Navigation satellite signal coming direction simulation method based on hybrid clustering algorithm

Abstract

The invention discloses a navigation satellite signal coming direction simulation method based on a hybrid clustering algorithm. The navigation satellite signal coming direction simulation method comprises the steps that 1, satellite coordinate values in a geodetic coordinate system are calculated; 2, projecting to an east-north-sky coordinate system, and calculating the distance between any two satellites; 3, setting a cluster number, an azimuth angle threshold value and a distance threshold value; step 4, clustering is carried out by using a DBSCAN algorithm; 5, calculating horizontal and vertical coordinate values of the central point of each cluster; step 6, clustering each satellite point by using a K-means algorithm; step 7, determining an optimized clustering result; 8, judging whether the maximum value of the azimuth angle difference value is smaller than or equal to the azimuth angle threshold value or not, and if yes, entering the step 9; otherwise, updating and optimizing the clustering result by using a K-means algorithm, and returning to the step 8; and step 9, outputting a current optimization clustering result. According to the invention, the incident angle difference between the false satellite signal and the real satellite signal can be significantly reduced, the state of the navigation satellite can be simulated more truly, and the possibility of identifying the false satellite signal is reduced.

Description

Technical Field

[0001] The invention belongs to the field of electronic information technology and relates to a navigation satellite signal simulation method based on a hybrid clustering algorithm. Background Art

[0002] With the continuous development of array antennas and array signal processing technology, satellite navigation anti-interference technology based on array antennas has gradually become popular. In order to better test the anti-interference ability of array antennas, it has become a new demand to build a simulation space with almost no noise signals and the navigation signal state is close to the real satellite signal state as a test environment. As a specially designed electromagnetic isolation environment, the "Starry Sky" microwave darkroom has its internal walls, ceilings and floors covered with high-efficiency electromagnetic wave absorbing materials, which can create an ideal test space with almost no electromagnetic interference. In addition, multiple signal transmitting antennas are installed on the top of the darkroom, which can simulate navigation satellite signals from different directions, and provide the necessary conditions for key performance tests such as array antenna beamforming and space-time anti-interference. However, if an independent transmitting antenna is deployed for each visible satellite, it will greatly increase the cost and difficulty of implementation. Therefore, the common practice is to install a few groups of antennas on the top of the darkroom in small quantities, and each transmitting antenna transmits several satellite signals. How many antennas should be deployed and which navigation satellite signals each antenna simulates to transmit have a great impact on the simulation effect of the direction of the navigation signal. However, there is no reasonable calculation method for the above problems, and there is a large room for optimization. The existing "Starry Sky" microwave darkroom transmitting antenna layout lacks scientificity, and usually uses a simple calculation of angle difference to determine the distribution state. Therefore, there is an urgent need to develop a more scientific and efficient navigation satellite signal approach simulation method to improve the microwave darkroom's ability to simulate the real satellite signal environment. At the same time, this method can also be used in the field of navigation deception technology for array antenna receivers to reduce the possibility of being identified by the array antenna signal approach estimation technology. Summary of the invention

[0003] The purpose of the present invention is to propose a navigation satellite signal arrival simulation method based on a hybrid clustering algorithm to solve the problem that the existing microwave darkroom test method is difficult to accurately simulate the actual satellite signal arrival state.

[0004] In order to achieve the above object, the present invention adopts the following technical solution:

[0005] A navigation satellite signal simulation method based on a hybrid clustering algorithm specifically comprises the following steps:

[0006] Step 1: According to the needs of the simulation scene, input the required simulation time and the current coordinate value of the satellite navigation receiver; obtain and analyze the ephemeris data through the satellite navigation receiver, online or offline, extract the satellite number and orbit parameter information, and calculate the satellite coordinate value in the geodetic coordinate system;

[0007] Step 2: According to the current coordinate value of the satellite navigation receiver in step 1 and the satellite coordinate value in the earth-fixed coordinate system, the distance between any two satellites in the two-dimensional Cartesian coordinate system is obtained, including the distance of the azimuth angle and the pitch angle;

