Clustered hybrid accumulation optimization method for multi-uav radar target signal detection

Abstract

The application discloses a kind of multi-unmanned aerial vehicle radar target signal detection's cluster hybrid accumulation optimization method, applied to radar signal technical field, for multi-unmanned aerial vehicle radar detection process in part, unmanned aerial vehicle spatial distribution sparse, cannot realize full coherent accumulation, and full non-coherent accumulation signal-to-noise ratio gain is lower Problem;The application constructs cluster hybrid accumulation processing architecture (intra-cluster coherent accumulation and inter-cluster non-coherent accumulation), the amount of calculation in the calculation hybrid accumulation process is analyzed with hybrid accumulation signal-to-noise ratio gain, and according to the two indexes, a cluster hybrid accumulation optimization model is established, and then a cluster processing method based on meta-heuristic bidirectional is designed, so as to obtain the optimized cluster processing result and realize the effective accumulation of echo signal, and finally realize target detection through CFAR detector.

Description

Technical Field

[0001] This invention belongs to the field of radar signal technology, and specifically relates to a radar echo signal hybrid accumulation technology. Background Technology

[0002] Multi-UAV (Unmanned Aerial Vehicle) radar systems, based on UAV platforms, utilize multiple UAV-borne radars to simultaneously receive signals from a single transmitter, achieving the accumulation and fusion of multiple echo signals. Improving the detection performance of multi-UAV radar systems hinges on achieving high-gain accumulation of echoes from multiple UAV-borne radars; simultaneously, it's necessary to minimize the computational load required for accumulation processing to enhance real-time detection. Signal accumulation methods can be broadly categorized into coherent and non-coherent accumulation. Coherent accumulation yields higher accumulation gain but requires more computation and a certain level of echo correlation. In UAV radar systems, some radar nodes are close together, resulting in strong target echo correlation, making coherent accumulation suitable. Conversely, some nodes are far apart, leading to weak echo correlation, making non-coherent accumulation preferable. Therefore, to obtain optimal accumulation gain, radar nodes need to be clustered, ensuring high correlation between echoes from nodes within a cluster for easier coherent accumulation. Inter-cluster echo correlation is lower, making non-coherent accumulation preferable.

[0003] M. Wang et al. proposed an inter-channel coherent accumulation method that can be used for intra-cluster coherent accumulation. To achieve non-coherent signal accumulation, C. Wang et al. proposed a CS-based inter-channel non-coherent accumulation algorithm that can be used for inter-cluster non-coherent accumulation. However, both of these methods neglect the problem of how to cluster nodes to achieve optimized signal clustering accumulation.

[0004] Several algorithms have been studied for node clustering. Fahad designed a clustering method based on the Grey Wolf algorithm to improve the stability of vehicle ad hoc networks during communication. Aadil designed a clustering method based on the K-Means algorithm, which extended the lifespan of UAV networks and reduced routing overhead. However, these methods focus more on communication stability or the longest possible UAV network lifespan, without considering the performance indicators and constraints of radar detection, and cannot be directly applied to the field of radar signal processing. Therefore, proposing a clustering optimization method for the mixed accumulation of radar echo signals is of great significance. Summary of the Invention

[0005] To address the problem of efficient accumulation of echo signals from multiple UAV-borne radars, this invention proposes a clustered hybrid accumulation optimization method for target signal detection from multiple UAV-borne radars, which achieves good results in improving the signal-to-noise ratio gain of hybrid accumulation and reducing computational load.

[0006] The technical solution adopted in this invention is: a clustered hybrid accumulation optimization method for multi-UAV radar target signal detection, comprising:

[0007] S1. Construct a clustered hybrid accumulation architecture to perform coherent accumulation of signals received by radar nodes within a cluster and non-coherent accumulation of signals between clusters;

[0008] S2. Perform pulse compression processing and multi-pulse accumulation on the echo signals received by multi-UAV onboard radar;

[0009] S3. Calculate the computational cost and signal-to-noise ratio gain of the clustering and mixing accumulation process of radar echo signals;

[0010] S4. Using the computational cost of the hybrid accumulation process and the signal-to-noise ratio gain of the hybrid accumulation as indicators, construct a clustering optimization model;

[0011] S5. Design a metaheuristic bidirectional clustering method to solve the clustering optimization problem and obtain the optimized clustering results;

[0012] S6. Perform clustered hybrid accumulation processing based on the clustering results to obtain the output results after hybrid accumulation of multiple UAV-borne radar echoes;

[0013] S7. Use the output results obtained from clustered hybrid accumulation to perform CFAR detection, and finally obtain the detection results.

