Target area identification method, system and equipment for complex electromagnetic environment

Abstract

The invention belongs to the technical field of target detection, and particularly relates to a complex electromagnetic environment-oriented target area identification method, system and equipment, and the method comprises the steps: constructing a two-dimensional plane to form a plane-coordinate system, taking the center of a target area as a coordinate origin, discretizing the target area into uniform grids, and carrying out the recognition of the target area. The grid covers a square area with the original point as the center, an incircle area of the grid serves as an out-of-domain area, an out-of-domain area indication matrix is generated, an arrival angle sector area indication matrix is generated through the out-of-domain indication matrix, then an intersection area indication matrix is generated through the arrival angle sector area indication matrix, and the intersection area indication matrix is used for indicating the intersection area. And finally, calculating the proportion of the intersection area of the arrival angle sectors in the target area through the intersection area indication matrix, setting a judgment threshold value, judging that the target is in the target area if the proportion is greater than or equal to the threshold value, otherwise, judging that the target is outside the target area. The system and the equipment are used for implementing the method. According to the method, the accuracy, the robustness and the calculation efficiency of target region identification are remarkably improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of target detection, and particularly relates to a target area recognition method, system and device for a complex electromagnetic environment. BACKGROUND

[0002] Angle of Arrival (AOA) technology is a key method in passive positioning, which determines the position of the transmitting source by measuring the angle of signal arrival at the receiving station. AOA technology uses multiple measurement sites to obtain the angle of arrival of the target signal, and then calculates the spatial position of the target through the principle of triangulation. Compared with other positioning technologies, AOA has the advantages of simple device structure, independence from time synchronization, and suitability for various signal types. However, AOA technology faces the problem of limited measurement accuracy in practical application. Due to factors such as antenna array precision limitation, multipath effect, environmental interference, etc., the AOA measurement value often has a large error, especially in complex environments, which may reach dozens of degrees, seriously affecting the positioning accuracy.

[0003] Institute of East China University of its application in the patent application documents "a keyless system positioning method based on Bluetooth RSSI" (application number: CN201710167473.4, publication number: CN107124696A) proposed a keyless system positioning method based on Bluetooth RSSI (Received Signal Strength Indicator, Received Signal Strength Indicator). The invention aims to determine the position of the car intelligent key through the Bluetooth signal strength, which is suitable for both in-car and out-of-car scenarios, and belongs to the field of wireless communication and intelligent control technology. This method uses Bluetooth RSSI for positioning, without the need for additional infrastructure, with lower cost and suitability for a variety of scenarios in and out of the car. However, this method relies on signal strength measurement, which is susceptible to errors introduced by environmental noise, multipath effect and obstacle attenuation, which may lead to inaccurate distance calculation.

[0004] Institute of Fraunhofer Application Research Promotion Association in its application in the patent documents "positioning concept on a path" (application number: DE102006061650, publication number: DE102006061650A1) proposed a path positioning method based on radio signals. This path feature matching-based method provides a low-cost, easy-to-implement positioning solution for specific scenarios such as repeated path navigation, but its dependence on pre-recorded paths, sensitivity to signal and path changes, and limited positioning accuracy limit its application in applications that require general, robust area recognition.

[0005] Tianjin University proposed an array antenna indoor positioning algorithm based on AOA and PDOA in its patent application "An Improved RSSI-Based Positioning Method Using Sector Transmission Model and Distance Optimization Technique" (Application Number: CN201710156453.7, Publication Number: CN106793087A). The invention aims to use the energy and phase information of the multipath signals received by the array antenna to determine the position of the indoor positioning object, belonging to the field of wireless communication and positioning technology. The main steps of this method are: this method uses array antenna combined with AOA and PDOA estimation, which can effectively handle multipath and non-line-of-sight effects, and improve indoor positioning accuracy through virtual base stations and optimization algorithms. The disadvantage is that the method relies on array antenna to increase the cost and complexity, and AOA and PDOA estimation is easily affected by interference and environmental changes, which may limit its performance in complex environments.

[0006] In the literature "An Improved RSSI-Based Positioning Method Using Sector Transmission Model and Distance Optimization Technique", an RSSI-based positioning algorithm is proposed (Zhu, H., & Alsharari, T. (2015). An improved RSSI-based positioning method using sector transmission model and distance optimization technique. International Journal of Distributed Sensor Networks, 2015, Article ID 587195.), aiming to determine the position of the blind node through the RSSI value. However, the accuracy of the environment modeling of this method depends on a large extent, and the granularity of sector division and the fitting accuracy of the transmission model directly affect the positioning performance. In addition, establishing an accurate sector transmission model requires a large amount of offline measurement data, which increases the cost and complexity of deployment and maintenance. SUMMARY

[0007] In order to overcome the above-mentioned prior art defects, the purpose of the present application is to provide a target area identification method, system and device for complex electromagnetic environment, which first determines the possible fan-shaped area of the target to be positioned corresponding to each positioning station according to the angle of arrival measurement value of the target to be positioned by each positioning station and the preset maximum error range; then calculates the intersection area of the fan-shaped areas within the outer circle range; finally, by analyzing the area distribution proportion of the intersection area under the division of the inner and outer circle boundaries, it is determined whether the target is located in the internal area or the external area, thereby improving the area identification accuracy and robustness under the condition of AOA measurement error, not relying on array antenna, reducing the cost and complexity, avoiding the problem of inaccurate distance calculation caused by RSSI due to environmental complexity, and reducing the accuracy of environmental modeling.

