Electromagnetic environment monitoring optimal point distribution method and system

Abstract

The application provides an electromagnetic environment monitoring optimization distribution method and system, and relates to the technical field of electromagnetic environment monitoring. In view of the low representativeness and precision of the selected monitoring points in the prior art, and the problem that the real environment of the region cannot be reflected, the application divides a region to be monitored into multiple grids, obtains electromagnetic radiation indexes of the grids, and constructs an original data matrix; the high optimal index value is determined according to the optimal index method, the original data matrix is normalized to obtain a decision matrix; the decision matrix is converted to obtain a closeness matrix; the similarity values between the grids are calculated based on the closeness matrix to construct a similarity matrix; the grids are classified based on the similarity matrix to obtain multiple sets of classification results; for each set of classification results, the consistency of the same type of grids and the difference of different types of grids are analyzed to determine a final classification result, and the final representative grid is determined according to the final classification result. The representativeness and precision of the selected final representative grid are high.

Description

Technical Field

[0001] This invention relates to the field of electromagnetic environment monitoring technology, and in particular to an optimized method and system for electromagnetic environment monitoring point deployment. Background Technology

[0002] Electromagnetic environment monitoring refers to the systematic measurement, data collection, and analysis of electromagnetic radiation levels within a specific area to understand the electromagnetic environment quality of that area. Optimized electromagnetic environment monitoring point layout involves using a small number of representative monitoring points to reflect the regional electromagnetic environment quality as reasonably, accurately, and completely as possible.

[0003] Currently, the selection of electromagnetic environment quality monitoring sites in China is mostly based on environmental functional zoning and the distribution of specific electromagnetic pollution sources. This involves setting up monitoring sites near relevant environmental functional zones and specific pollution sources, and conducting regular manual monitoring. However, this approach fails to consider the differences in electromagnetic radiation levels at different locations within the same functional zone, leading to the omission of key monitoring sites.

[0004] Patent CN120146247A discloses a method for optimizing the layout of electromagnetic environment testing points. This method includes determining multiple grid sizes based on building structure information and preset grid information, and dividing the target computer room into multiple grids based on these grid sizes to obtain multiple grid division schemes. Test sampling points are determined in each grid based on the computer room layout information; these sampling points are the points furthest from all electronic devices in the computer room within their respective grids. This indicates that determining test sampling points through grid division and distance filtering does not consider the inherent differences in electromagnetic radiation levels between grids, potentially leading to insufficient sampling points in high-radiation areas and redundant sampling points in low-radiation areas.

[0005] However, existing technologies suffer from issues such as omissions of key monitoring points, insufficient deployment, and redundancy, which result in low representativeness and accuracy of the selected monitoring points, failing to reflect the true environment of the region.

[0006] Therefore, developing an optimized method and system for electromagnetic environment monitoring point selection is of great significance for improving the representativeness and accuracy of the selected monitoring points. Summary of the Invention

[0007] To address the problem that existing technologies often have low representativeness and accuracy in selecting monitoring points, failing to reflect the true environment of a region, this invention proposes an optimized method for electromagnetic environment monitoring point selection, specifically including the following steps:

[0008] S1. The area to be monitored is divided into multiple grids using a grid method;

[0009] S2. Obtain the electromagnetic radiation index of each grid and construct the original data matrix based on the electromagnetic radiation index of each grid.

[0010] S3. Determine the high-optimal index values ​​in the original data matrix according to the optimal index method, and normalize the original data matrix according to the high-optimal index values ​​to obtain the decision matrix.

[0011] S4. Calculate the degree of similarity between the electromagnetic radiation index of each grid in the decision matrix and the value of the high-optimal index to obtain the degree of similarity matrix;

[0012] S5. Based on the proximity matrix, calculate the similarity value between every two grids, and construct a similarity matrix based on the similarity value;

[0013] S6. Classify multiple grids based on the similarity matrix to obtain multiple sets of classification results;

[0014] S7. For each group of classification results, analyze the consistency of grids of the same type and the differences between grids of different types to determine the final classification result;

[0015] S8. Based on the final classification results, determine the representative grid of the area to be monitored.

[0016] Furthermore, the monitoring area is divided into multiple grids using a grid method, specifically including: uniformly dividing the monitoring area into grids according to a preset area value; and / or, dividing the monitoring area into non-uniform grid regions according to regional characteristic factors.

[0017] Furthermore, the elements in the original data matrix include the maximum value and average value of the electromagnetic radiation index of each grid. The optimal index value in the original data matrix is ​​determined according to the optimal index method, specifically including: selecting the maximum value among the maximum values ​​of the electromagnetic radiation index of each grid in the original data matrix; selecting the maximum value among the average values ​​of the electromagnetic radiation index of each grid in the original data matrix; and taking the maximum value among the maximum values ​​of the electromagnetic radiation index of each grid and the maximum value among the average values ​​of the electromagnetic radiation index of each grid as the optimal index value.

