Large-scale phased array surface temperature measurement point rarefaction method based on genetic algorithm

Abstract

The invention discloses a large-scale phased array surface temperature measurement point rarefaction method based on a genetic algorithm, and the method comprises the following steps: S1, solving a phased array surface temperature field through a finite element method, and enabling the temperature field to serve as a data source for collecting temperature information of temperature measurement points and a standard for reconstructing the temperature field; s2, performing first-step sparsification on the temperature measurement points, and taking the temperature measurement points as initial values for position optimization of the temperature measurement points; s3, the positions of the temperature measurement points are optimized through a genetic algorithm, so that the precision of the reconstructed temperature field is higher; and S4, performing second-step sparsification on the temperature measurement points after position optimization, removing the temperature measurement points which have small influence on reconstruction precision, and realizing high-precision reconstruction of the temperature field with fewer temperature measurement points. According to the method, the layout positions of the array surface temperature measurement points of the large-scale active phased array are optimized, the number of the temperature measurement points is reduced, and the aim of reconstructing the array surface temperature field with less temperature measurement points and high precision is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of antennas, and in particular to a method for sparsely determining temperature measurement points on a large-scale phased array surface based on a genetic algorithm. Background Art

[0002] Large-scale active phased array antennas are widely used in space communications, navigation, remote sensing, electronic reconnaissance and other fields. They have the advantages of high gain, low sidelobe level, controllable beam, and high communication rate, and are important equipment in the field of modern communications. However, when large-scale active phased array antennas work for a long time, the heat power consumption generated by the T / R components will cause the temperature of the antenna array surface to change, and the array surface will produce thermal deformation, resulting in the displacement of the antenna array element position and the change of the antenna phase difference, thereby causing the deterioration of the antenna radiation performance, such as the main lobe beam pointing deviation, 3dB beam width widening, and sidelobe level increase.

[0003] An important method to improve the radiation performance of large phased array antennas is to monitor the temperature information of the phased array surface by placing temperature measurement points on the array surface, calculate the position offset of the array elements, and compensate the antenna beam accordingly. Therefore, the accurate description of the temperature field of the large-scale active phased array surface is the basis for accurate compensation of the phased array beam performance. How to accurately reconstruct the temperature field of the array surface is of great significance to the regulation and compensation of the antenna electrical performance.

[0004] At present, the temperature field reconstruction method of large-scale mechanical and electronic equipment is to deploy a network of temperature measurement points on the equipment and reconstruct the temperature field by interpolation based on the detected temperature information. The existing temperature measurement point sparseness and position optimization methods are mostly used for mechanical and electronic equipment such as parabolic antennas and radio telescopes. There is a lack of temperature measurement point optimization methods for large-scale active phased arrays. In addition, the existing interpolation methods are not suitable for accurately describing the temperature characteristics of T / R components and cannot accurately reconstruct the temperature characteristics of the phased array array surface. At the same time, under the limitations of weight load and equipment cost, sparse number of temperature measurement points and high-precision reconstruction of the array surface temperature field using limited temperature measurement points are also a problem that needs to be solved urgently. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a method for sparsely configuring temperature measurement points on a large-scale phased array based on a genetic algorithm, optimize the layout of temperature measurement points on a large-scale active phased array, sparsely configure the number of temperature measurement points, and achieve the purpose of reconstructing the temperature field on the array surface with high precision using fewer temperature measurement points. In order to achieve the above-mentioned purpose and other advantages of the present invention, a method for sparsely configuring temperature measurement points on a large-scale phased array based on a genetic algorithm is provided, comprising:

[0006] S1. Solving the temperature field of the phased array surface by finite element method, the temperature field is used as the data source for collecting temperature information at the temperature measurement points and the standard for reconstructing the temperature field;

[0007] S2, perform the first step of sparseness on the temperature measurement points and use them as the initial value for optimizing the position of the temperature measurement points;

[0008] S3, optimize the temperature measurement point position through genetic algorithm to make the reconstructed temperature field more accurate;

[0009] S4. Perform the second step of sparseness on the temperature measurement points after position optimization, remove the temperature measurement points that have little impact on the reconstruction accuracy, and achieve high-precision reconstruction of the temperature field with fewer temperature measurement points.