[0008] In order to achieve the above-mentioned purpose, the technical scheme adopted by the present application is:

[0009] A target area identification method for complex electromagnetic environment, comprising the following steps:

[0010] Step 1: mapping the target area of the positioning scene to a two-dimensional plane by plane projection to form a plane coordinate system, taking the center of the target area as a calculation reference point, and moving the target area in the plane coordinate system to make the calculation reference point coincide with the coordinate origin;

[0011] Step 2: discretizing the moved target area into uniform grids, and covering the square area with the origin as the center and the inscribed circle radius r outer ;

[0012] Step 3: taking the circular area with the origin as the center and r outer as the radius as the out-of-domain area, i.e. the total detection range, and generating an out-of-domain area indication matrix;

[0013] Step 4: generating an angle of arrival sector area indication matrix based on the out-of-domain indication matrix;

[0014] Step 5: generating an intersection area indication matrix based on the angle of arrival sector area indication matrix;

[0015] Step 6: calculating the proportion of the angle of arrival sector intersection area within the target area based on the intersection area indication matrix, and setting a determination threshold, if the proportion is greater than or equal to the threshold, it is determined that the target is located in the target area, otherwise it is determined that the target is located outside the target area.

[0016] The positioning scene contains N S positioning stations and N T targets to be positioned S j coordinate of the positioning station, T represents the target to be positioned k Coordinates defining two concentric circular regions, the center being the calculation reference point, the inner circle being the target region, and the radius being R in , and the outer circle radius being R out ;

[0017] Positioning stations S j are uniformly distributed on the boundary of the inner circle, and the target to be positioned T k is randomly distributed in the scene;

[0018] Each positioning station S j performs angle of arrival measurement on the signal received from the target to be positioned T k , and the true angle of arrival of the target to be positioned T j from the positioning station S k is defined as the angle between the vector pointing from the positioning station S j to the target to be positioned T k and the reference direction, denoted as θ true,j,k ;

[0019] The measurement error is defined as ∈ j,k , and the measurement error ∈ j,k obeys a Gaussian distribution with mean 0 and standard deviation σ AOA , or is uniformly distributed in the range [-Δθ max , +Δθ max ], then the angle of arrival measurement value θ j of the target to be positioned T k by the positioning station S meas,j,k is expressed as:

[0020] θ meas,j,k = θ true,j,k + ∈ j,k

[0021] where ∈ j,k is the measurement error, and the maximum error value is Δθ max degrees.

[0022] The latitude and longitude of the calculation reference point in step 1 are (λ c , φ c ), and the coordinates after planar projection are (x c , y c ). The coordinates of any point with latitude and longitude (λ, φ) after translation are:

[0023]

[0024] where R = 6378137 m is the long semi-axis of the Earth.

[0025] The grid in step 2 is centered at the origin and has an inscribed circle with radius r outer . m The coordinates (x n , y res ) of each point in the grid are determined by the following formula:

[0026]

[0027] where N is the resolution of the grid, indicating the number of grid points in the horizontal and vertical directions of the plane coordinate system; an N res × N res uniform grid is generated by these two parameters, covering the horizontal and vertical coordinate ranges from -r outer to r outer .

[0028] The out-of-domain region indicator matrix M outer (x, y) is defined as follows:

[0029]

[0030] where (x, y) represents the plane coordinates of any point in the grid, r outer is the preset inscribed circle radius, and I(·) in the function is an indicator function, which is defined as: when the condition in the parentheses is met, the function value is 1; otherwise, the function value is 0.

[0031] The step 4 specifically includes the following steps:

[0032] Calculate the relative vector of any grid point pointing to the current positioning station S s , and the horizontal coordinate component Δx and the vertical coordinate component Δy of the relative vector:

[0033]

[0034] where x m , y n represent the coordinates of any grid point, and x s , y s represent the plane coordinates of the current positioning station S grid(x,y) . Calculate the theoretical angle of arrival of any grid point relative to the current positioning station S s , denoted as θ grid(x,y) , using the horizontal coordinate component Δx and the vertical coordinate component Δy:

[0035]

[0036]

[0037] Wherein, the theoretical angle of arrival is defined as the angle of clockwise rotation from the positive longitudinal axis to the direction of the relative vector, ranging between 0° and 360°; is the angle between the relative vector and the positive transverse axis;

[0038] Calculate the measurement angle of arrival value θ s of the current positioning site S s measured The angle difference Δθ s between the theoretical angle of arrival value θ grid (x, y) of the grid point relative to the current positioning site S s .

[0039] Δθ s = |((θ s measured - θ grid (x, y) + 180°) (mod 360°)) - 180°|

[0040] According to the angle difference Δθ j of each positioning site S s calculated and the preset angle of arrival robustness range θ error , the angle of arrival sector area indication matrix of the current positioning site S s is generated If the distance from the out-of-domain area indication matrix M outer (x, y) to the origin is less than or equal to the preset radius r outer , the corresponding element of M outer (x, y) is 1, indicating that the point is within the total detection range; otherwise, it is 0, indicating that the point is outside the total detection range; the calculation formula of the angle of arrival sector area indication matrix is:

[0041]

[0042] Wherein, I(Δθ s ≤ θ error ) is an indication function.