[0018] Furthermore, the original data matrix is ​​normalized based on the high-optimality index value to obtain a decision matrix. Specifically, this includes: dividing each element in the original data matrix by the high-optimality index value to obtain the corresponding normalized value; and arranging all the normalized data in the row and column order of the original data matrix to obtain the decision matrix.

[0019] Furthermore, the proximity of each grid's electromagnetic radiation index to the high-optimal index value in the decision matrix is ​​calculated to obtain a proximity matrix. Specifically, this includes: for each grid in the decision matrix, taking the arithmetic mean of all electromagnetic radiation indices in that grid to obtain the proximity value between that grid and the high-optimal index value; and arranging all the proximity values ​​of all grids in order of grid number to obtain a proximity matrix.

[0020] Furthermore, based on the proximity matrix, the similarity value between every two grids is calculated, and a similarity matrix is ​​constructed according to the similarity value. Specifically, this includes: using the least arithmetic mean method to calculate the similarity coefficient between any two corresponding values ​​in the proximity matrix, which is taken as the similarity value; arranging all similar values ​​according to the grid number to obtain the similarity matrix; wherein, the similarity matrix is ​​a symmetric matrix with all diagonal values ​​being 1.

[0021] Furthermore, multiple grids are classified based on the similarity matrix to obtain multiple sets of classification results. Specifically, this includes: selecting multiple similarity values ​​in the similarity matrix in descending order, and using these multiple similarity values ​​as classification thresholds in turn; and determining multiple sets of classification results containing different numbers of categories based on the multiple classification thresholds.

[0022] Furthermore, for each group of classification results, the consistency of grids within the same category and the differences between grids of different categories are analyzed to determine the final classification result. Specifically, this includes: for each group of classification results, calculating the dispersion of electromagnetic radiation indices of grids within the same category and obtaining consistency results; for each group of classification results, performing statistical tests on electromagnetic radiation indices of grids of different categories and obtaining difference results; and comprehensively analyzing the consistency results and the difference results, selecting the group with the highest consistency result and the largest difference result as the final classification result.

[0023] Furthermore, based on the final classification results, the final representative grid for the area to be monitored is determined, specifically including: for each class of grids in the final classification results, calculating the intra-class mean of the electromagnetic radiation index of all grids within that class; in each class of grids, selecting the grids whose electromagnetic radiation index is closest to the intra-class mean of that class's index, and determining the candidate representative grids for that class; if the relative error between the electromagnetic radiation index of the candidate representative grid and the average value of the electromagnetic radiation index of all grids in the area to be monitored is less than the set control target value, the candidate representative grid is determined as the final representative grid for the area to be monitored.

[0024] The present invention also provides an electromagnetic environment monitoring point optimization deployment system, wherein the system executes the electromagnetic environment monitoring point optimization deployment method described in any of the preceding claims, and the system includes:

[0025] The grid division module is used to divide the area to be monitored into multiple grids using a grid method.

[0026] The original data matrix construction module is connected to the grid division module and is used to obtain the electromagnetic radiation index of each grid and construct the original data matrix based on the electromagnetic radiation index of each grid.

[0027] The decision matrix construction module is connected to the original data matrix construction module. It is used to determine the high-optimal index values ​​in the original data matrix according to the optimal index method, and to normalize the original data matrix according to the high-optimal index values ​​to obtain the decision matrix.

[0028] A proximity matrix construction module, connected to the decision matrix construction module, is used to calculate the proximity of each grid electromagnetic radiation index in the decision matrix to the high-optimality index value, thereby obtaining a proximity matrix.

[0029] A similarity matrix construction module, connected to the proximity matrix construction module, is used to calculate the similarity value between every two grids based on the proximity matrix, and construct a similarity matrix based on the similarity value;

[0030] The classification module, connected to the similarity matrix construction module, is used to classify multiple grids based on the similarity matrix to obtain multiple sets of classification results;

[0031] The final classification result determination module, connected to the classification module, is used to analyze the consistency of grids of the same type and the differences between grids of different types for each group of classification results, and determine the final classification result.

[0032] The representative grid determination module, connected to the final classification result determination module, is used to determine the final representative grid of the area to be monitored based on the final classification result.