[0014] The beneficial effects of this invention are as follows: This invention constructs a clustered hybrid accumulation processing architecture (intra-cluster coherent accumulation and inter-cluster non-coherent accumulation), analyzes the computational complexity and hybrid accumulation signal-to-noise ratio gain in the process of calculating hybrid accumulation, and establishes a clustered hybrid accumulation optimization model based on these two indicators. Then, a bidirectional clustering processing method based on metaheuristics is designed to obtain the optimized clustering processing results and achieve effective accumulation of echo signals. Finally, target detection is achieved through a CFAR detector. Attached Figure Description

[0015] Figure 1 This is a flowchart of an embodiment of the present invention.

[0016] Figure 2 For hybrid accumulation processing architecture;

[0017] Figure 3 Here is a flowchart of the meta-heuristic bidirectional clustering algorithm;

[0018] Figure 4 This indicates the initial position of the UAV-borne radar.

[0019] Figure 5 The iterative curve of the metaheuristic bidirectional clustering algorithm;

[0020] Figure 6 This is the clustering result of the UAV-borne radar;

[0021] in, Figure 6 (a) shows the clustering results of the metaheuristic bidirectional clustering algorithm. Figure 6 (b) shows the clustering results when the number of cluster heads is set to 3. Figure 6 (c) shows the clustering results when the number of cluster heads is set to 6. Figure 6 (d) shows the clustering results when the number of cluster heads is set to 9;

[0022] Figure 7 This is a bar chart comparing the performance of the metaheuristic bidirectional clustering algorithm and the K-means algorithm of this invention.

[0023] Figure 8 This is for detecting performance curves. Detailed Implementation

[0024] This invention is verified using Matlab simulation experiments. The correctness and effectiveness of this invention are verified on the scientific computing software Matlab R2021b. The embodiments of this invention are further described below with reference to the accompanying drawings.

[0025] like Figure 1 As shown, this invention proposes a clustered hybrid accumulation optimization method for multi-UAV radar target signal detection, which is implemented through the following process:

[0026] S1. Construct a clustered hybrid accumulation architecture to perform coherent accumulation of signals received by radar nodes within a cluster and non-coherent accumulation of signals between clusters;

[0027] S2. Perform pulse compression processing and multi-pulse accumulation on the echo signals received by multi-UAV onboard radar;

[0028] S3. Establish a model to calculate the computational cost and accumulation gain of the signal in clustered hybrid accumulation;

[0029] S4. Using the computational cost of the hybrid accumulation process and the signal-to-noise ratio gain of the hybrid accumulation as indicators, construct a clustering optimization model;

[0030] S5. Design a metaheuristic bidirectional clustering method to solve the clustering optimization model and obtain the optimized clustering results;

[0031] S6. Perform clustered fusion accumulation based on the clustering results to obtain the signal after fusion accumulation;

[0032] S7. Use the signals obtained from clustered and mixed accumulation to perform CFAR (Constant False Alarm Rate Detector) detection, and finally obtain the detection results.

[0033] Step S1: In a multi-UAV radar system consisting of one transmitter and N UAV nodes, the UAV nodes act as receivers of radar signals, collaboratively receiving and processing signals to improve detection performance. For example... Figure 2 As shown, based on a multi-UAV radar system, this invention constructs a clustered hybrid accumulation processing architecture. If the clustering method obtained by solving the clustering model in step S4 results in N UAVs being divided into M clusters, where N is greater than or equal to M, and the i-th cluster contains n... i There are 10 drone nodes. The echo signal received by the j-th drone in the i-th cluster is... The signal is obtained by accumulating between pulses. Then, coherent accumulation of signals is performed on the UAVs in the i-th cluster to obtain the signal. Non-coherent accumulation is performed between clusters, and CFAR detection is finally performed based on the accumulation results.

[0034] Step S2, in the multi-UAV radar system, the external radiation source transmits the signal as follows:

[0035] s(t)=u(t)exp(j2πf c t)

[0036] Where u(t) is the baseband signal, t is the fast time, and f c This is the carrier frequency for the transmitted signal.