[0010] Preferably, in step S2, the number and arrangement of the temperature measuring points are non-dominatedly sorted, and a temperature measuring point arrangement with a small number of temperature measuring points and a better temperature field reconstruction effect is selected to achieve the first step of sparseness of the temperature measuring points. The specific steps are:

[0011] S21, discretizing the phased array surface into grid points, using the grid points to represent the position on the surface, and dividing the grid points into two parts: grid points on the T / R components and grid points at the gaps between the T / R components according to the positions of the T / R components on the surface;

[0012] S22, evenly arrange different numbers of temperature measuring points on the array surface and calculate the error of the reconstructed temperature field. The upper limit of the number of temperature measuring points is the number of T / R components on the array surface. The positions of the temperature measuring points are taken at the grid points on the T / R components. They are evenly arranged on the array surface. The array surface temperature field is reconstructed by interpolation combined with heat transfer theory, and the root mean square error between the reconstructed temperature field and the original temperature field is calculated.

[0013] S23. Perform a non-dominated sorting of the uniform arrangement according to the number of temperature measurement points and the root mean square error, find out the points on the Pareto front in the non-dominated sorting, take the arrangement and number of temperature measurement points at the inflection point of the Pareto front curve as the initial value for optimizing the position of the temperature measurement points, instead of arranging the temperature measurement points on all T / R components, to complete the first step of sparseness of the temperature measurement points.

[0014] Preferably, the method of reconstructing the array temperature field by interpolation in combination with the heat transfer theory in step S22 specifically includes: obtaining the temperature of all T / R components on the array surface by using the inverse distance interpolation according to the temperature measurement point data; calculating the temperature at the gap between the T / R components according to the heat transfer theory based on the temperature of the T / R components obtained by interpolation, and obtaining a temperature value along the x and y directions respectively. Since the structure of a single T / R component is symmetrical about x and y, eight temperature values ​​of the gap around the single T / R component can be obtained according to the symmetry;

[0015] Based on all the above temperature information, the temperature of the entire array is obtained by inverse interpolation using a smaller search radius.

[0016] Preferably, in step S3, the accuracy of reconstructing the temperature field is used as the reconstruction objective function, and a genetic algorithm is used to optimize the positions of the temperature measurement points based on the initial arrangement of the temperature measurement points. The specific steps are as follows:

[0017] S31. Generate a temperature measurement point position group according to the number of temperature measurement points in the initial value. The temperature measurement point position is represented by the grid number of the discrete grid points on the phased array. When the temperature measurement point position information is generated, an error correction mechanism is used. When the generated temperature measurement point position is on the T / R component gap, the temperature measurement point position is changed to the T / R component with the closest spatial distance. If there are multiple temperature measurement points on the same T / R component, one of them will be retained, and the remaining temperature measurement points will be arranged on the T / R component without temperature measurement points. Generate a corresponding number of temperature measurement point position groups according to the set population size. In order to ensure the diversity of the population and improve the solution efficiency of the genetic algorithm, take 10% of the individuals in the population and round them down to evenly arrange the temperature measurement points on the array surface T / R components.

[0018] S32, reconstructing the temperature field of the phased array surface using an interpolation reconstruction method combined with heat transfer according to the temperature information collected by the temperature measurement point position group, and calculating the fitness value of each temperature measurement point group according to the objective function expression;

[0019] S33, by using the tournament selection method, the temperature measurement point group is divided into several groups, the number of temperature measurement point groups in each group is the same, and the temperature measurement point group with the largest fitness value in each group is selected to enter the temperature measurement point position group for the next iteration solution;

[0020] S34, generating new temperature measurement point location information by simulating the temperature measurement point location data screened out by binary cross processing;

[0021] S35, using polynomial mutation to generate new temperature measurement point position information based on the simulation of binary crossover;

[0022] S36. A new temperature measurement point group is formed by combining the screened temperature measurement point group and the newly generated temperature measurement point group, and then the fitness value is calculated. The above process is iterated until the fitness value converges. At this time, the layout position of the temperature measurement point group with the largest fitness value is the optimal.

[0023] Preferably, the specific steps of step S4 are as follows:

[0024] S41, setting the maximum allowable reconstruction accuracy variation range ΔRMS;

[0025] S42, traversing and deleting each temperature measurement point in the optimal solution, and calculating the reconstruction accuracy after deleting the point;

[0026] S43, selecting the temperature measurement point arrangement corresponding to the minimum reconstruction accuracy change, and determining whether the reconstruction accuracy change exceeds ΔRMS;

[0027] S44, repeat S42 to S43 until the reconstruction accuracy changes beyond ΔRMS, and take the last temperature measurement point arrangement as the optimal temperature measurement point arrangement after sparseness.