[0033] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0034] This invention divides the area to be monitored into multiple grids, acquires the electromagnetic radiation index of each grid, and constructs an original data matrix based on the electromagnetic radiation index of each grid. The optimal index method is used to determine the high-optimal index values ​​in the original data matrix, and the original data matrix is ​​normalized based on these high-optimal index values ​​to obtain a decision matrix. The similarity between the electromagnetic radiation index of each grid in the decision matrix and the high-optimal index values ​​is calculated to obtain a similarity matrix. Based on the similarity matrix, the similarity value between every two grids is calculated, and a similarity matrix is ​​constructed based on the similarity value. Multiple grids are classified based on the similarity matrix to obtain multiple sets of classification results. For each set of classification results, the consistency of grids of the same type and the differences between grids of different types are analyzed to determine the final classification result. Based on the final classification result, representative grids for the area to be monitored are determined. The optimal index method is used to normalize the original data, eliminating differences in data dimensions, allowing direct comparison of the electromagnetic radiation levels of different grids. The grids are then classified using a similarity matrix, and consistency and difference analyses are performed on multiple classification results to ensure that the electromagnetic environment characteristics of grids of the same type are highly similar and the differences between grids of different types are significant. Selecting the final representative grid based on the final classification results can accurately reflect the overall characteristics of the corresponding category of grids, and has high representativeness and accuracy. Attached Figure Description

[0035] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0036] Figure 1 This is a flowchart of an electromagnetic environment monitoring optimization point layout method provided by an embodiment of the present invention;

[0037] Figure 2 This is a schematic diagram of an electromagnetic environment monitoring optimized deployment system provided in an embodiment of the present invention. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0039] The specific embodiments of the present invention will be described below.

[0040] To address the problem that existing technologies often suffer from low representativeness and accuracy in selecting monitoring points, failing to reflect the true environment of the area, this invention divides the monitoring area into multiple grids, acquires electromagnetic radiation indicators for each grid, and constructs an original data matrix. High-optimal indicator values ​​are determined using the optimal indicator method, and the original data matrix is ​​normalized to obtain a decision matrix. The decision matrix is ​​then transformed to obtain a proximity matrix. Based on the proximity matrix, the similarity value between every two grids is calculated, constructing a similarity matrix. The grids are then classified based on the similarity matrix, yielding multiple classification results. For each classification result, the consistency of grids within the same category and the differences between grids of different categories are analyzed to determine the final classification result. Based on the final classification result, representative monitoring points are selected from each category of grids. This invention achieves high representativeness and accuracy in selecting monitoring points.

[0041] Example 1

[0042] This invention provides a method for optimizing the deployment of electromagnetic environment monitoring points. Figure 1 This is a flowchart of an electromagnetic environment monitoring point optimization method provided by an embodiment of the present invention, such as... Figure 1 As shown, the specific steps include the following:

[0043] S1. The area to be monitored is divided into multiple grids using the grid method.

[0044] The area to be monitored refers to a specific geographical area where electromagnetic environment quality monitoring is required. The grid method involves dividing the area to be monitored into several grids and collecting, analyzing, or calculating data for each grid. After determining the area to be monitored, monitoring points are evenly distributed around the center of each grid, with no fewer than five monitoring points in each grid. The background electromagnetic environment quality data (i.e., the electromagnetic radiation index of each grid) is obtained through these monitoring points.

[0045] The continuous monitoring area is divided into multiple grids using a grid method. Within each grid, radio frequency (RF) field strength data is acquired at each monitoring point using manual monitoring or electromagnetic monitoring vehicle patrols. This RF field strength data serves as an indicator of electromagnetic radiation. For example, manual monitoring involves using a portable RF field strength meter to read 10 consecutive, stable values ​​for 15 seconds at each monitoring point. Electromagnetic monitoring vehicle patrols involve a vehicle-mounted environmental RF field strength meter measurement system based on a Geographic Information System (GIS) to record data for each monitoring point in real time. The vehicle-mounted environmental RF field strength meter measurement system combines the geographical location information of the monitoring points with a digital map using GIS technology, enabling simultaneous acquisition of RF field strength data and corresponding geographical location information for each monitoring point. By setting multiple monitoring points within each grid, the random errors of individual monitoring points can be reduced. Simultaneously, the grid division covers the entire monitoring area, avoiding the omission of key monitoring areas.

[0046] Specifically, the monitoring area is divided into multiple grids using a grid method, including: uniformly dividing the monitoring area into grids based on a preset area value; and / or, dividing the monitoring area into non-uniform grid regions based on regional characteristic factors.

[0047] After determining the area to be monitored, the area is divided into uniform and / or non-uniform grids. Uniform grid division involves dividing the area into multiple grids of equal area at preset intervals, with each grid serving as an independent monitoring unit. For example, the preset area value might be 1000m × 1000m. Non-uniform grid division involves adjusting the grid area based on regional characteristics, such as environmental functional zoning and population density. For instance, smaller, denser grids are used in densely populated areas with concentrated electromagnetic pollution sources (such as residential or industrial areas), while larger grids are used in sparsely populated areas with fewer pollution sources (such as the urban fringe), creating grid units of varying sizes.