[0008] Step 3: According to the satellite signal in the simulated scene, the number of clusters k, the azimuth angle threshold Th and the distance threshold Eps are set according to the simulation requirements and the number of false signal transmitting antennas that can be provided;

[0009] Step 4: Use the DBSCAN algorithm based on density distribution to cluster each satellite data point using the distances of the azimuth and elevation angles between any two satellites calculated in step 2. The clustering results include the number of clusters and each cluster. Secondly, determine whether the number of clusters is equal to the number of clusters k set in step 3. If yes, execute step 5. Otherwise, use the binary search method to update the distance threshold Eps and repeat step 4.

[0010] Step 5: According to the clustering results of step 4, calculate the horizontal and vertical coordinate values ​​of the center point of each cluster;

[0011] Step 6: Take the center point of each cluster obtained in step 5 as the initial center point, use the K-means algorithm, and cluster each satellite data point using the distance of the azimuth and elevation angle between any two satellites calculated in step 2 to obtain the clustering result;

[0012] Step 7: Calculate the average intra-cluster distances of the clustering results of step 4 and step 6 respectively, and select the clustering result corresponding to the smaller average intra-cluster distance as the optimized clustering result; the optimized clustering result includes the clustering of navigation satellite data points, the horizontal and vertical coordinates of the center point of each cluster, the average distance from each satellite point to the center point of its cluster, and the maximum difference between the elevation angle and azimuth angle of the satellite in each cluster;

[0013] Step 8: Calculate the maximum azimuth difference of each cluster in the optimized clustering result, and determine the maximum difference, recorded as In the formula, α mi With α mjIndicates the azimuth of the two satellite signals in the mth cluster relative to the satellite navigation receiver; determines whether the maximum azimuth difference value M is less than or equal to the azimuth threshold value Th set in step 3, if so, proceeds to step 9; otherwise, uses the K-means algorithm to re-cluster the cluster corresponding to the maximum azimuth value M, merges the re-clustered cluster with all other clusters in the current optimized clustering result as a new optimized clustering result, and returns to execute step 8;

[0014] Step 9: Output the current optimized clustering result as the final clustering result.

[0015] Furthermore, step 2 specifically includes the following sub-steps:

[0016] Step 21, convert the current coordinate value of the satellite navigation receiver in step 1 and the satellite coordinate value in the earth-fixed coordinate system to a northeast sky coordinate system with the position of the satellite navigation receiver as the origin; calculate the azimuth and elevation angle of each satellite relative to the satellite navigation receiver in this coordinate system, and then convert it to a two-dimensional Cartesian coordinate system for representation; specifically, assuming that the azimuth angle of the i-th satellite is α i , the pitch angle is β i , with an azimuth angle α i As the horizontal axis, the pitch angle β i As the ordinate, project the direction angle of the i-th satellite signal into the two-dimensional Cartesian coordinate system;

[0017] Step 22: In a two-dimensional Cartesian coordinate system, let the satellite data set to be clustered be X = {x i |x i ∈R 2 ,i=1,2,…,n}, where x i =(α i ,β i ), n is the number of visible satellites relative to the satellite navigation receiver; calculate x i With x j The distance d(x i ,x j )=||x i -x j ||,x i 、x j represents the azimuth and elevation information of any two representative satellites in the satellite data set X to be clustered, and ||·|| represents the 2-norm.

[0018] Furthermore, in step 22, the two data points x are calculated in a two-dimensional Cartesian coordinate system. i 、x jFor the pitch angle, the norm of the two satellites is taken directly after the difference of the pitch angles; for the azimuth angle, if the azimuth angle difference between the two satellites is less than or equal to 180°, the distance between the azimuth angles of the two satellites is d(α i ,α j )=||α i -α j ||, otherwise, d(α i ,α j )=360-||α i -α j ||.