[0043] The step 5 specifically includes the following steps: the intersection area of the angle of arrival sector area of the current positioning site S s is the current intersection area, and the current intersection area indication matrix M' intersect is used to represent; first, the current intersection area indication matrix M' intersect is initialized as the out-of-domain area indication matrix M outer ;

[0044] Then, all the positioning sites S j are traversed, and the current intersection area indication matrix M' intersect is updated by the angle of arrival sector area indication matrix of each positioning site logical AND operation is performed;

[0045] by performing intersection operation on all the angle-of-arrival sector area indication matrices M j the final intersection area indication matrix M intersect represents the common area of the angle-of-arrival sector areas of all the positioning stations S j , i.e. the set of grid points that satisfy the angle-of-arrival measurement error range of all the stations S j .

[0046] The step 6 specifically comprises the following steps:

[0047] First, generate an inner circle area indication matrix M inner (x,y) to indicate whether the coordinates (x,y) of a grid point are inside or on the boundary of the inner circle;

[0048]

[0049] where r inner is the preset inner circle radius, and I(·) is still the indicator function;

[0050] Next, calculate the proportion of the intersection area of the angle-of-arrival sectors inside the inner circle, denoted as R inner , and the calculation formula is:

[0051]

[0052] In the formula, the numerator M intersect ∩M inner represents the number of grid points that are simultaneously inside the intersection area of the angle-of-arrival sectors and the inner circle; and the denominator M intersect represents the total number of grid points in the intersection area of the angle-of-arrival sectors;

[0053] Finally, compare the calculated proportion R inner with the preset determination threshold t to determine the final area determination result, and the determination rule is as follows:

[0054]

[0055] If the proportion R inner of the intersection area of the angle-of-arrival sectors inside the inner circle is greater than or equal to t, it is determined that the target is inside the target area; otherwise, it is determined that the target is outside the target area.

[0056] A target area identification system for complex electromagnetic environment, comprising:

[0057] ​The pre-processing module maps a target region on the earth surface to a two-dimensional plane by plane projection to form a plane coordinate system, and moves the target region under the plane coordinate system so that a calculation reference point is coincident with a coordinate origin;

[0058] The grid discretization module discretizes the moved target region into uniform grids, and the grids cover a square region with the origin as a center and a radius r outer as an inscribed circle radius;

[0059] The matrix generation module generates a domain-outside region indication matrix based on a circular region with the origin as a center and r outer as a radius, generates an angle-of-arrival sector region indication matrix based on the domain-outside region indication matrix, and generates an intersection region indication matrix based on the angle-of-arrival sector region indication matrix;

[0060] The region determination module calculates a proportion of the angle-of-arrival sector intersection region in the target region based on the intersection region indication matrix, sets a determination threshold, and determines that the target is located in the target region when the proportion is greater than or equal to the threshold, and otherwise determines that the target is located outside the target region.

[0061] A target region identification device for a complex electromagnetic environment includes:

[0062] A memory is configured to store a computer program for implementing a target region identification method for a complex electromagnetic environment;

[0063] A processor is configured to implement the target region identification method for a complex electromagnetic environment when executing the computer program.

[0064] Compared with the prior art, the present application has the following advantages:

[0065] 1. The high-resolution grid is used in step 2 to discretize the target region, which realizes fine modeling and quantization of the target region and the intersection region of the multi-site AOAs, significantly improves the area calculation accuracy, and provides a reliable basis for the final region attribution determination.

[0066] 2. The dynamic site combination selection mechanism is used in step 6, which realizes intelligent selection and use of effective site combinations to form the intersection region even in the case of missing intersection region caused by serious deviation of the site angle-of-arrival measurement, significantly enhances the adaptability and robustness of the algorithm in the complex electromagnetic environment.

[0067] 3. The area distribution proportion based on the intersection region under the division of the inner and outer circle boundaries is used in step 6 to realize the conversion of the angle-of-arrival error into a quantifiable region coverage proportion, which is used as the determination basis for the region attribution, avoids the dependence on a single intersection point, and improves the accuracy and reliability of the region identification.

[0068] 4. The step 4 of the present application adopts the method of the arrival angle pyramid error range to realize the area identification, realizes directly dealing with the large error of the arrival angle measurement under the complex electromagnetic environment, converts the single point measurement into the fan area which can tolerate the error, and determines the possible area of the target through the intersection of the multi-station sector, significantly improves the accuracy and robustness of the area identification, and effectively solves the problem that the traditional accurate point positioning method is invalid when the error is too large.

[0069] 5. The step 1 of the present application adopts the method of converting the input latitude and longitude coordinates into the two-dimensional plane coordinates with the reference point as the origin for calculation, realizes the simplification of the complex earth surface geometry problem into the calculation in the standard two-dimensional Cartesian coordinate system, greatly simplifies the complexity of the subsequent grid generation, arrival angle calculation and area intersection analysis, and improves the universality and calculation efficiency of the algorithm.

[0070] In summary, the present application effectively solves the key problem that the traditional positioning method is invalid due to the too large error of the arrival angle measurement under the complex electromagnetic environment by combining the high-resolution grid modeling, dynamic station combination selection, area judgment based on the area ratio and the area identification of the arrival angle pyramid error range, and simplifying the latitude and longitude coordinates into the two-dimensional plane calculation, significantly improves the accuracy, robustness and calculation efficiency of the target area identification. BRIEF DESCRIPTION OF DRAWINGS

[0071] Figure 1 It is a scene assumption schematic diagram of the area identification of the present application.

[0072] Figure 2 It is a scene simulation diagram of the area identification of the present application.

[0073] Fig. 3 (a) is a performance comparison curve diagram of the present application and the AOA positioning method when the target to be positioned is distributed inside the target area.

[0074] Fig. 3 (b) is a performance comparison curve diagram of the present application and the AOA positioning method when the target to be positioned is distributed outside the target area.