[0037] During the observation time T of the multi-UAV-borne radar, the radar receives a total of K pulses. The echo signal of the j-th UAV in the i-th cluster receiving the k-th pulse can be expressed as:

[0038]

[0039] Among them, A ij This represents the complex amplitude of the echo signal reaching the j-th UAV receiving node in the i-th cluster after reflection from the target, where k = 1, 2, 3, ..., N, and represents the k-th pulse. After down-conversion, the echo signal is:

[0040]

[0041] The initial radial distance between the external radiation source and the target is The initial radial distance of the j-th UAV in the i-th cluster relative to the target is: The radial velocity v of the external radiation source relative to the target e The initial radial velocity of the j-th UAV in the i-th cluster relative to the target is v. ij The distance the signal emitted by the external radiation source travels to reach the target is:

[0042]

[0043] Among them, Tp This represents the pulse repetition time.

[0044] The distance the external radiation source signal travels from the target to the UAV is:

[0045]

[0046] Among them, R ij v is the radial distance of the target relative to the external radiation source. UVA Let be the radial velocity of the target relative to the external radiation source. Let τ be the time delay τ from signal transmission to reception. ij (k) is:

[0047]

[0048] Where c represents the speed of light.

[0049] The signal obtained after matched filtering of the echo signal is:

[0050]

[0051] Where ξ represents the fast time symbol in the integral, and h(t) is the matched filter of the transmitted signal.

[0052] The echo from the matched filter is processed by FFT along the slow time dimension to obtain the multi-pulse accumulation result s. ij p (f,k).

[0053] Step 3: Calculate the computational cost and accumulation gain of the signal in clustered hybrid accumulation. During the hybrid accumulation process, when performing intra-cluster channel alignment, the distance alignment search cost is... The search volume for Doppler alignment is The computational cost of phase alignment between the searched range offset and Doppler offset is c. p Then the computational cost of coherent accumulation of the i-th cluster with the j-th node as the reference can be expressed as:

[0054]

[0055] Where, n i Let be the number of nodes in the i-th cluster. In the M clusters, starting with the m-th cluster... k Using the cluster head as a reference, phase alignment is not required when performing noncoherent accumulation between clusters. The computational cost can be expressed as:

[0056]

[0057] In hybrid accumulation, N nodes are divided into M clusters, and the computational cost of hybrid accumulation is...

[0058]

[0059] Clustering method {n1,n2,…,n m The hybrid accumulation gain under} can be expressed as

[0060]

[0061] (1-0.140lg(N ncl )+0.018310(lg(N ncl )) 2 )

[0062] Among them, P D For the required detection probability, P fa N represents the false alarm rate. ncl The equivalent total number of nodes under the clustering scheme {n1,n2,…,nm}:

[0063]

[0064] Where, f(n) i ,P fa ,P D The expression for ) is:

[0065]

[0066] The expression for ζ is:

[0067]

[0068] The expression for ψ is:

[0069]

[0070] Step 4: An optimization model was established based on two metrics: computational complexity of hybrid accumulation and signal-to-noise ratio gain. The gain G of hybrid accumulation... a The weighted summation of the mixed accumulation computation CA: ω1CA+ω2G a As the objective function, the maximum distance between two node echo signals satisfying the coherent condition is used as the intra-cluster node distance constraint to construct an optimization model. The optimization model is expressed as:

[0071] max W=ω1CA+ω2G a

[0072]

[0073] k = 1, 2, 3, ..., M

[0074] i,j=1,2,3,......,n k

[0075] Where ω1 is the weight of the mixed accumulation computation index and ω2 is the weight of the mixed accumulation signal-to-noise ratio gain index, satisfying ω1+ω2=1. Let M be the distance between the i-th drone and the j-th drone within the k-th cluster, and M be the number of drone clusters. k Let K be the number of drones in the k-th cluster. The maximum distance between the echo signals from two nodes that satisfies the coherent condition is given. The hybrid accumulation signal-to-noise ratio gain and computational complexity in the objective function are normalized indices.

[0076] Step 5: For the optimized model of clustered hybrid accumulation, a metaheuristic bidirectional clustering algorithm is designed to achieve better clustering results. This algorithm takes the position, velocity, and prior information of the target from the UAV-borne radar as input, such as... Figure 3 As shown, the algorithm mainly consists of three parts: initialization, fusion processing, and splitting processing. The initialization operation obtains the number of UAV-borne radar nodes N and generates the clustering result initialization matrix C.