[0019] Furthermore, in step 5, for the pitch angle, the average value of the maximum and minimum pitch angles of all satellites in each cluster is calculated as the vertical coordinate of the center point of the cluster; for the azimuth angle, according to the difference between the two azimuth angles of all satellites in each cluster, the two satellite points with the largest azimuth angle difference are taken as the left and right boundary points of the cluster, and then the horizontal coordinate of its center point is calculated based on the left and right boundary points.

[0020] In a second aspect, the present invention provides an electronic device, the electronic device comprising:

[0021] A memory for storing executable instructions;

[0022] A processor, configured to implement the method according to any one of claims 1 to 4 when executing the executable instructions or computer programs stored in the memory.

[0023] In a third aspect, the present invention provides a computer-readable storage medium storing executable instructions or a computer program, wherein the executable instructions, when executed by a processor, implement the method according to any one of claims 1 to 4.

[0024] Compared with the prior art, the present invention has the following technical effects:

[0025] The present invention can classify satellite signals from similar satellites into the same cluster according to the simulation requirements of navigation satellite signals, determine the best false signal emission angle, and reduce the difference in the incident angle between real satellite signals and false satellite signals. At the same time, when implementing cluster division, it combines K-means and DBSCAN algorithms, organically integrates the two and adds threshold restrictions. Specifically, it has the following advantages:

[0026] (1) This method automatically clusters the directions of navigation satellite signals and calculates the center point of each cluster for arranging the transmitting antenna of the false signal. The calculation results of this method can significantly reduce the difference in the incident angle between the false satellite signal and the real satellite signal. On the one hand, it can more realistically simulate the state of the navigation satellite and provide a realistic environment for array antenna testing. On the other hand, it can reduce the possibility of false satellite signals being identified by array antenna technology.

[0027] (2) The DBSCAN algorithm and the K-means algorithm are organically combined. On the one hand, the problem that the DBSCAN algorithm is sensitive to the neighborhood radius and the minimum number of samples in the cluster and cannot directly specify the number of clusters is solved; on the other hand, the problem that the K-means algorithm is sensitive to the position of the initial cluster center is solved. Based on the processing of the clustering module, the present invention can provide a transmitting antenna layout and signal transmission distribution scheme that conforms to the actual situation according to the number of existing antennas and the accuracy required by the experiment, and meet the requirements of the simulation accuracy of the incoming angle under different experimental tasks.

[0028] (3) The present invention is mainly implemented through data preprocessing and hybrid clustering algorithm. The method does not require the deployment of additional hardware equipment, and the method process is streamlined, the calculation is efficient, the demand for computing resources is low, and it is easy to deploy and use. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is the cluster analysis process;

[0030] Figure 2 is the distribution of experimental satellites;

[0031] Figure 3 This is the clustering effect obtained using the K-means algorithm;

[0032] Figure 4 This is the clustering effect obtained using the DBSCAN algorithm;

[0033] Figure 5 It is the clustering effect obtained using the hybrid clustering algorithm;

[0034] Figure 6 This is a comparison of the clustering effects of the three algorithms.

[0035] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments. DETAILED DESCRIPTION

[0036] like Figure 1 As shown, the present invention proposes a navigation satellite signal simulation method based on a hybrid clustering algorithm, which specifically includes the following steps:

[0037] Step 1: According to the needs of the simulation scene, input the required simulation time and the current coordinate value of the satellite navigation receiver; obtain and analyze the ephemeris data through the satellite navigation receiver, online or offline, extract the satellite number and orbit parameter information, and calculate the satellite coordinate value in the earth-fixed coordinate system.

