overfitting is reduced and the prediction accuracy is improved.

[0096] In an embodiment of the present invention, the optimization control method of the antenna array includes: obtaining the wave-transmitting rainproof coating parameters and the conductive coating parameters of the single antenna in the antenna array; obtaining the omnidirectional gain, directional beam gain, beam width and sidelobe suppression of the antenna array corresponding to the wave-transmitting rainproof coating parameters and the conductive coating parameters; merging the wave-transmitting rainproof coating parameters and the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression into a feature set; converting the feature set into a level parameter and sorting them to obtain a feature correlation parameter; selecting a feature correlation sub-parameter of the feature correlation parameter by a two-way search method; inputting the feature correlation sub-parameter into the screening model to obtain the screened feature correlation sub-parameter; inputting the screened feature correlation sub-parameter into the optimization model to obtain a trained optimization model; inputting the wave-transmitting rainproof coating parameters and the conductive coating parameters of multiple different single antennas into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression. It provides a new simulation approach for selecting the coating material and properties of antenna monomers, improves the optimization control capability of antenna arrays, improves the accuracy of antenna material optimization selection, saves test time and reduces costs.

[0097] It should be noted that, for the method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that this embodiment is not limited by the order of the actions described, because according to this embodiment, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by this embodiment.

[0098] Reference Figure 2 , shows a structural block diagram of an embodiment of an antenna array optimization control system of this embodiment, which may specifically include the following modules:

[0099] The first acquisition module 301 is used to acquire the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the single antenna in the antenna array;

[0100] The second acquisition module 302 is used to obtain the omnidirectional gain, directional beam gain, beam width and sidelobe suppression of the antenna array corresponding to the wave-transmitting and rain-proof coating parameters and the conductive coating parameters;

[0101] A merging module 303 is used to merge the wave-transmitting and rain-proof coating parameters and the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression into a feature set;

[0102] A sorting module 304 is used to convert the feature set into a level parameter and sort it to obtain a feature correlation parameter;

[0103] A selection module 305, for selecting a feature correlation sub-parameter of a feature correlation parameter by a bidirectional search method;

[0104] A screening module 306, configured to input the feature correlation sub-parameters into the screening model to obtain screened feature correlation sub-parameters;

[0105] A training module 307, used for inputting the screened feature correlation sub-parameters into the optimization model to obtain a trained optimization model;

[0106] The output module 308 is used to input the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of multiple different single antennas into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

[0107] Preferably, the sorting module comprises:

[0108] A sorting submodule is used to calculate the average value of each data in the feature set and sort it, and determine whether the average value meets a preset threshold range;

[0109] An acquisition submodule, used for acquiring a level parameter corresponding to a preset threshold range if the average value meets the preset threshold range;

[0110] The conversion submodule is used to convert the level parameter and the data in the corresponding feature set into feature correlation parameters.

[0111] Preferably, the selection module includes:

[0112] The first selection submodule is used to start from an empty set of feature correlation parameters by a sequence forward selection method, and each time select a certain feature correlation parameter to add to the feature subset of the feature correlation parameters, so that the evaluation function reaches the optimal value;

[0113] The second selection submodule is used to start from the full set of feature correlation parameters through a sequential backward selection method, and each time eliminate a feature correlation sub-parameter from the full set of feature correlation parameters, so that the evaluation function reaches the optimal value after eliminating the feature correlation sub-parameter, and the feature correlation sub-parameter is selected.

[0114] Preferably, the screening module comprises:

[0115] Establish submodules to build weak learners and loss functions of regression tree models;

[0116] The screening submodule is used to input the feature correlation sub-parameters into the weak learner, initialize the weak learner, iteratively optimize each regression tree model by gradient descent, iteratively generate multiple base learners, and in each iteration, calculate the gradient and second-order derivative of the current regression tree model to obtain the gradient and second-order derivative of the new regression tree model; construct a new regression tree model by the gradient and second-order derivative of the new regression tree model until the loss function converges to obtain the screened feature correlation sub-parameters.

[0117] Preferably, the training module includes:

[0118] A first construction submodule is used to construct a plurality of decision tree models, and an optimization model is formed by the decision tree models;

[0119] The first segmentation submodule is used to randomly select feature data at each splitting node of each decision tree model for segmentation according to the filtered feature correlation sub-parameters until the function converges to obtain an initial prediction result;

[0120] The first analysis submodule is used to analyze the initial prediction results through the improved voting operation to complete the training of the optimization model.