[0028] Compared with the prior art, the beneficial effect of the present invention is as follows: by forming an optimization model with the array surface temperature field reconstruction error as the optimization target, and then using a genetic algorithm suitable for solving such optimization problems to solve it, the scheme of the present invention can achieve higher temperature field reconstruction accuracy with fewer temperature measurement points, and provide a new optional scheme for exploring the phased array temperature measurement point layout method and energy saving and consumption reduction. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a flow chart of a method for sparse and position optimization of temperature measurement points on a large-scale phased array based on a genetic algorithm according to the present invention;

[0030] Figure 2 A schematic diagram of a phased array structure according to a method for sparse and position optimization of temperature measurement points on a large-scale phased array surface based on a genetic algorithm according to the present invention;

[0031] Figure 3 It is a schematic diagram of the arrangement of T / R components on the phased array surface according to the method for sparse and position optimization of temperature measurement points on the large-scale phased array surface based on genetic algorithm of the present invention;

[0032] Figure 4 It is a schematic diagram of the distance inverse interpolation method in step S2 of the method for sparse and position optimization of large-scale phased array array surface temperature measurement points based on genetic algorithm according to the present invention;

[0033] Figure 5 It is a schematic diagram of calculating the temperature at the gap of the T / R component in combination with the heat transfer theory in step S2 of the method for sparse and position optimization of temperature measurement points on a large-scale phased array based on a genetic algorithm according to the present invention;

[0034] Figure 6 The genetic algorithm solution process in step S3 of the method for sparse and position optimization of large-scale phased array surface temperature measurement points based on genetic algorithm according to the present invention;

[0035] Figure 7 The layout positions of temperature measurement points before and after optimization of the phased array antenna according to the large-scale phased array array surface temperature measurement point sparseization and position optimization method based on genetic algorithm of the present invention. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] Reference Figure 1 A method for sparsely determining temperature measurement points on a large-scale phased array based on a genetic algorithm comprises the following steps:

[0038] Step S1: Based on the phased array structure and its T / R component heat consumption, the finite element method is used to accurately solve the temperature field of the phased array surface, and the temperature field is used as the data source for collecting temperature information at the temperature measurement points and the standard for reconstructing the temperature field;

[0039] like Figure 2 As shown, the phased array structure used in the example has three layers from top to bottom, namely the dielectric layer, the T / R component and the alloy base plate, and the temperature measurement points will be arranged on the T / R component.

[0040] Step S2: non-dominated sorting of the number and arrangement of the temperature measurement points is performed to achieve the first step of sparseness of the temperature measurement points, and to select appropriate initial arrangement positions and numbers of the temperature measurement points;

[0041] S21, discretize the phased array surface into grid points, use the grid points to represent the position on the array surface, and divide the grid points into two parts: grid points on the T / R components and grid points at the gap between the T / R components;

[0042] like Figure 3 As shown in the figure, the phased array has a total of 10×10 T / R components, each T / R component is divided into 39×39 grids, and the T / R component gap is divided into 3×39 grids. The discretized grid of the entire array consists of these two parts, totaling 420×420 grids.

[0043] S22, evenly arrange different numbers of temperature measurement points on the array surface and calculate the error of the reconstructed temperature field. The upper limit of the number of temperature measurement points is the number of T / R components on the array surface. The positions of the temperature measurement points are taken at the grid points on the T / R components. They are evenly arranged on the array surface. The array surface temperature field is reconstructed by interpolation combined with heat transfer theory, and the root mean square error between the reconstructed temperature field and the original temperature field is calculated.