[0038] Step 2: According to the current coordinate value of the satellite navigation receiver in step 1 and the satellite coordinate value in the earth-fixed coordinate system, the distance between any two satellites in the two-dimensional Cartesian coordinate system is obtained, including the distance of the azimuth angle and the pitch angle. Specifically, the following sub-steps are included:

[0039] Step 21, convert the current coordinate value of the satellite navigation receiver in step 1 and the satellite coordinate value in the earth-fixed coordinate system to the northeast sky coordinate system with the position of the satellite navigation receiver as the origin; calculate the azimuth and elevation angle of each satellite relative to the satellite navigation receiver in this coordinate system, and then convert it to a two-dimensional Cartesian coordinate system for representation. Specifically, assuming that the azimuth angle of the i-th satellite is α i , the pitch angle is β i , with an azimuth angle α i As the horizontal axis, the pitch angle β i As the ordinate, project the direction angle of the i-th satellite signal into the two-dimensional Cartesian coordinate system;

[0040] Step 22: In a two-dimensional Cartesian coordinate system, let the satellite data set to be clustered be X = {x i |x i ∈R 2 ,i=1,2,…,n}, where x i =(α i ,β i ), n is the number of visible satellites relative to the satellite navigation receiver; calculate x i With x j The distance d(x i ,x j )=||x i -x j ||,x i 、x j represents the azimuth and elevation information of any two representative satellites in the satellite data set X to be clustered, and ||·|| represents the 2-norm.

[0041] Specifically, find two data points x in a two-dimensional Cartesian coordinate system. i 、x jFor the elevation angle, the angle and angle difference range is 0°~90°, and the norm can be directly obtained by taking the difference between the elevation angles of the two satellites; for the azimuth angle, the azimuth angle range is -180°~180°, and the angle difference range is 0°~180°. Therefore, when calculating the distance for the azimuth angle, if the azimuth angle difference between the two satellites is less than or equal to 180°, then the distance between the azimuth angles of the two satellites is d(α i ,α j )=||α i -α j ||, otherwise, d(α i ,α j )=360-||α i -α j ||.

[0042] Step 3: According to the satellite signal in the simulated scene, the number of clusters k, the azimuth angle threshold Th and the distance threshold Eps are set according to the need for angular simulation and the number of false signal transmitting antennas that can be provided.

[0043] Step 4: Use the DBSCAN algorithm based on density distribution to cluster each satellite data point using the distance between any two satellites calculated in step 2 (including the distance in azimuth and elevation). The clustering result includes the number of clusters and each cluster. Secondly, determine whether the number of clusters is equal to the number of clusters k set in step 3. If so, execute step 5. Otherwise, use the binary search method to update the distance threshold Eps and repeat step 4.

[0044] Step 5: Based on the clustering results of step 4, calculate the horizontal and vertical coordinate values ​​of the center point of each cluster.

[0045] Specifically, for the pitch angle, the average value of the maximum and minimum pitch angles of all satellites in each cluster is calculated as the vertical coordinate of the center point of the cluster; for the azimuth angle, according to the difference between the two azimuth angles of all satellites in each cluster, the two satellite points with the largest azimuth angle difference are taken as the left and right boundary points of the cluster, and then the horizontal coordinate of its center point is calculated based on the left and right boundary points.

[0046] Step 6: Take the center point of each cluster obtained in step 5 as the initial center point, use the K-means algorithm, and use the distance between any two satellites calculated in step 2 (including the distance of azimuth and elevation angle) to cluster the satellite data points to obtain the clustering results.

[0047] Step 7: Calculate the average intra-cluster distance of the clustering results of step 4 and step 6 respectively, and select the clustering result corresponding to the smaller average intra-cluster distance as the optimized clustering result. The optimized clustering result includes the clustering of navigation satellite data points, the horizontal and vertical coordinates of the center point of each cluster, the average distance from each satellite point to the center point of its cluster, and the maximum difference in the elevation angle and azimuth angle of the satellite in each cluster.

[0048] Step 8: Calculate the maximum azimuth difference of each cluster in the optimized clustering result, and determine the maximum difference, recorded as In the formula, α mi With α mj Represents the azimuth of the two satellite signals in the mth cluster relative to the satellite navigation receiver; determine whether the maximum azimuth difference M is less than or equal to the azimuth threshold Th set in step 3, if so, proceed to step 9; otherwise, use the K-means algorithm to re-cluster the cluster corresponding to the maximum azimuth M, merge the re-clustered cluster with all other clusters in the current optimized clustering result as a new optimized clustering result, and return to execute step 8.