[0121] Preferably, the output module comprises:

[0122] A second construction submodule is used to input the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas into multiple decision tree models constructed by the optimization model;

[0123] The second segmentation submodule is used to randomly select feature data at each splitting node of each decision tree model to segment the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas until the function converges to obtain an initial prediction result;

[0124] The second analysis submodule is used to analyze the initial prediction result through the improved voting operation to obtain the final omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

[0125] As for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0126] For the specific definition of the optimization control system of the antenna array, please refer to the definition of the optimization control method of the antenna array above, which will not be repeated here. Each module in the optimization control system of the above antenna array can be implemented in whole or in part by software, hardware and a combination thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0127] The above-mentioned optimization control system of the antenna array can be used to execute the optimization control method of the antenna array provided in any of the above-mentioned embodiments, and has corresponding functions and beneficial effects.

[0128] In one embodiment, a computer device is provided, whose internal structure diagram can be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for simulating daylighting is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0129] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0130] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0131] Obtain the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the single antenna in the antenna array;

[0132] Obtaining the omnidirectional gain, directional beam gain, beam width and sidelobe suppression of the antenna array corresponding to the wave-transmitting and rain-proof coating parameters and the conductive coating parameters;

[0133] The wave-transmitting and rainproof coating parameters and the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression are combined into a feature set;

[0134] Converting the feature set into a level parameter and sorting the level parameters to obtain a feature correlation parameter;

[0135] Selecting feature correlation sub-parameters of feature correlation parameters by bidirectional search;

[0136] Inputting the feature correlation sub-parameter into the screening model to obtain the screened feature correlation sub-parameter;

[0137] Inputting the screened feature correlation sub-parameters into the optimization model to obtain a trained optimization model;

[0138] The wave-transmitting and rain-proof coating parameters and conductive coating parameters of multiple different single antennas are input into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

[0139] Preferably, converting the feature set into a level parameter and sorting the level parameters to obtain a feature correlation parameter includes:

[0140] Calculate the average value of each data in the feature set and sort them, and determine whether the average value meets the preset threshold range;

[0141] If the average value meets the preset threshold range, the level parameter corresponding to the preset threshold range is obtained;

[0142] The level parameter and the data in the corresponding feature set are converted into feature correlation parameters.

[0143] Preferably, the step of selecting the feature correlation sub-parameters of the feature correlation parameter by bidirectional search includes:

[0144] Starting from the empty set of feature correlation parameters through the sequence forward selection method, a certain feature correlation parameter is selected each time to be added to the feature subset of the feature correlation parameters so that the evaluation function reaches the optimal value;

[0145] Starting from the full set of feature correlation parameters through sequential backward selection, a feature correlation sub-parameter is eliminated from the full set of feature correlation parameters each time, so that the evaluation function reaches the optimal value after eliminating the feature correlation sub-parameter, and the feature correlation sub-parameter is selected.

[0146] Preferably, the inputting the feature correlation sub-parameters into the screening model to obtain the screened feature correlation sub-parameters comprises:

[0147] Establish the weak learner and loss function of the CART regression tree model;

[0148] The feature correlation sub-parameter is input into the weak learner, the weak learner is initialized, each regression tree model is iteratively optimized by gradient descent, and multiple base learners are iteratively generated. In each iteration, the gradient and second-order derivative of the current regression tree model are calculated to obtain the gradient and second-order derivative of the new regression tree model; a new regression tree model is constructed by the gradient and second-order derivative of the new regression tree model until the loss function converges to obtain the screened feature correlation sub-parameter.

[0149] Preferably, the step of inputting the screened feature correlation sub-parameters into an optimization model to obtain a trained optimization model comprises:

[0150] Constructing multiple decision tree models, and forming an optimization model through the decision tree models;

[0151] At each splitting node of each decision tree model, feature data is randomly selected for segmentation according to the selected feature correlation sub-parameters until the function converges to obtain the initial prediction result;

[0152] The initial prediction results are analyzed through the improved voting operation to complete the training of the optimization model.