[0048] The uniform grid division in the above embodiments covers the entire monitoring area with a fixed area, avoiding any omissions and ensuring the integrity of the monitoring data. Non-uniform grid division, combined with regional characteristic factors, uses a denser grid in key areas to obtain more refined data, while expanding the grid area in secondary areas reduces unnecessary monitoring costs and solves the problem of traditional uniform division ignoring regional differences. Furthermore, both uniform and non-uniform grid division methods can be used simultaneously; that is, some areas are divided with a uniform grid, while other areas are adjusted with a non-uniform grid based on regional characteristic factors. This is suitable for both areas with simple and complex environmental characteristics, balancing comprehensiveness and specificity in monitoring.

[0049] S2. Obtain the electromagnetic radiation index of each grid and construct the original data matrix based on the electromagnetic radiation index of each grid.

[0050] Electromagnetic radiation indices refer to key parameters reflecting the electromagnetic radiation level within each grid. Referring to the above embodiment, electromagnetic radiation indices can be represented by radio frequency integrated field strength data. The original data matrix is ​​constructed using the electromagnetic radiation indices of each grid as elements.

[0051] For each grid segment, radio frequency (RF) field strength data of monitoring points is acquired through manual or vehicle-mounted surveys. The monitoring data for each grid is statistically analyzed, and the maximum and average RF field strengths for that grid are calculated. These maximum and average RF field strengths are used as electromagnetic radiation indicators. The maximum and average values ​​of all grids are arranged in grid order to construct an original data matrix. By constructing the original data matrix, the dispersed grid indicators are provided with standardized input for subsequent data processing.

[0052] S3. Determine the high-optimal index values ​​in the original data matrix according to the optimal index method, and normalize the original data matrix according to the high-optimal index values ​​to obtain the decision matrix.

[0053] The optimal index method is a multi-objective decision-making and evaluation method based on the principle of near-optimal level. The optimal index method makes decisions and evaluates multi-objective systems based on the principle of near-optimal level. It can extract high-optimal index values ​​from the original data matrix and standardize the data through normalization processing.

[0054] For example, the elements in the original data matrix include two electromagnetic radiation indices: the maximum value of the integrated radio frequency field strength data for each grid and the average value of the integrated radio frequency field strength data. The high-quality index value is the maximum value corresponding to the two electromagnetic radiation indices in the original matrix.

[0055] The optimal index values ​​in the original data matrix are determined using the optimal index method. Specifically, this involves: selecting the maximum value among the maximum values ​​of electromagnetic radiation indices for each grid in the original data matrix; selecting the maximum value among the average values ​​of electromagnetic radiation indices for each grid in the original data matrix; and using both the maximum value of the maximum values ​​and the maximum value of the average values ​​of electromagnetic radiation indices for each grid as the optimal index values. By selecting the maximum values ​​of both the maximum and average values ​​of electromagnetic radiation indices as the optimal index values, a clear reference value for the optimal level is provided for subsequent analysis, giving a unified standard for comparison of electromagnetic radiation levels in each grid. This avoids the bias of subjective experience in judging the optimal level, making subsequent optimization of monitoring points more objective and accurate.

[0056] The original data matrix is ​​normalized based on the high-optimality index value to obtain a decision matrix. Specifically, this includes: dividing each element in the original data matrix by the high-optimality index value to obtain the corresponding normalized value; and arranging all the normalized data in the row and column order of the original data matrix to obtain the decision matrix.

[0057] For the two electromagnetic radiation indices in the original data matrix, normalization is performed using their corresponding superior index values. For the maximum value index, the maximum value of each grid cell is divided by the maximum value among all grid maximum values; for the average value index, the average value of each grid cell is divided by the maximum average value among all grid average values. This division operation is performed on each element of the original data matrix to obtain dimensionless normalized values. All normalized values ​​are arranged in the row and column order of the original data matrix, maintaining the correspondence between grid cells and indices, forming a decision matrix. For example, the first row corresponds to the normalized maximum and average values ​​of the electromagnetic radiation index for the first grid cell. The decision matrix is ​​the standardized matrix formed after normalization of the original data matrix; each element in the matrix reflects the relative closeness of the electromagnetic radiation index to its corresponding superior index value.

[0058] The grid indicators in the original data matrix cannot be directly compared due to differences in units or magnitudes. For example, the maximum and average values ​​of the electromagnetic radiation indicator are of different magnitudes. After normalization, the data are all relative values ​​in the range [0,1], allowing both types of indicators to be analyzed on the same dimension. The elements in the decision matrix intuitively reflect the relative closeness of each grid indicator to the high-optimal indicator value. For example, the closer an element in the decision matrix is ​​to 1, the closer it is to the high-optimal indicator value.

[0059] S4. Calculate the degree of similarity between the electromagnetic radiation index of each grid in the decision matrix and the value of the high-optimal index to obtain the similarity matrix.