[0049] Step 9: Output the current optimized clustering result as the final clustering result.

[0050] In order to prove the feasibility and effectiveness of the present invention, a simulation experiment was conducted using 120 sets of ephemeris data, and experimental results were obtained.

[0051] At present, the mainstream array antenna is four-element or seven-element. In the experiment, the cluster number k is set to 5, the azimuth angle threshold Th is set to 72°, and the distance threshold Eps is set to 100. K-means, DBSCAN, and the method of the present invention are used to cluster 120 groups of different ephemeris data. The representative results are as follows: Figure 2-Figure 5 shown. Figure 2 The satellite distribution of a certain set of data is shown, with dots of different colors representing different satellites. Figures 3 to 5 The clustering effects of K-means, DBSCAN, and the method proposed in the present invention are shown respectively. The dots of the same color represent the same cluster of satellites, and the red cross "×" represents the corresponding deployment azimuth of the false signal transmitting antenna. It can be seen that K-means and DBSCAN can both successfully cluster into 5 clusters. However, in the K-means clustering results, the satellite corresponding to the azimuth angle of -136.9° and the pitch angle of 21.6° is actually far away from the two satellites on its right, and is mistakenly divided into one cluster. In the DBSCAN clustering results, the 5 satellites with azimuth angles between -65° and 45° have a large angle span and are mistakenly divided into the same cluster. The hybrid clustering algorithm of the present invention effectively solves the above problems, and the clustering effect is significantly improved.

[0052] In order to quantitatively analyze the effect of the proposed clustering method, the average intra-cluster distance of the clustering results of 120 groups of data using the three methods is statistically analyzed as follows: Figure 6 shown. Figure 6 The orange curve represents the average intra-cluster distance of the clustering results of the K-means algorithm, the green curve represents the DBSCAN algorithm, and the blue curve represents the hybrid clustering algorithm. The smaller the value, the higher the similarity within the cluster, that is, the better the clustering effect. As can be seen from the figure below, compared with the K-means and DBSCAN algorithms, the hybrid clustering algorithm has a smaller average intra-cluster distance and higher intra-cluster similarity in the 120 groups of data, and the satellite signals in the cluster are closer in angle. Therefore, the false signal emitted according to the hybrid clustering result is more realistic than the real satellite signal, which proves the effectiveness of the proposed method.

Claims

1. A navigation satellite signal simulation method based on a hybrid clustering algorithm, characterized in that: The specific steps include: Step 1: According to the needs of the simulation scene, input the required simulation time and the current coordinate value of the satellite navigation receiver; obtain and analyze the ephemeris data through the satellite navigation receiver, online or offline, extract the satellite number and orbit parameter information, and calculate the satellite coordinate value in the geodetic coordinate system; Step 2: According to the current coordinate value of the satellite navigation receiver in step 1 and the satellite coordinate value in the earth-fixed coordinate system, the distance between any two satellites in the two-dimensional Cartesian coordinate system is obtained, including the distance of the azimuth angle and the pitch angle; Step 3: According to the satellite signal in the simulated scene, the number of clusters k, the azimuth angle threshold Th and the distance threshold Eps are set according to the simulation requirements and the number of false signal transmitting antennas that can be provided; Step 4: Use the DBSCAN algorithm based on density distribution to cluster each satellite data point using the distances of the azimuth and elevation angles between any two satellites calculated in step 2. The clustering results include the number of clusters and each cluster. Secondly, determine whether the number of clusters is equal to the number of clusters k set in step 3. If yes, execute step 5. Otherwise, use the binary search method to update the distance threshold Eps and repeat step 4. Step 5: According to the clustering results of step 4, calculate the horizontal and vertical coordinate values ​​of the center point of each cluster; Step 6: Take the center point of each cluster obtained in step 5 as the initial center point, use the K-means algorithm, and cluster each satellite data point using the distance of the azimuth and elevation angle between any two satellites calculated in step 2 to obtain the clustering result; Step 7: Calculate the average intra-cluster distances of the clustering results of step 4 and step 6 respectively, and select the clustering result corresponding to the smaller average intra-cluster distance as the optimized clustering result; the optimized clustering result includes the clustering of navigation satellite data points, the horizontal and vertical coordinates of the center point of each cluster, the average distance from each satellite point to the center point of its cluster, and the maximum difference between the elevation angle and azimuth angle of the satellite in each cluster; Step 8: Calculate the maximum azimuth difference of each cluster in the optimized clustering result, and determine the maximum difference, recorded as In the formula, α mi With α mj Indicates the azimuth of the two satellite signals in the mth cluster relative to the satellite navigation receiver; determines whether the maximum azimuth difference M is less than or equal to the azimuth threshold Th set in step 3, if so, proceeds to step 9; Otherwise, the cluster corresponding to the maximum azimuth angle M is re-clustered using the K-means algorithm, and the re-clustered cluster is merged with all other clusters in the current optimized clustering result as a new optimized clustering result, and the process returns to step 8; Step 9: Output the current optimized clustering result as the final clustering result.