[0075] Figure 4 It is a performance comparison curve diagram of the present application and the AOA positioning method when the target to be positioned is uniformly distributed inside and outside the target area. DETAILED DESCRIPTION

[0076] The present application will be described in detail below with reference to the drawings.

[0077] A target area identification method for complex electromagnetic environment includes the following scene assumptions:

[0078] Assume that the positioning scene contains N S positioning stations and N TTarget to be localized Positioning station S j Coordinates, Target to be localized T k Coordinates, defining two concentric circular regions, the inner circle is the target region, radius R in (meters), the outer circle radius is R out (meters);

[0079] Positioning station S j Uniformly distributed on the boundary of the inner circle, target to be localized T k Randomly distributed in the scene, in order to simulate the positioning situation under different distances, part of the targets are located in the inner circle region, and the other part of the targets are located in the annular region between the inner circle and the outer circle;

[0080] Each positioning station S j performs Angle of Arrival (AOA) measurement on the signal received from the target to be localized T k The true AOA of the target to be localized T j from the positioning station S k is defined as the angle between the vector pointing from the positioning station S j to the target to be localized T k and the reference direction (e.g. the north direction), denoted as θ true,j,k ;

[0081] In actual measurement, the AOA measurement value will be affected by random error, define the measurement error as ∈ j,k , the measurement error ∈ j,k obeys Gaussian distribution with mean 0 and standard deviation σ AOA , or is uniformly distributed in the range [-Δθ max , +Δθ max ], then the AOA measurement value θ j of the target to be localized T k by the positioning station S meas,j,k is represented as:

[0082] θ meas,j,k = θ true,j,k + ∈ j,k

[0083] where ∈ j,k is the measurement error, and the maximum error value is Δθ max degrees;

[0084] This scenario aims to simulate the challenges and methods of localizing the target to be localized T k through AOA measurement data with error under different spatial distributions. A typical positioning scenario is shown in Figure 1 .

[0085] The target area recognition method has the following specific steps:

[0086] Step 1: Boundary preprocessing

[0087] For the area recognition algorithm with unknown positioning results, in order to improve the calculation efficiency and accuracy of the area discrimination algorithm based on angle of arrival (AOA), the input boundary latitude and longitude coordinates are preprocessed to convert them into coordinates in a plane coordinate system, and the calculation reference point (the center of the target area, i.e. the center of the concentric circle) is further moved to the coordinate origin. Since direct distance and angle calculation in the spherical coordinate system will introduce significant errors due to the curvature of the earth, especially in large-scale areas or near the poles, the target area of the positioning scene is mapped to a two-dimensional plane through plane projection, making the calculation more accurate. The center of the target area in the plane coordinate system is denoted as the calculation reference point, and the target area in the plane coordinate system is moved so that the calculation reference point moves to the coordinate origin, which can effectively simplify the subsequent geometric calculation process. In the plane rectangular coordinate system, the distance between two points is calculated by the square root of the sum of the squares of the coordinate differences. By placing the calculation reference point at the origin, the vector calculation related to this point (such as calculating the vector of the grid point relative to the station) becomes intuitive and easy to handle, for example, when calculating the angle of arrival (AOA), the relative coordinate difference is directly used for arctangent operation. This coordinate translation not only simplifies the calculation formula and reduces the calculation steps, but also helps to control the numerical errors that may occur in the calculation process, thereby further improving the efficiency and stability of the algorithm.

[0088] Let the latitude and longitude of the center of the target area be (λ c ,φ c ), and the coordinates after plane projection be (x c ,y c ), then the coordinates of any point with latitude and longitude (λ,φ) after translation are:

[0089]

[0090] r=R·cosφ c

[0091]

[0092] where R = 6378137 m is the length of the semi-major axis of the earth.

[0093] Step 2: Plane coordinate system and grid discretization

[0094] When performing region discrimination, the target region is set on a two-dimensional plane, with the center of the monitored area as the origin. To perform effective analysis and calculation of this region, the entire target region is discretized into a uniform grid. The grid covers a circle centered at the origin with an inscribed circle of radius r. outer A square region. The coordinates (x, y) of each point in the grid. m ,y n It is determined by the following formula:

[0095]

[0096] in, It defines the extent of the grid coverage; N res N represents the resolution of the grid, indicating the number of grid points in each of the horizontal and vertical directions of the planar coordinate system (in this embodiment, N). res Set to 200). An N is generated using these two parameters. res ×N res A uniform grid, covering from -r outer to r outer The range of the horizontal and vertical coordinates.

[0097] Step 3: Delineation of the total monitoring area and generation of the indicator matrix for the outer region

[0098] After discretizing the target region into a uniform grid, the primary task is to define an overall monitoring range. This step aims to ensure that all subsequent calculations and region determinations are strictly confined to a circle centered at the origin with a preset radius of r. outer Within the circular region. To accurately represent this range, an out-of-domain region indicator matrix is ​​defined.

[0099] The extra-area region indication matrix is ​​essentially a two-dimensional Boolean matrix with dimensions identical to the generated uniform grid. The value of each element in the extra-area region indication matrix (corresponding to a point in the grid) explicitly indicates whether that point is located inside or above the boundary of the defined total monitoring range (i.e., the extra-area region).

[0100] Define the out-of-domain region indicator matrix M outer (x,y) are as follows:

[0101]

[0102] Where (x,y) represents the planar coordinates of any point in the grid, r outer This is the preset inscribed circle radius. The I(·) in the function is an indicator function, defined as follows: the function value is 1 when the condition within the parentheses is met; otherwise, the function value is 0.