[0153] Preferably, the step of inputting the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression comprises: inputting the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas into multiple decision tree models constructed by the optimization model;

[0154] At each splitting node of each decision tree model, feature data is randomly selected to segment the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas until the function converges to obtain an initial prediction result;

[0155] The initial prediction results are analyzed through the improved voting operation to obtain the final omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

[0156] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented;

[0157] Obtain the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the single antenna in the antenna array;

[0158] Obtaining the omnidirectional gain, directional beam gain, beam width and sidelobe suppression of the antenna array corresponding to the wave-transmitting and rain-proof coating parameters and the conductive coating parameters;

[0159] The wave-transmitting and rainproof coating parameters and the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression are combined into a feature set;

[0160] Converting the feature set into a level parameter and sorting the level parameters to obtain a feature correlation parameter;

[0161] Selecting feature correlation sub-parameters of feature correlation parameters by bidirectional search;

[0162] Inputting the feature correlation sub-parameter into the screening model to obtain the screened feature correlation sub-parameter;

[0163] Inputting the screened feature correlation sub-parameters into the optimization model to obtain a trained optimization model;

[0164] The wave-transmitting and rain-proof coating parameters and conductive coating parameters of multiple different single antennas are input into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

[0165] Preferably, converting the feature set into a level parameter and sorting the level parameters to obtain a feature correlation parameter includes:

[0166] Calculate the average value of each data in the feature set and sort them, and determine whether the average value meets the preset threshold range;

[0167] If the average value meets the preset threshold range, the level parameter corresponding to the preset threshold range is obtained;

[0168] The level parameter and the data in the corresponding feature set are converted into feature correlation parameters.

[0169] Preferably, the step of selecting the feature correlation sub-parameters of the feature correlation parameter by bidirectional search includes:

[0170] Starting from the empty set of feature correlation parameters through the sequence forward selection method, a certain feature correlation parameter is selected each time to be added to the feature subset of the feature correlation parameters so that the evaluation function reaches the optimal value;

[0171] Starting from the full set of feature correlation parameters through sequential backward selection, a feature correlation sub-parameter is eliminated from the full set of feature correlation parameters each time, so that the evaluation function reaches the optimal value after eliminating the feature correlation sub-parameter, and the feature correlation sub-parameter is selected.

[0172] Preferably, the inputting the feature correlation sub-parameters into the screening model to obtain the screened feature correlation sub-parameters comprises:

[0173] Establish the weak learner and loss function of the CART regression tree model;

[0174] The feature correlation sub-parameter is input into the weak learner, the weak learner is initialized, each regression tree model is iteratively optimized by gradient descent, and multiple base learners are iteratively generated. In each iteration, the gradient and second-order derivative of the current regression tree model are calculated to obtain the gradient and second-order derivative of the new regression tree model; a new regression tree model is constructed by the gradient and second-order derivative of the new regression tree model until the loss function converges to obtain the screened feature correlation sub-parameter.

[0175] Preferably, the step of inputting the screened feature correlation sub-parameters into an optimization model to obtain a trained optimization model comprises:

[0176] Constructing multiple decision tree models, and forming an optimization model through the decision tree models;

[0177] At each splitting node of each decision tree model, feature data is randomly selected for segmentation according to the selected feature correlation sub-parameters until the function converges to obtain the initial prediction result;

[0178] The initial prediction results are analyzed through the improved voting operation to complete the training of the optimization model.

[0179] Preferably, the step of inputting the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression comprises: inputting the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas into multiple decision tree models constructed by the optimization model;

[0180] At each splitting node of each decision tree model, feature data is randomly selected to segment the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas until the function converges to obtain an initial prediction result;

[0181] The initial prediction results are analyzed through the improved voting operation to obtain the final omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

[0182] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0183] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0184] Those skilled in the art will appreciate that the embodiments of the present embodiment may be provided as methods, devices, or computer program products. Therefore, the present embodiment may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present embodiment may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0185] The present embodiment is described with reference to the flowchart and / or block diagram of the apparatus, terminal device (system), and computer program product according to the present embodiment. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0186] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction method, which is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0187] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable terminal device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0188] Although the preferred embodiment of the present embodiment has been described, other changes and modifications may be made to these embodiments once those skilled in the art have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiment and all changes and modifications falling within the scope of the present embodiment.

[0189] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, device, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a..." do not exclude the existence of other identical elements in the process, device, article or terminal device including the elements.

[0190] The above is a detailed introduction to an antenna array optimization control method and an antenna array optimization control system, a computer device and a storage medium provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. An optimization control method for an antenna array, characterized in that: include: Obtain the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the single antenna in the antenna array; Obtaining the omnidirectional gain, directional beam gain, beam width and sidelobe suppression of the antenna array corresponding to the wave-transmitting and rain-proof coating parameters and the conductive coating parameters; The wave-transmitting and rainproof coating parameters and the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression are combined into a feature set; Converting the feature set into a level parameter and sorting the level parameters to obtain a feature correlation parameter; Selecting feature correlation sub-parameters of feature correlation parameters by bidirectional search; Inputting the feature correlation sub-parameters into a screening model to obtain screened feature correlation sub-parameters; Inputting the screened feature correlation sub-parameters into the optimization model to obtain a trained optimization model; The wave-transmitting and rain-proof coating parameters and conductive coating parameters of multiple different single antennas are input into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