[0060] Specifically, the proximity of each grid's electromagnetic radiation index to the high-optimal index value in the decision matrix is ​​calculated to obtain a proximity matrix. This includes: taking the arithmetic mean of all electromagnetic radiation indices in each grid in the decision matrix to obtain the proximity value between the grid and the high-optimal index value; and arranging the proximity values ​​of all grids in order of grid number to obtain a proximity matrix.

[0061] The proximity score is a quantification of the overall proximity of a single grid cell in the decision matrix to the high-quality index value in the two aforementioned indicators. The calculation requires normalization elements of the decision matrix, and the core is to take the arithmetic mean of the normalized values ​​of the two indicators for a single grid cell. The proximity score matrix is ​​a one-dimensional matrix formed by arranging the proximity score values ​​of all grid cells according to their grid numbers, providing the core input for subsequent cluster analysis.

[0062] In the proximity matrix, smaller element values ​​indicate that the grid is closer to the optimal level for the corresponding indicator, making it more likely to become a representative grid during subsequent screening. Larger element values ​​indicate a greater deviation from the optimal level and weaker representativeness. The proximity matrix intuitively reflects the specific gap between each grid and the high-optimal level, overcoming the limitation of the decision matrix which only reflects proximity, and providing a more comprehensive quantitative basis for grid classification.

[0063] S5. Calculate the similarity value between every two grids based on the proximity matrix, and construct a similarity matrix based on the similarity value.

[0064] In the proximity matrix, each row corresponds to one grid, and each column corresponds to the difference value of one electromagnetic radiation index. For example, the first column represents the difference between the maximum value of the electromagnetic radiation index in each grid and the superior index, and the second column represents the difference between the average value of the electromagnetic radiation index in each grid and the superior index. The similarity value between any two grids is calculated by quantifying the consistency of the deviations of the two grids across all indices.

[0065] Specifically, the similarity value between every two grids is calculated based on the proximity matrix, and a similarity matrix is ​​constructed based on the similarity value. This includes: calculating the similarity coefficient between corresponding values ​​of any two grids in the proximity matrix using the least arithmetic mean method, and using this coefficient as the similarity value; arranging all similarity values ​​according to their grid numbers to obtain the similarity matrix. The similarity matrix is ​​a symmetric matrix with all diagonal values ​​being 1.

[0066] The minimum arithmetic mean method integrates multi-indicator deviation information through arithmetic averaging, and uses the logic of the minimum deviation average corresponding to the highest similarity to ensure the objectivity and comprehensiveness of grid similarity calculation. By calculating similarity values, the difference in radiation levels between grids is transformed from a qualitative judgment to quantitative data, avoiding the subjectivity of empirical classification. The similarity matrix is ​​the core input data for subsequent direct clustering methods; based on the similarity matrix, grids can be divided into different categories according to different thresholds.

[0067] S6. Classify multiple grids based on the similarity matrix to obtain multiple sets of classification results.

[0068] Specifically, multiple grids are classified based on the similarity matrix to obtain multiple sets of classification results. This includes: selecting multiple similarity values ​​from the similarity matrix in descending order and using these similarity values ​​as classification thresholds; and determining multiple sets of classification results containing different numbers of categories based on these thresholds. By obtaining multiple sets of classification results with different numbers of categories, multiple alternative schemes are provided for subsequent analysis of consistency within the same category and differences between different categories, ensuring that the final classification conforms to the actual characteristics of the electromagnetic environment and meets the accuracy requirements of monitoring point deployment.

[0069] S7. For each group of classification results, analyze the consistency of grids of the same type and the differences between grids of different types to determine the final classification result.

[0070] Specifically, for each group of classification results, the consistency of grids within the same category and the differences between grids of different categories are analyzed to determine the final classification result. This includes: for each group of classification results, calculating the dispersion of electromagnetic radiation indices of grids within the same category and obtaining consistency results; for each group of classification results, performing statistical tests on electromagnetic radiation indices of grids of different categories and obtaining difference results; and comprehensively analyzing the consistency results and the difference results, selecting the group with the highest consistency result and the largest difference result as the final classification result.

[0071] S8. Based on the final classification results, determine the representative grid of the area to be monitored.

[0072] Specifically, for each grid class in the final classification results, the mean value of the electromagnetic radiation index of all grids in that class is calculated. In each grid class, the grid whose electromagnetic radiation index is closest to the mean value of that class is selected to determine the candidate representative grid for that class. If the relative error between the electromagnetic radiation index of the candidate representative grid and the average value of the electromagnetic radiation index of all grids in the area to be monitored is less than the set control target value, the candidate representative grid is determined as the final representative grid of the area to be monitored.

[0073] By first matching the intra-class mean to ensure the representativeness of the candidate representative grids selected in each class, and then verifying the error of the monitored area for each candidate representative grid, the representativeness of the final representative grid is guaranteed. This dual screening avoids both bias in selecting similar points and overall bias in the monitored area. Monitoring is then conducted based on the monitoring points in the final representative grid, reducing the number of monitoring points while accurately reflecting the electromagnetic environment quality of the monitored area.