2. The navigation satellite signal source simulation method based on a hybrid clustering algorithm as claimed in claim 1, characterized in that: Step 2 specifically includes the following sub-steps: Step 21, convert the current coordinate value of the satellite navigation receiver in step 1 and the satellite coordinate value in the earth-fixed coordinate system to a northeast sky coordinate system with the position of the satellite navigation receiver as the origin; calculate the azimuth and elevation angle of each satellite relative to the satellite navigation receiver in this coordinate system, and then convert it to a two-dimensional Cartesian coordinate system for representation; specifically, assuming that the azimuth angle of the i-th satellite is α i , the pitch angle is β i , with an azimuth angle α i As the horizontal axis, the pitch angle β i As the ordinate, project the direction angle of the i-th satellite signal into the two-dimensional Cartesian coordinate system; Step 22: In a two-dimensional Cartesian coordinate system, let the satellite data set to be clustered be X = {x i |x i ∈R 2 ,i=1,2,…,n}, where x i =(α i ,β i ), n is the number of visible satellites relative to the satellite navigation receiver; calculate x i With x j The distance d(x i ,x j )=||x i -x j ||,x i 、x j represents the azimuth and elevation information of any two representative satellites in the satellite data set X to be clustered, and ||·|| represents the 2-norm.

3. The navigation satellite signal source simulation method based on a hybrid clustering algorithm as claimed in claim 2, characterized in that: In step 22, find the two data points x in the two-dimensional Cartesian coordinate system. i 、x j For the pitch angle, the norm of the two satellites is taken directly after the difference of the pitch angles; for the azimuth angle, if the azimuth angle difference between the two satellites is less than or equal to 180°, the distance between the azimuth angles of the two satellites is d(α i ,α j )=||α i -α j ||, otherwise, d(α i ,α j )=360-||α i -α j ||.

4. The navigation satellite signal source simulation method based on a hybrid clustering algorithm as claimed in claim 1, characterized in that: In step 5, for the pitch angle, the average value of the maximum and minimum pitch angles of all satellites in each cluster is calculated as the ordinate of the center point of the cluster; For the azimuth, according to the difference between the two azimuths of all satellites in each cluster, the two satellite points with the largest azimuth difference are used as the left and right boundary points of the cluster, and then the horizontal coordinate of the center point is calculated based on the left and right boundary points.

5. An electronic device, characterized in that: The electronic device comprises: A memory for storing executable instructions; A processor, configured to implement the method according to any one of claims 1 to 4 when executing the executable instructions or computer programs stored in the memory.

6. A computer-readable storage medium storing executable instructions or a computer program, characterized in that: When the executable instructions are executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images