[0103] Step 4: Calculation of the AOA sector region indicator matrix

[0104] To use angle of arrival (AOA) information for region discrimination, it is necessary to calculate the distance of each point in the grid relative to each positioning station S. j The theoretical AOA. Assume the current processing is of the s-th location site S. s Its plane coordinates are For any point in the grid, first calculate the distance S from that grid point to the current location station. s The relative vector. The horizontal component Δx and the vertical component Δy of this relative vector are respectively:

[0105]

[0106] The grid points relative to the current positioning station S are calculated using the horizontal coordinate component Δx and the vertical coordinate component Δy. s The theoretical AOA, denoted as θ grid(x,y) The theoretical AOA is defined as the angle from the positive vertical axis to the direction of the relative vector, rotating clockwise, and ranges from 0° to 360°. The calculation formula is:

[0107]

[0108] in, It calculates the angle (in radians) between the relative vector and the positive horizontal axis, by multiplying by... Convert to degrees. Since AOA is calculated starting from the square of the vertical axis and in a clockwise direction, this included angle needs to be subtracted from 90°, and the result should be modulo 360° to ensure that the angle is within the range of 0° to 360°.

[0109] Calculate the current location station S s Measurement of AOA value θ s measured Relative to the grid point and the current location station S s Theoretical AOA value θ grid The angle difference Δθ between (x, y) s .

[0110] Δθ s =|((θ) s measured -θ grid (x,y)+180°)(mod360°))-180°|

[0111] It uses the difference between two angles θ s measured -θ grid(x, y) plus 180° modulo 360°, to normalize the difference to the range [0°, 360°). Then, subtract 180° from the result to map it to the interval (-180°, 180°] to get the signed shortest angle difference. Finally, take the absolute value to ensure that Δθ s is non-negative and ranges between 0° and 180°, effectively handling the periodicity of angles.

[0112] For each positioning site S j , the angle difference Δθ s between the calculated and the preset AOA robustness range θ error (50° in this example) is calculated. s The AOA sector region indicator matrix M for the current positioning site S outer is generated. outer If the distance between (x, y) and the origin is less than or equal to the preset radius r outer , then the corresponding element of M s (x, y) is 1 (true), indicating that the point is within the total detection range; otherwise, it is 0 (false), indicating that the point is outside the total detection range. The calculation formula of the AOA sector region indicator matrix is:

[0113]

[0114] where I(Δθ error ≤ θ error ) is an indicator function that takes the value 1 when the angle difference Δθ is within the allowed error range θ outer , and 0 otherwise. Multiplying the result of this indicator function with the out-of-domain region indicator matrix M s (x, y) ensures that only the grid points that satisfy both the AOA error condition and are located in the out-of-domain region have a value of 1 in the AOA sector region indicator matrix for the current positioning site S

[0115] Step 5: Generation of the common region indicator matrix for multiple sites

[0116] After calculating the AOA sector region indicator matrix M for each site, the next step is to determine the intersection region of the AOA sector regions. The intersection region represents the set of grid points that satisfy the AOA measurement error range for all sites. The intersection region is represented using the intersection region indicator matrix M intersect .

[0117] First, the current intersection region indicator matrix M′ intersect is initialized to the out-of-domain region indicator matrix M outer , which means that the initial intersection region is the total monitoring range.

[0118] Then, iterate through all the location sites S. j The current intersection region indicator matrix M′ intersect AOA sector area indication matrix for each location site Perform a logical AND operation (i.e., find the intersection). The logical AND operation ensures that only grid points that are simultaneously within the current intersection region and the AOA sector of the current site are retained in the updated intersection region.

[0119] By analyzing all location sites S j AOA sector region indicator matrix The intersection operation is performed sequentially, and the final intersection region indicator matrix M is obtained. intersect This indicates that all location stations S j The common area of ​​the AOA sector region, that is, the area that satisfies the conditions of all stations S j The set of grid points for the AOA measurement error range.

[0120] Step Six: Region Determination Rules

[0121] After calculating all stations S j Intersection region indicator matrix M intersect Then, based on the relationship between this intersection area and the preset internal area, it is determined whether the target is located in the "internal area" or the "external area". The internal area is the defined inner circular area, and the external area is the defined annular area between the inner and outer circles.

[0122] First, generate an inner circular region indicator matrix M. inner (x,y) indicates whether the coordinates (x,y) of the grid point are located inside the inner circle or on the boundary;

[0123] For cases where the positioning result is unknown, the target location is determined by forming a cone-shaped area using the angle of arrival (AOA). Each positioning station S j The AOA is considered as an extent, which forms a sector region in space. The vertices of the sector region are located at the positioning station S. j The location and sector opening angle are determined by the preset AOA robustness range θ. error Decision. Thus, each location station S j The measured values ​​form a possible target location area.

[0124]

[0125] Where, r inner It is the preset inner circle radius, and I(·) is still the indicator function.

[0126] Next, calculate the proportion of the intersection region of the AOA sectors within the inner circle, denoted as R.inner , the calculation formula is:

[0127]

[0128] In the formula, the numerator M intersect represents the number of grid points located in the intersection region of AOA sectors and inside the inner circle. The denominator M inner represents the total number of grid points in the intersection region of AOA sectors. intersect

[0129] Finally, the calculated proportion R inner is compared with a preset determination threshold t, thereby determining the final region determination result, and in this embodiment, t = 0.55. The determination rule is as follows:

[0130]

[0131] If the proportion R inner of the intersection region of AOA sectors in the inner circle is greater than or equal to 0.55 (this threshold value can be adjusted according to actual needs), it is determined that the target is located in the target region; otherwise, it is determined that the target is located outside the target region.