[0077] C = [1,2,3,…,N] T

[0078] The superscript T indicates transpose;

[0079] In the clustering matrix C, the element in the m-th row represents the number of the UAV-borne radar node that is assigned to the m-th cluster. At this point, the initialized clustering matrix indicates that the current N UAV nodes each form a cluster.

[0080] The fusion process selects two clusters from all clusters for fusion, operating in the direction of rounding down to zero. The splitting process selects one cluster from all clusters and splits its nodes into independent clusters, operating in the direction of rounding down to zero. The algorithm randomly performs fusion and splitting processes during iteration, iterating to obtain a better clustering result. The specific operations of the fusion and splitting processes are as follows:

[0081] Step 5.1 Fusion Processing:

[0082] In the fusion process, to improve the algorithm's optimization ability, two clusters are selected from all clusters for fusion. The fusion target value matrix is ​​calculated based on the objective function, and the cluster target value change matrix is ​​further calculated.

[0083] △Va(i,j)=Va(i,j)-Va′

[0084] Where Va(i,j) represents the clustering target value after the fusion of the i-th and j-th clusters, and Va′ represents the clustering target value in the previous generation.

[0085] When the clustering target change matrix contains elements greater than 0, the fusion probability matrix is:

[0086]

[0087] Otherwise, set the fusion probability matrix as follows:

[0088]

[0089] Among them, V + The matrix of positive elements of the clustering objective value change matrix is ​​represented as:

[0090]

[0091] M' represents the current cluster number.

[0092] The fusion probability matrix calculates the probability of each fusion method being selected in the current iteration. During the fusion process, the two clusters to be fused in this process are selected based on the fusion probability matrix. The fusion probability matrix is ​​then transformed into a fusion probability distribution matrix:

[0093]

[0094] Where J is the index of the last element of the second dimension in the fusion probability distribution matrix, and a random number rand uniformly distributed between 0 and 1 is generated to search for the index [i]. r ,j r ]satisfy:

[0095]

[0096] Then the i-th r Cluster j and the jth r Each cluster is merged into one cluster, and the clustering result is compared with the previous generation's clustering result to retain the better clustering result.

[0097] Step 5.2 Splitting process:

[0098] In the splitting process, several clusters containing more than 2 points are randomly selected for splitting. Following a greedy strategy, the current best clustering method is retained throughout the iterations, while the current generation's best solution set is excluded from splitting. The splitting process is based on the quality of each cluster, with poorer clusters having a higher probability of being split. The index values ​​of all M clusters in the current clustering method are calculated using the objective function, resulting in an index value vector V for the M clusters. T The index value vector reflects the degree of good or bad of each cluster under the evaluation system given by the objective function. The index value vector is then normalized.

[0099]

[0100] Calculate the split probability vector based on the index value vector:

[0101]

[0102] Among them, there are Calculate the split probability distribution vector

[0103]

[0104] Generate a random number rand that is uniformly distributed between 0 and 1, and look up the index m. r satisfy

[0105]

[0106] Then the mth r Each node in a cluster is split into a separate cluster.

[0107] After the above processing, a clustering result can be obtained, which achieves good results in improving the hybrid accumulation signal-to-noise ratio gain and reducing the amount of computation.

[0108] Step 6: Perform signal clustering, mixing, and accumulation based on the final clustering results. This involves accumulating the echo signals s received by nodes within the cluster. ij p Coherent accumulation is performed on (f,k) to obtain the intra-cluster accumulation signal. Non-coherent accumulation is then performed on the intra-cluster accumulation signals of each cluster to obtain the final mixed accumulation result.

[0109] Step 7: Perform CFAR detection based on the mixed accumulation results to obtain the final detection result.

[0110] Figure 4 The current position of the UAV-borne radar in the embodiment is given. Figure 5 Iterative curves of the clustering results calculated by the metaheuristic bidirectional clustering method are presented. Figure 4 , Figure 5 The corresponding parameters are shown in Table 1. As can be seen from the figure, the meta-heuristic bidirectional clustering algorithm can achieve good clustering results through iteration. Figure 6 The results of UAV-borne radar clustering are shown, among which, Figure 6 (a) shows the clustering results obtained by the metaheuristic bidirectional clustering method. Figure 6 (b) shows the clustering results when the number of cluster heads is set to 3. Figure 6 (c) shows the clustering results when the number of cluster heads is set to 6. Figure 6 (d) shows the clustering results when the number of cluster heads is set to 9. Figure 7 A bar chart comparing the clustered hybrid accumulation performance is provided. Figure 8The detection curves corresponding to different clustering results obtained by the metaheuristic bidirectional clustering method and the K-means algorithm are presented. Those skilled in the art will understand that, at a given signal-to-noise ratio, a higher detection probability and a higher signal accumulation gain indicate better performance of the proposed method.