[0044] The expression of the root mean square error between the reconstructed temperature field and the original temperature field is:

[0045]

[0046] Where, T c,i is the temperature value of the reconstructed temperature field, T f,iis the temperature value of the original temperature field, and n is the number of grid points on the array surface;

[0047] According to the temperature data collected at the temperature measurement point, the temperature of all T / R components on the array surface is obtained by using distance inverse interpolation. The mathematical expression of the distance inverse interpolation is:

[0048]

[0049] Where n is the number of temperature measurement points within the search radius, x is the spatial coordinate of the point to be interpolated, and x i (i=1,2,…,n) is the spatial coordinate of the i-th temperature measurement point within the search radius R, d i =‖xx i ‖ is the spatial distance from the interpolation point to the i-th temperature measurement point within the search radius R, and k is the power parameter, such as Figure 4 As shown, a schematic diagram of distance inverse interpolation calculation is given;

[0050] According to the temperature of the T / R component obtained by interpolation, the temperature at the gap of the T / R component is calculated according to the heat transfer theory, and a temperature value is calculated along the x and y directions. Since the structure of a single T / R component is symmetrical about x and y, eight temperature values ​​of the gap around a single T / R component can be obtained according to the symmetry. The expression for calculating the temperature at the gap of the T / R component according to the heat transfer theory is:

[0051]

[0052] In the formula, is the heat consumption rate of the heat source, λ heat is the heat transfer coefficient, δ i is the distance between the gap between the T / R components and the temperature measurement point on the T / R, such as Figure 5 As shown, a schematic diagram of the heat transfer theory for calculating the temperature at the gap of the T / R component is given;

[0053] Based on all the above temperature information, the temperature of the entire array is obtained by inverse interpolation using a smaller search radius.

[0054] S23, performing non-dominated sorting on the uniform arrangement according to the number of temperature measurement points and the root mean square error, finding the points on the Pareto front in the non-dominated sorting, taking the arrangement and number of temperature measurement points at the inflection point of the Pareto front curve as the initial value for optimizing the position of the temperature measurement points, replacing the form of arranging temperature measurement points on all T / R components, and completing the first step of sparseness of the temperature measurement points;

[0055] Step S3: Taking the root mean square error between the reconstructed temperature field and the original temperature field as the objective function, a genetic algorithm is used to optimize the positions of the temperature measurement points based on the initial arrangement of the temperature measurement points, so as to make the reconstructed temperature field more accurate. The flowchart of the genetic algorithm is shown in FIG. Figure 6 As shown;

[0056] S31, based on the number of temperature measurement points in the initial value, generate a temperature measurement point position group. The temperature measurement point position is represented by the grid number of the discrete grid points on the phased array. When generating the temperature measurement point position information, an error correction mechanism is used. When the generated temperature measurement point is located at the gap between the T / R components, the temperature measurement point position is changed to the T / R component with the closest spatial distance. If there are multiple temperature measurement points on the same T / R component, retain one of them and place the remaining temperature measurement points on the T / R component without temperature measurement points. Generate a corresponding number of temperature measurement point position groups based on the set population size. To ensure the diversity of the population and improve the efficiency of the genetic algorithm, take 10% of the individuals in the population (round down) and evenly place temperature measurement points on the T / R components on the array surface.

[0057] S32, reconstructing the temperature field of the phased array surface using the interpolation reconstruction method combined with heat transfer according to the temperature information collected by the temperature measurement point position group, and calculating the fitness value of each temperature measurement point group according to the objective function expression;

[0058] S33, using a tournament selection method, these temperature measurement point groups are divided into a number of groups, the number of temperature measurement point groups in each group is the same, and the temperature measurement point group with the largest fitness value in each group is selected to enter the temperature measurement point position group for the next iteration solution;

[0059] S34, use the simulated binary cross processing to filter out the temperature measurement point location data to generate new temperature measurement point location information. The filtered temperature measurement point location data is exchanged, and the calculation process is as follows:

[0060]

[0061] In the formula, and There are two groups of temperature measurement points selected by the tournament. is the position of the temperature measurement point on the antenna array; and are two newly generated groups of temperature measurement points. is the position of the temperature measurement point on the antenna array. Where β is dynamically and randomly determined by the distribution factor η:

[0062]

[0063] In the formula, rand is a random number between 0 and 1 that follows a uniform distribution, and η is a custom parameter between 0 and 1. The larger η is, the greater the probability that the generated new temperature measurement point group is close to the screened temperature measurement point group.