[0132] A target region identification system for a complex electromagnetic environment comprises:

[0133] A preprocessing module maps a target region on the earth's surface to a two-dimensional plane by plane projection to form a plane coordinate system, records the center of the target region as a calculation reference point, moves the target region in the plane coordinate system so that the calculation reference point coincides with the coordinate origin, and is used to realize step 1 of the method.

[0134] A grid discretization module discretizes the moved target region into uniform grids, and the grid coverage is a square region with the origin as the center and the inscribed circle radius r outer , and is used to realize step 2 of the method.

[0135] A matrix generation module takes a circular region with the origin as the center and r outer as the radius as the out-of-domain region, i.e., the total detection range, generates an out-of-domain region indication matrix, generates an angle of arrival sector region indication matrix based on the out-of-domain indication matrix, and generates an intersection region indication matrix based on the angle of arrival sector region indication matrix, and is used to realize steps 3-5 of the method.

[0136] A region determination module calculates the proportion of the intersection region of angles of arrival sectors in the target region based on the intersection region indication matrix, sets a determination threshold, and determines that the target is located in the target region if the proportion is greater than or equal to the threshold, otherwise, it is determined that the target is located outside the target region, and is used to realize step 6 of the method. ​

[0137] A target area recognition device for a complex electromagnetic environment, comprising:

[0138] a memory for storing a computer program implementing a target area recognition method for a complex electromagnetic environment;

[0139] a processor for implementing the target area recognition method for a complex electromagnetic environment when executing the computer program.

[0140] Experimental analysis

[0141] To comprehensively verify the performance of the target area recognition method for a complex electromagnetic environment proposed by the present application, the present application compares and analyzes it with the AOA positioning BRPLE algorithm. The AOA positioning BRPLE algorithm is a mature bias cancellation pseudo-linear estimation (BRPLE) method, mainly used for accurate three-dimensional AOA point positioning. Although its core goal is to provide high-precision position estimation, its positioning result can be converted into area judgment (i.e. judging whether the estimated point falls within a specific area), thereby serving as an effective benchmark for the present area recognition task.

[0142] The target area recognition method for a complex electromagnetic environment proposed by the present application directly faces the area judgment task, and by expanding the AOA measurement range of each positioning site S j Setting the AOA error range θ error , the AOA measurement range of each positioning site S j is expanded to a positive and negative θ error degree conical area centered on the measured AOA, and the overlap of the intersection of the conical areas of multiple positioning sites S j and the target area is analyzed to achieve direct judgment of the target area.

[0143] In the target area classification under single measurement, the algorithm performance mainly reflects in the accuracy (Accuracy). The present application mainly compares and analyzes the present application and the AOA positioning BRPLE algorithm from the influence factor AOA measurement error.

[0144] The feasibility, effectiveness and performance of the present application are verified through the following simulation experiment.

[0145] It is assumed that the positioning scene is constructed in a two-dimensional Cartesian coordinate system with the origin (0, 0) as the center. The scene contains two concentric circular areas, with the inner circle radius being 400 (meters) and the outer circle radius being 1000 (meters).

[0146] The number of positioning sites is 2, which are uniformly distributed on the boundary of the inner circle. The target to be positioned Randomly distributed throughout the scene. To simulate and verify the region recognition performance of this invention under different spatial distributions, we set up 2000 targets in this experiment. These targets T to be located k The targets are uniformly distributed within a predefined region to ensure the statistical realism and universality of the simulation results, thereby enabling a comprehensive evaluation of the region identification accuracy of this invention under different target location conditions. Specifically, the following three target distribution strategies are considered: Strategy 1, some targets are located inside the inner circular region, while the remaining targets are distributed within the annular region between the inner and outer circles; Strategy 2, all targets are located inside the inner circular region; Strategy 3, outer field concentration: all targets are distributed within the annular region between the inner and outer circles.

[0147] Each location station S j The received data from the target T to be located k The signal is used to measure the Angle of Arrival (AOA). Station S j To target T k The true AOA is defined as the angle between the vector pointing from the site to the target and the reference direction (0 degrees with true north as the reference, increasing clockwise), denoted as θ. true,j,k In the simulation, the AOA measurement value will be affected by random errors. These errors are set to a positive and negative symmetrical distribution centered on the true AOA. Specifically, the AOA measurement error range is set to start from 0 degrees and increase in increments of 5 degrees to 30 degrees.

[0148] This scenario aims to simulate the challenge of region identification of targets under varying spatial distributions and measurement uncertainties. Typical scenarios include... Figure 2 As shown. For clarity, Figure 2 Only 20 targets T to be located were depicted. k To avoid 2000 targets T to be located k Visually too dense, it affects observation and identification.