[0074] For example, electromagnetic environment monitoring is conducted on streets in a city. The radio frequency (RF) field strength data for each grid is obtained through manual monitoring or patrol by an electromagnetic monitoring vehicle. Based on the obtained RF field strength data for each grid, an original data matrix X is constructed, as shown below:

[0075] ;

[0076] In the original data matrix X, the first column represents the maximum value of the RF integrated field strength data in each grid, and the second column represents the average value of the RF integrated field strength data in each grid. There are 197 grids in total. An element of 0.2 in the original data matrix X indicates that both the maximum and average RF integrated field strength data of the grid are less than 0.2. When the maximum and average RF integrated field strength data are greater than 0.2, the actual values ​​are used. The unit of the RF integrated field strength data is volts per meter (V / m).

[0077] The high-quality index value is determined based on the original data matrix X. The formula for calculating the high-quality index value is as follows:

[0078] ;

[0079] , ;

[0080] Where y represents the high-quality index matrix, x1 represents the high-quality index value in the first column of the original data matrix, and x2 represents the high-quality index value in the second column of the original data matrix.

[0081] The original data matrix is ​​normalized based on the high-quality index values ​​to obtain the decision matrix. The normalization formula is as follows:

[0082] z=x ij / x j ;

[0083] z represents the normalized data (elements in the decision matrix), x ij x represents an element in the original data matrix. j The optimal index values ​​are j=1,2; i=1,2…197.

[0084] The resulting decision matrix Z is:

[0085] ;

[0086] The decision matrix contains elements ranging from 0 to 1. Transforming the decision matrix yields a proximity matrix, which is a one-dimensional matrix. Each element in the proximity matrix represents a degree of closeness, C. i The calculation formula is as follows:

[0087] ;

[0088] Among them, Z ij Let j = 1, 2; i = 1, 2…197 represent the elements in the decision matrix.

[0089] The least arithmetic mean method is used to calculate the similarity coefficient between corresponding values ​​of any two grids in the proximity matrix, which serves as the similarity value. For example, the formula for calculating the similarity coefficient is as follows:

[0090] ;

[0091] Where, r mn Let M represent the similarity coefficient, and M represent the number of electromagnetic radiation indices corresponding to each grid in the proximity matrix. Based on the one-dimensional proximity matrix described above, M=1. Cmk and Cnk represent the proximity between the m-th and n-th rows of the proximity matrix, respectively. The resulting m×n similarity matrix is:

[0092] ;

[0093] The above similarity matrices were obtained using direct clustering and the F-statistic:

[0094] When the value is 1, there are 58 categories;

[0095] When the value is 0.99 (F=137.72), it is divided into 31 categories;

[0096] When the value is 0.98 (F=32.74), it is divided into 21 categories;

[0097] When the value is 0.97 (F=143.95), it is divided into 16 categories;

[0098] When the value is 0.96 (F=40.62), it is divided into 15 categories;

[0099] When the value is 0.95 (F=18.11), it is divided into 12 categories;

[0100] When the value is 0.94 (F=13.06), it is divided into 8 categories;

[0101] When the value is 0.90 (F=200.78), it is divided into 7 categories;

[0102] When the value is 0.88 (F=40.98), it is divided into 5 categories;

[0103] When the value is 0.86 (F=67.70), it is divided into 4 categories;

[0104] When the value is 0.78 (F=49.21), it is divided into 3 categories;

[0105] When the value is 0.75 (F=77.61), it is divided into 2 categories;

[0106] When the value is 0.57, it is classified into category 1;

[0107] Here, λ is the similarity coefficient threshold used in cluster analysis to classify network categories. By setting different similarity coefficient thresholds, the fineness of grid clustering is controlled, ultimately resulting in different numbers of grid categories. The similarity coefficient threshold is selected from the elements of the similarity matrix. The F-statistic is a core quantitative indicator used to evaluate the reasonableness of the clustering results. The F-statistic is the ratio of the between-class mean square to the within-class mean square. The larger the F-value, the greater the difference between classes, and the better the classification. Furthermore, based on practical experience... The classification with F=0.90 (F=200.78) is the most suitable, and the final classification result is determined.

[0108] For example, the final classification result contains 7 categories. For each category of grids, the mean value of the electromagnetic radiation index of all grids in that category is calculated. In each category of grids, the grids whose electromagnetic radiation index is closest to the mean value of that category are selected to determine the candidate representative grids of that category. If the relative error between the electromagnetic radiation index of the candidate representative grid and the average value of the electromagnetic radiation index of all grids in the area to be monitored is less than the set control target value, the candidate representative grid is determined as the final representative grid of that category.