[0064] S35, using polynomial mutation to generate new temperature measurement point position information based on the simulated binary crossover, the calculation process is as follows:

[0065] v' k =v k +δ×(u k -l k )

[0066]

[0067] In the formula, δ1=(v k -l k ) / (u k -l k ),δ2=(u k -v k ) / (u k -l k ), u is a random number in the interval [0,1], η m Take 20, v k represents the original temperature measurement point group, v′ k Indicates the new temperature measurement point group after the layout location information is modified. k Indicates the lower limit of the temperature measurement point selection range, u k It indicates the upper limit of the temperature measurement point selection range, which is set according to the size of the phased array surface;

[0068] S36, a new temperature measurement point group is formed by the screened temperature measurement point group and the newly generated temperature measurement point group, and then the fitness value is calculated. The above process is iterated until the fitness value converges. At this time, the layout position of the temperature measurement point group with the largest fitness value is optimal. At this time, the root mean square error of the reconstructed temperature field decreases from 0.00247 to 0.00059.

[0069] Step S4: performing the second step of sparseness on the temperature measurement points after position optimization, removing the temperature measurement points that have little influence on the reconstruction accuracy, and reconstructing the temperature field with high accuracy using fewer temperature measurement points;

[0070] S41, setting the maximum allowable reconstruction accuracy variation range ΔRMS;

[0071] S42, traversing and deleting each temperature measurement point in the optimal solution, and calculating the reconstruction accuracy after deleting the point;

[0072] S43, selecting the temperature measurement point arrangement corresponding to the minimum reconstruction accuracy change, and determining whether the reconstruction accuracy change exceeds ΔRMS;

[0073] S44, repeat S42 to S43 until the reconstruction accuracy changes beyond ΔRMS, and take the last temperature measurement point arrangement as the optimal temperature measurement point arrangement after sparseness. The final optimization result is as follows Figure 7 As shown in the figure, the red dots represent the temperature measurement points. The number of temperature measurement points is optimized from 100 to 51. At this time, the reconstruction error of the temperature field increases from 0.00059 to 0.00067.

[0074] In summary, the present invention improves the interpolation method according to the heating characteristics of the phased array heat source T / R component, selects the preliminary number and layout positions of the temperature measuring points according to the non-dominated sorting of the reconstructed temperature field accuracy and the number of temperature measuring points, and establishes an optimization model with the inverse of the root mean square error between the original temperature field and the reconstructed temperature field as the optimization target. Then, a genetic algorithm is used to solve the model, and the solution result is the optimal layout position of the temperature measuring points under the current number of temperature measuring points. The temperature measuring points in the layout position that have little influence on the reconstruction accuracy are deleted, so as to achieve higher temperature field reconstruction accuracy with fewer temperature measuring points.

[0075] The number of devices and processing scales described here are used to simplify the description of the present invention, and the application, modification and variation of the present invention will be obvious to those skilled in the art.

[0076] Although the embodiments of the present invention have been disclosed as above, they are not limited to the applications listed in the specification and implementation modes. They can be fully applied to various fields suitable for the present invention. For those familiar with the art, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the scope of equivalents, the present invention is not limited to the specific details and the illustrations shown and described herein.

Claims

1. A method for sparse temperature measurement points on a large-scale phased array based on a genetic algorithm, characterized in that: The following steps are involved: S1. Solving the temperature field of the phased array surface by finite element method, the temperature field is used as the data source for collecting temperature information at the temperature measurement points and the standard for reconstructing the temperature field; S2, perform the first step of sparseness on the temperature measurement points and use them as the initial value for optimizing the position of the temperature measurement points; S3, optimize the temperature measurement point position through genetic algorithm to make the reconstructed temperature field more accurate; S4. Perform the second step of sparseness on the temperature measurement points after position optimization, remove the temperature measurement points that have little impact on the reconstruction accuracy, and achieve high-precision reconstruction of the temperature field with fewer temperature measurement points.