[0149] Fig. 3(a) corresponds to the scenario where targets are only distributed in the inner region (Strategy 2), while Fig. 3(b) shows the scenario where targets are only distributed in the outer region (Strategy 3). The horizontal axis of both figures represents the range of AOA measurement error (positive and negative degrees), and the vertical axis represents the accuracy of region identification (%). The blue solid line represents the present invention (area-based region identification technique with AOA cone error range, hereinafter referred to as the area method since the area of intersection is used to determine the region to which the target belongs), and the orange dashed line represents the AOA positioning method (BRPLE). As shown in Fig. 3(a), when the AOA measurement error gradually increases from 0 degrees to 30 degrees, the accuracy of both algorithms decreases as the AOA measurement error increases in this scenario. When there is no error (0 degrees), the accuracy of the AOA positioning method (BRPLE) is close to 100%, slightly higher than that of the present invention (about 92%). However, as the measurement error increases, the present invention exhibits significant robustness, with a more gradual decrease in accuracy. When the measurement error reaches 30 degrees, the accuracy of the present invention remains above 80%, while the accuracy of the AOA positioning method (BRPLE) drops sharply to about 60%, showing sensitivity to larger measurement errors. This indicates that when targets are concentrated in the inner region, although the AOA positioning method (BRPLE) performs well under ideal conditions, its performance is greatly affected by measurement errors, and the present invention exhibits significantly higher accuracy and robustness under different measurement errors. Specifically, even in the extreme case where the AOA measurement error range reaches ±30°±30°, the present invention still maintains an accuracy of about 81%, which is much higher than the about 60% of the traditional AOA positioning method.

[0150] Figure 4 For the performance comparison of the algorithm when targets are uniformly distributed in the inner and outer regions, under ideal error-free conditions, the AOA positioning method (BRPLE) exhibits an accuracy close to 100%, indicating that under precise measurement, the AOA positioning-based method can accurately identify the target region. The accuracy of the present invention is also close to 95% when there is no error, although it is slightly lower than that of the AOA positioning method (BRPLE), but it is still at a high level.

[0151] As the AOA measurement error increases, the region identification accuracy of both algorithms shows a downward trend, which is expected, as measurement uncertainty reduces the reliability of identification. Robustness of the present invention: The blue solid line represents the present invention, which exhibits stronger robustness. Its accuracy decline trend is very gradual; even when the AOA measurement error reaches a large value of 30 degrees, its accuracy remains above 85%. This indicates that the present invention has low sensitivity to measurement errors and maintains good stability in practical applications facing inaccurate measurements. Sensitivity of the "AOA Positioning Method (BRPLE)": The orange dashed line represents the "AOA Positioning Method (BRPLE)," which is more sensitive to measurement errors. Although it performs excellently when there are no errors, its accuracy decreases more significantly with increasing errors. When the measurement error reaches 30 degrees, its accuracy has significantly decreased to around 70%. The present invention (blue solid line in the figure) also exhibits excellent robustness and consistently high accuracy in scenarios where the target is uniformly distributed inside and outside. Although its accuracy is slightly lower than that of the AOA positioning method (approximately 96%) when there is no error, its performance decline is much slower as the AOA measurement error increases. Even when the measurement error reaches ±30°, the accuracy of this invention can still be maintained at approximately 87%, significantly higher than the approximately 70% of the "AOA positioning method (BRPLE)". This fully demonstrates that this invention can provide a more stable and reliable area identification capability in complex electromagnetic environments and in the face of unavoidable AOA measurement errors.

Claims

1. A target area identification method for complex electromagnetic environments, characterized in that, Includes the following steps: Step 1: Map the target area of ​​the positioning scene onto a two-dimensional plane to form a planar coordinate system through planar projection. Record the center of the target area as the calculation reference point. Move the target area under the planar coordinate system so that the calculation reference point coincides with the origin of the coordinate system. Step 2: Discretize the moved target region into a uniform grid, with the grid covering the origin and an inscribed circle of radius r. outer A square region; Step 3, with the origin as the center, r outer A circular region with a radius of is taken as the outer region, i.e. the total detection range, and an outer region indication matrix is ​​generated; Step 4: Generate the arrival angle sector region indication matrix based on the outside indication matrix; Step 5: Generate the intersection region indication matrix based on the arrival angle sector region indication matrix; Step 6: Calculate the proportion of the intersection region of the arrival corner sector within the target area based on the intersection region indicator matrix, and set a judgment threshold. If the proportion is greater than or equal to the threshold, the target is determined to be within the target area; otherwise, the target is determined to be outside the target area.

2. The method according to claim 1, characterized in that, The positioning scenario includes N S Location stations and N T Individual target to be located Indicates location station S j coordinate, Indicates the target T to be located k Coordinates are defined by two concentric circular regions. The center of the inner circle is the calculation reference point, and the inner circle is the target region with radius R. in The outer radius is R out ; Location station S j The target T is evenly distributed on the inner circle boundary. k Randomly distributed throughout the scene; Each location station S j The received data from the target T to be located k The signal is used to measure the angle of arrival, and the positioning station S is located. j To locate the target T k The true angle of arrival is defined as the angle from the location station S j Point to target T k The angle between the vector and the reference direction is denoted as θ. true,j,k ; Define measurement error as ∈ j,k The measurement error ∈ j,k It follows a mean of 0 and a standard deviation of σ. AOA A Gaussian distribution, or in [-Δθ] max ,+Δθ max If the locations are evenly distributed within the range, then the positioning stations S j Treating the target T k Angle of arrival measurement θ meas,j,k Represented as: i meas,j,k =θ true,j,k +∈ j,k Where, ∈ j,k The measurement error has a maximum value of Δθ. max Spend.

3. The method according to claim 1, characterized in that, In step 1, the latitude and longitude of the reference point are calculated as (λ). c ,φ c The coordinates after projection onto the plane are (x...). c ,y c The coordinates of any point with latitude and longitude (λ, φ) after translation are: Where R = 6378137m is the Earth's semi-major axis.