[0111] Table 1 Parameters

[0112]

[0113] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Various modifications and variations can be made to the invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the invention should be included within the scope of the claims of the invention.

Claims

1. A clustered hybrid accumulation optimization method for multi-UAV radar target signal detection, characterized in that, include: S1. Construct a clustered hybrid accumulation architecture, perform coherent accumulation of signals received by radar nodes within a cluster, and non-coherent accumulation of signals between clusters; S2. Perform pulse compression processing and multi-pulse accumulation on the echo signals received by multi-UAV onboard radar; S3. Calculate the computational cost and signal-to-noise ratio gain of the clustering and mixing accumulation process of radar echo signals; S4. Using the computational cost of the hybrid accumulation process and the signal-to-noise ratio gain of the hybrid accumulation as indicators, construct a clustering optimization model; S5. Design a metaheuristic bidirectional clustering method to solve the clustering optimization problem and obtain the optimized clustering results; The implementation process of step S5 includes: S51. Obtain the number of UAV-borne radar nodes through initialization operations. Generate clustering result initialization matrix : ； Clustering matrix The Middle The row element indicates that it is divided into the first row. The clustering matrix represents the number of the UAV-borne radar node in each cluster. Each drone node forms a cluster; S52. Based on the fusion probability matrix, select the two clusters to be fused in this processing. The process for determining the fusion probability is as follows: The fusion target value matrix is ​​calculated based on the objective function, and the cluster target value change matrix is ​​further calculated: ； in, Indicates the first and the The cluster target value after merging individual clusters, This represents the target value of the previous generation's clustering; The fusion probability is determined based on the target value change matrix; S53. Randomly select several clusters containing more than 2 points and split them according to the splitting probability; The process of determining the split probability is as follows: All the current clustering methods... Each cluster's index value was calculated using the objective function, resulting in information about... The index value vector of each cluster ,according to Determine the splitting probability; S6. Perform clustered hybrid accumulation processing based on the clustering results to obtain the output results after hybrid accumulation of multiple UAV-borne radar echoes; S7. CFAR detection is performed using the output results obtained from clustered hybrid accumulation, and the final detection result is obtained.

2. The clustered hybrid accumulation optimization method for multi-UAV radar target signal detection according to claim 1, characterized in that, The computational complexity of the mixed accumulation process in step S3 is expressed as follows: ； in, Indicates the number of clusters. Indicates the first The computational cost of coherent accumulation of individual clusters, This represents the computational cost of performing noncoherent accumulation between clusters.

3. The clustered hybrid accumulation optimization method for multi-UAV radar target signal detection according to claim 2, characterized in that, The hybrid accumulation signal-to-noise ratio gain in step S3 is expressed as: ； in, For the required detection probability, False alarm rate This represents the equivalent total number of nodes under the current clustering method.

4. The clustered hybrid accumulation optimization method for multi-UAV radar target signal detection according to claim 3, characterized in that, The clustering optimization model described in step S4 is expressed as follows: ； in, For the weighting of the mixed accumulation calculation index, Weights for the hybrid accumulation signal-to-noise ratio gain metric. , For the first Within the cluster, the first The drone and the first The distance of the drone For the first Number of drones per cluster The maximum distance between the echo signals of two nodes that satisfy the coherent condition.

5. The clustered hybrid accumulation optimization method for multi-UAV radar target signal detection according to claim 4, characterized in that, The fusion probability is determined based on the target value change matrix, specifically: When the clustering target change matrix contains elements greater than 0, the fusion probability matrix is: ； Otherwise, set the fusion probability matrix as follows: ； in, This is the matrix of positive elements of the clustering target value change matrix.

6. The clustered hybrid accumulation optimization method for multi-UAV radar target signal detection according to claim 5, characterized in that, The formula for calculating the splitting probability is: ； in, express The normalized result.

Images