[0109] This embodiment divides the area to be monitored into multiple grids, obtains the electromagnetic radiation index of each grid, and constructs an original data matrix based on the electromagnetic radiation index of each grid. The optimal index method is used to determine the high-optimal index values ​​in the original data matrix, and the original data matrix is ​​normalized based on these high-optimal index values ​​to obtain a decision matrix. The similarity between the electromagnetic radiation index of each grid in the decision matrix and the high-optimal index values ​​is calculated to obtain a similarity matrix. Based on the similarity matrix, the similarity value between every two grids is calculated, and a similarity matrix is ​​constructed based on the similarity values. Multiple grids are classified based on the similarity matrix to obtain multiple sets of classification results. For each set of classification results, the consistency of grids of the same type and the differences between grids of different types are analyzed to determine the final classification result. Based on the final classification result, representative grids for the area to be monitored are determined. Normalizing the original data using the optimal index method eliminates differences in data dimensions, allowing direct comparison of the electromagnetic radiation levels of different grids. The grids are then classified using a similarity matrix, and consistency and difference analyses are performed on multiple classification results to ensure that the electromagnetic environment characteristics of grids of the same type are highly similar and the differences between grids of different types are significant. Selecting the final representative grid based on the final classification results can accurately reflect the overall characteristics of the corresponding category of grids, and has high representativeness and accuracy.

[0110] Example 2

[0111] This invention also provides an electromagnetic environment monitoring point optimization system. Figure 2 This is a schematic diagram of an electromagnetic environment monitoring optimized deployment system provided in an embodiment of the present invention, as shown below. Figure 2 As shown, the system includes:

[0112] The grid division module 110 is used to divide the area to be monitored into multiple grids using a grid method;

[0113] The original data matrix construction module 120 is connected to the grid division module 110 and is used to obtain the electromagnetic radiation index of each grid and construct the original data matrix based on the electromagnetic radiation index of each grid.

[0114] The decision matrix construction module 130 is connected to the original data matrix construction module 120. It is used to determine the high-optimal index values ​​in the original data matrix according to the optimal index method, and to normalize the original data matrix according to the high-optimal index values ​​to obtain the decision matrix.

[0115] The proximity matrix construction module 140 is connected to the decision matrix construction module 130 and is used to calculate the proximity of each grid electromagnetic radiation index in the decision matrix to the high-optimal index value, so as to obtain the proximity matrix.

[0116] The similarity matrix construction module 150 is connected to the proximity matrix construction module 140 and is used to calculate the similarity value between every two grids based on the proximity matrix, and construct a similarity matrix based on the similarity value;

[0117] The classification module 160 is connected to the similarity matrix construction module 150 and is used to classify multiple grids based on the similarity matrix to obtain multiple sets of classification results;

[0118] The final classification result determination module 170, connected to the classification module 160, is used to analyze the consistency of grids of the same type and the differences between grids of different types for each group of classification results, and determine the final classification result.

[0119] The representative grid determination module 180 is connected to the final classification result determination module 170 and is used to determine the final representative grid of the area to be monitored based on the final classification result.

[0120] The electromagnetic environment monitoring optimization deployment system implements the electromagnetic environment monitoring optimization deployment method described in any of the above embodiments, and has the beneficial effects of any of the above embodiments, which will not be repeated here.

[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing the deployment of electromagnetic environment monitoring points, characterized in that, include: S1. The area to be monitored is divided into multiple grids using a grid method; S2. Obtain the electromagnetic radiation index of each grid and construct the original data matrix based on the electromagnetic radiation index of each grid. S3. Determine the high-optimal index values ​​in the original data matrix according to the optimal index method, and normalize the original data matrix according to the high-optimal index values ​​to obtain the decision matrix. S4. Calculate the degree of similarity between the electromagnetic radiation index of each grid in the decision matrix and the value of the high-optimal index to obtain the degree of similarity matrix; S5. Based on the proximity matrix, calculate the similarity value between every two grids, and construct a similarity matrix based on the similarity value; S6. Classify multiple grids based on the similarity matrix to obtain multiple sets of classification results; S7. For each group of classification results, analyze the consistency of grids of the same type and the differences between grids of different types to determine the final classification result; S8. Based on the final classification results, determine the final representative grid of the area to be monitored.

2. The electromagnetic environment monitoring optimized deployment method according to claim 1, characterized in that, The area to be monitored is divided into multiple grids using a grid method, specifically including: The area to be monitored is divided into uniform grids according to a preset area value; And / or, The monitored area is divided into non-uniform grid regions based on regional characteristic factors.

3. The method for optimizing the deployment of electromagnetic environment monitoring points according to claim 1, characterized in that, The elements in the original data matrix include the maximum and average values ​​of the electromagnetic radiation index for each grid. The optimal index values ​​in the original data matrix are determined using the optimal index method, specifically including: Select the maximum value among the maximum values ​​of electromagnetic radiation indices for each grid in the original data matrix; Select the maximum value among the average values ​​of electromagnetic radiation indices for each grid in the original data matrix; The maximum value among the maximum values ​​of the electromagnetic radiation index of each grid and the maximum value among the average values ​​of the electromagnetic radiation index of each grid are taken as the high-quality index value.