4. The method according to claim 1, characterized in that, In step 2, the mesh coverage is centered at the origin and has an inscribed circle radius of r. outer A square region, the coordinates (x, y) of each point in the grid. m ,y n It is determined by the following formula: in, N res The resolution of the grid represents the number of grid points in the horizontal and vertical directions of the planar coordinate system; N is generated using these two parameters. res ×N res A uniform grid, covering from -r outer to r outer The range of the horizontal and vertical coordinates.

5. The method according to claim 1, characterized in that, In step 3, the extra-domain region indication matrix M outer (x,y) is defined as follows: Where (x,y) represents the planar coordinates of any point in the grid, r outer It is the preset inscribed circle radius. I(·) in the function is an indicator function, which is defined as follows: when the condition inside the parentheses is met, the function value is 1; otherwise, the function value is 0.

6. The method according to claim 1, characterized in that, Step 4 specifically includes the following steps: Calculate the direction from any grid point to the current location station S s The relative vector, whose abscissa component Δx and ordinate component Δy: Where, x m ,y n Represents the coordinates of any grid point. Indicates the current location station S s Plane coordinates; Calculate the position of any grid point relative to the current location station S using the x-axis component Δx and the y-axis component Δy. s The theoretical angle of arrival, denoted as θ grid(x,y) : The theoretical angle of arrival is defined as the angle from the positive vertical axis to the direction of the relative vector, ranging from 0° to 360°. The angle between the relative vector and the positive direction of the horizontal axis; Calculate the current location station S s The measured angle value θ w measured Relative to the grid point and the current location station S s The theoretical angle of arrival θ grid The angle difference Δθ between (x, y) s : Dth s =|((θs measured -θ grid (x,y)+180°)(mod360°))-180°| Based on the calculated location station S j angular difference Δθ s and the preset robustness range of the angle of arrival θ error Generate the current location station S s Arrival angle sector area indication matrix If the extra-domain region indicator matrix M outer The distance from (x, y) to the origin is less than or equal to the preset radius r. outer Then M outer The corresponding element of (x,y) is 1, indicating that the point is within the total detection range; otherwise, it is 0, indicating that the point is outside the total detection range; the formula for calculating the arrival sector region indication matrix is: Where I(Δθ) s ≤θ error ) is an indicator function.

7. The method according to claim 1, characterized in that, Step 5 specifically includes the following steps: Current location station S s The intersection region of the reached corner sector regions is the current intersection region, and the current intersection region is used to indicate the matrix M′. intersect This means that, firstly, the current intersection region indicator matrix M′ is... intersect Initialized as an out-of-domain region indicator matrix M outer ; Then, iterate through all the location sites S. j The current intersection region indicator matrix M′ intersect With the arrival angle sector area indication matrix of each positioning station Perform logical AND operation; By analyzing all location sites S j Arrival angle sector area indication matrix The intersection operation is performed sequentially, and the final intersection region indicator matrix M is obtained. intersect This indicates that all location stations S j The common area of ​​the arrival angle sector region, that is, the area that satisfies all stations S j The set of grid points within the range of the angle of arrival measurement error.

8. The method according to claim 1, characterized in that, Step 6 specifically includes the following steps: First, generate an inner circular region indicator matrix M. inner (x,y) indicates whether the coordinates (x,y) of the grid point are located inside the inner circle or on the boundary; Where, r inner It is the preset inner circle radius, and I(·) is still an indicator function; Next, calculate the proportion of the intersection region of the reaching corner sectors within the inner circle, denoted as R. inner The calculation formula is: In the formula, molecule M intersect ∩M inner This represents the number of grid points that simultaneously reside within the intersection region of the arrival corner sector and the interior of the inner circle; the denominator M intersect This indicates the total number of grid points reaching the intersection area of ​​the corner sectors; Finally, based on the calculated ratio R inner The final region discrimination result is determined by comparing it with a preset judgment threshold t. The judgment rules are as follows: If the proportion R of the intersection area of ​​the corner sectors within the inner circle is... inner If the value is greater than or equal to t, the target is determined to be within the target area; otherwise, the target is determined to be outside the target area.

9. A target area identification system for complex electromagnetic environments based on the method of any one of claims 1-8, characterized in that, include: The preprocessing module maps the target area on the Earth's surface onto a two-dimensional plane to form a planar coordinate system through planar projection. The center of the target area is recorded as the calculation reference point. The target area under the planar coordinate system is moved so that the calculation reference point coincides with the origin of the coordinate system. The mesh discretization module discretizes the moved target region into a uniform mesh, with the mesh coverage centered at the origin and containing an inscribed circle of radius r. outer A square region; The matrix generation module will use the origin as the center and r outer A circular region with a radius of is defined as the outer region, i.e., the total detection range, and an outer region indicator matrix is ​​generated; an arrival corner sector region indicator matrix is ​​generated based on the outer region indicator matrix; and an intersection region indicator matrix is ​​generated based on the arrival corner sector region indicator matrix. The region determination module calculates the proportion of the intersection region of the arrival corner sector within the target region based on the intersection region indicator matrix, and sets a determination threshold. If the proportion is greater than or equal to the threshold, the target is determined to be within the target region; otherwise, the target is determined to be outside the target region.

10. A target area identification device for complex electromagnetic environments, characterized in that, include: Memory: Used to store a computer program for implementing the target area identification method for complex electromagnetic environments as described in any one of claims 1-8; Processor: Used to implement the target area identification method for complex electromagnetic environments as described in any one of claims 1-8 when executing the computer program.

Images