2. The method according to claim 1, characterized in that: The step of converting the feature set into a level parameter and sorting the level parameters to obtain a feature correlation parameter includes: Calculate the average value of each data in the feature set and sort them, and determine whether the average value meets the preset threshold range; If the average value meets the preset threshold range, the level parameter corresponding to the preset threshold range is obtained; The level parameter and the data in the corresponding feature set are converted into feature correlation parameters.

3. The method according to claim 1, characterized in that: The step of selecting the feature correlation sub-parameters of the feature correlation parameter by bidirectional search includes: Starting from the empty set of feature correlation parameters through the sequence forward selection method, a certain feature correlation parameter is selected each time to be added to the feature subset of the feature correlation parameters so that the evaluation function reaches the optimal value; Starting from the full set of feature correlation parameters through sequential backward selection, a feature correlation sub-parameter is eliminated from the full set of feature correlation parameters each time, so that the evaluation function reaches the optimal value after eliminating the feature correlation sub-parameter, and the feature correlation sub-parameter is selected.

4. The method according to claim 1, characterized in that The step of inputting the feature correlation sub-parameter into a screening model to obtain the screened feature correlation sub-parameter comprises: Establish the weak learner and loss function of the CART regression tree model; The feature correlation sub-parameter is input into the weak learner, the weak learner is initialized, each regression tree model is iteratively optimized by gradient descent, and multiple base learners are iteratively generated. In each iteration, the gradient and second-order derivative of the current regression tree model are calculated to obtain the gradient and second-order derivative of the new regression tree model; a new regression tree model is constructed by the gradient and second-order derivative of the new regression tree model until the loss function converges to obtain the screened feature correlation sub-parameter.

5. The method according to claim 4, characterized in that The step of inputting the screened feature correlation sub-parameters into the optimization model to obtain the trained optimization model comprises: Constructing multiple decision tree models, and forming an optimization model through the decision tree models; At each splitting node of each decision tree model, feature data is randomly selected for segmentation according to the selected feature correlation sub-parameters until the function converges to obtain the initial prediction result; The initial prediction results are analyzed through the improved voting operation to complete the training of the optimization model.

6. The method according to claim 4, characterized in that The step of inputting the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression comprises: inputting the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas into multiple decision tree models constructed by the optimization model; At each splitting node of each decision tree model, feature data is randomly selected to segment the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the multiple different single antennas until the function converges to obtain an initial prediction result; The initial prediction results are analyzed through the improved voting operation to obtain the final omnidirectional gain, directional beam gain, beamwidth and sidelobe suppression.

7. An optimization control system for an antenna array, characterized in that: include: The first acquisition module is used to acquire the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of the single antenna in the antenna array; A second acquisition module is used to obtain the omnidirectional gain, directional beam gain, beam width and sidelobe suppression of the antenna array corresponding to the wave-transmitting and rain-proof coating parameters and the conductive coating parameters; A merging module, used to merge the wave-transmitting and rain-proof coating parameters and the conductive coating parameters and the corresponding omnidirectional gain, directional beam gain, beam width and sidelobe suppression into a feature set; A sorting module, used for converting the feature set into a level parameter and sorting the level parameters to obtain a feature correlation parameter; A selection module, used for selecting a feature correlation sub-parameter of a feature correlation parameter by a bidirectional search method; A screening module, used for inputting the feature correlation sub-parameters into a screening model to obtain screened feature correlation sub-parameters; A training module, used for inputting the screened feature correlation sub-parameters into the optimization model to obtain a trained optimization model; The output module is used to input the wave-transmitting and rain-proof coating parameters and the conductive coating parameters of multiple different single antennas into the trained optimization model to obtain the output omnidirectional gain, directional beam gain, beam width and sidelobe suppression.

8. The system according to claim 7, characterized in that The sorting module comprises: A sorting submodule is used to calculate the average value of each data in the feature set and sort it, and determine whether the average value meets a preset threshold range; An acquisition submodule, used for acquiring a level parameter corresponding to a preset threshold range if the average value meets the preset threshold range; The conversion submodule is used to convert the level parameter and the data in the corresponding feature set into feature correlation parameters.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the antenna array optimization control method according to any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the antenna array optimization control method according to any one of claims 1 to 6 are implemented.

Images