2. A method for sparsely determining temperature measurement points on a large-scale phased array based on a genetic algorithm as claimed in claim 1, characterized in that: In step S2, the number and arrangement of the temperature measurement points are non-dominatedly sorted, and a temperature measurement point arrangement with a small number of temperature measurement points and a better temperature field reconstruction effect is selected to achieve the first step of sparseness of the temperature measurement points. The specific steps are: S21, discretizing the phased array surface into grid points, using the grid points to represent the position on the surface, and dividing the grid points into two parts: grid points on the T / R components and grid points at the gaps between the T / R components according to the positions of the T / R components on the surface; S22, evenly arrange different numbers of temperature measuring points on the array surface and calculate the error of the reconstructed temperature field. The upper limit of the number of temperature measuring points is the number of T / R components on the array surface. The positions of the temperature measuring points are taken at the grid points on the T / R components. They are evenly arranged on the array surface. The array surface temperature field is reconstructed by interpolation combined with heat transfer theory, and the root mean square error between the reconstructed temperature field and the original temperature field is calculated. S23. Perform a non-dominated sorting of the uniform arrangement according to the number of temperature measurement points and the root mean square error, find out the points on the Pareto front in the non-dominated sorting, take the arrangement and number of temperature measurement points at the inflection point of the Pareto front curve as the initial value for optimizing the position of the temperature measurement points, instead of arranging the temperature measurement points on all T / R components, to complete the first step of sparseness of the temperature measurement points.

3. The method for sparsely determining temperature measurement points on a large-scale phased array based on a genetic algorithm as claimed in claim 2, characterized in that: The method of reconstructing the array temperature field by interpolation in combination with the heat transfer theory in step S22 specifically includes: obtaining the temperature of all T / R components on the array surface by using the inverse distance interpolation according to the temperature measurement point data; calculating the temperature at the gap between the T / R components according to the heat transfer theory based on the temperature of the T / R components obtained by interpolation, and obtaining a temperature value along the x and y directions respectively. Since the structure of a single T / R component is symmetrical about x and y, eight temperature values ​​of the gap around the single T / R component can be obtained according to the symmetry; Based on all the above temperature information, the temperature of the entire array is obtained by inverse interpolation using a smaller search radius.

4. The method for sparsely determining temperature measurement points on a large-scale phased array based on a genetic algorithm as claimed in claim 1, characterized in that: In step S3, the accuracy of reconstructing the temperature field is used as the reconstruction objective function, and the genetic algorithm is used to optimize the positions of the temperature measurement points based on the initial arrangement of the temperature measurement points. The specific steps are as follows: S31. Generate a temperature measurement point position group according to the number of temperature measurement points in the initial value. The temperature measurement point position is represented by the grid number of the discrete grid points on the phased array. When the temperature measurement point position information is generated, an error correction mechanism is used. When the generated temperature measurement point position is on the T / R component gap, the temperature measurement point position is changed to the T / R component with the closest spatial distance. If there are multiple temperature measurement points on the same T / R component, one of them will be retained, and the remaining temperature measurement points will be arranged on the T / R component without temperature measurement points. Generate a corresponding number of temperature measurement point position groups according to the set population size. In order to ensure the diversity of the population and improve the solution efficiency of the genetic algorithm, take 10% of the individuals in the population and round them down to evenly arrange the temperature measurement points on the array surface T / R components. S32, reconstructing the temperature field of the phased array surface using an interpolation reconstruction method combined with heat transfer according to the temperature information collected by the temperature measurement point position group, and calculating the fitness value of each temperature measurement point group according to the objective function expression; S33, by using the tournament selection method, the temperature measurement point group is divided into several groups, the number of temperature measurement point groups in each group is the same, and the temperature measurement point group with the largest fitness value in each group is selected to enter the temperature measurement point position group for the next iteration solution; S34, generating new temperature measurement point location information by simulating the temperature measurement point location data screened out by binary cross processing; S35, using polynomial mutation to generate new temperature measurement point position information based on the simulation of binary crossover; S36. A new temperature measurement point group is formed by combining the screened temperature measurement point group and the newly generated temperature measurement point group, and then the fitness value is calculated. The above process is iterated until the fitness value converges. At this time, the layout position of the temperature measurement point group with the largest fitness value is the optimal.

5. The method for sparsely determining temperature measurement points on a large-scale phased array based on a genetic algorithm as claimed in claim 1, characterized in that: The specific steps of step S4 are as follows: S41, setting the maximum allowable reconstruction accuracy variation range ΔRMS; S42, traversing and deleting each temperature measurement point in the optimal solution, and calculating the reconstruction accuracy after deleting the point; S43, selecting the temperature measurement point arrangement corresponding to the minimum reconstruction accuracy change, and determining whether the reconstruction accuracy change exceeds ΔRMS; S44, repeat S42 to S43 until the reconstruction accuracy changes beyond ΔRMS, and take the last temperature measurement point arrangement as the optimal temperature measurement point arrangement after sparseness.

Images