4. The method for optimizing the deployment of electromagnetic environment monitoring points according to claim 3, characterized in that, The original data matrix is ​​normalized based on the high-quality index value to obtain a decision matrix, specifically including: Divide each element in the original data matrix by the high-quality index value to obtain the corresponding normalized value; Arrange all normalized data in the row and column order of the original data matrix to obtain the decision matrix.

5. The method for optimizing the deployment of electromagnetic environment monitoring points according to claim 1, characterized in that, Calculate the degree of similarity between the electromagnetic radiation index of each grid in the decision matrix and the value of the high-optimal index to obtain a similarity matrix, which specifically includes: For each grid in the decision matrix, the electromagnetic radiation index corresponding to each grid is calculated by taking the arithmetic mean of all electromagnetic radiation indices in that grid, and the degree of closeness between that grid and the high-quality index value is obtained. Arrange all the proximity values ​​of the grids in order of grid number to obtain the proximity matrix.

6. The method for optimizing the deployment of electromagnetic environment monitoring points according to claim 1, characterized in that, Calculate the similarity value between every two grid cells based on the proximity matrix, and construct a similarity matrix based on the similarity values, specifically including: The least arithmetic mean method is used to calculate the similarity coefficient of corresponding values ​​between any two grids in the proximity matrix, which is then used as the similarity value. Arrange all similar values ​​according to their grid numbers to obtain the similarity matrix; The similarity matrix is ​​a symmetric matrix whose diagonal values ​​are all 1.

7. The method for optimizing the deployment of electromagnetic environment monitoring points according to claim 1, characterized in that, Based on the similarity matrix, multiple grids are classified to obtain multiple sets of classification results, specifically including: Multiple similarity values ​​in the similarity matrix are selected sequentially from largest to smallest, and these multiple similarity values ​​are used as classification thresholds in turn. Multiple classification results containing different numbers of categories were determined based on multiple classification thresholds.

8. The method for optimizing the deployment of electromagnetic environment monitoring points according to claim 7, characterized in that, For each group of classification results, the consistency of grids within the same class and the differences between grids of different classes are analyzed to determine the final classification result, specifically including: For each group of classification results, the dispersion of electromagnetic radiation indices of the same type of grid is calculated, and consistency results are obtained; For each group of classification results, statistical tests were performed on the electromagnetic radiation indices of different grid types, and the differences were obtained. By comprehensively analyzing the consistency results and the discrepancies results, the group with the highest consistency result and the largest discrepancy result is selected as the final classification result.

9. The method for optimizing the deployment of electromagnetic environment monitoring points according to claim 8, characterized in that, Based on the final classification results, the final representative grid for the area to be monitored is determined, specifically including: For each grid class in the final classification results, calculate the intra-class mean of the electromagnetic radiation index of all grids in that class; Within each grid class, the grids whose electromagnetic radiation index is closest to the intra-class mean of that index are selected to determine the candidate representative grids for that class. If the relative error between the electromagnetic radiation index of the candidate representative grid and the average value of the electromagnetic radiation index of all grids in the area to be monitored is less than the set control target value, the candidate representative grid is determined as the final representative grid of the area to be monitored.

10. An electromagnetic environment monitoring point optimization deployment system, characterized in that, The system executes the electromagnetic environment monitoring optimized deployment method according to any one of claims 1-9, and the system comprises: The grid division module is used to divide the area to be monitored into multiple grids using a grid method. The original data matrix construction module is connected to the grid division module and is used to obtain the electromagnetic radiation index of each grid and construct the original data matrix based on the electromagnetic radiation index of each grid. The decision matrix construction module is connected to the original data matrix construction module. It is used to determine the high-optimal index values ​​in the original data matrix according to the optimal index method, and to normalize the original data matrix according to the high-optimal index values ​​to obtain the decision matrix. A proximity matrix construction module, connected to the decision matrix construction module, is used to calculate the proximity of each grid electromagnetic radiation index in the decision matrix to the high-optimality index value, thereby obtaining a proximity matrix. A similarity matrix construction module, connected to the proximity matrix construction module, is used to calculate the similarity value between every two grids based on the proximity matrix, and construct a similarity matrix based on the similarity value; The classification module, connected to the similarity matrix construction module, is used to classify multiple grids based on the similarity matrix to obtain multiple sets of classification results; The final classification result determination module, connected to the classification module, is used to analyze the consistency of grids of the same type and the differences between grids of different types for each group of classification results, and determine the final classification result. The representative grid determination module, connected to the final classification result determination module, is used to determine the final representative grid of the area to be monitored based on the final classification result.

Images