A pre-detection tracking method based on a cascade increment value function

Abstract

The application discloses a kind of detection front tracking methods based on cascade incremental value function, first utilize multi-channel data, based on three guide vector method to the radial velocity value of potential target is roughly estimated, then utilize three guide vector iterative method to the radial velocity value of potential target is searched accurately;On this basis, the construction of frame cascade incremental value function is completed;Finally, the real target track is obtained by backtracking multiple frame data;The method of the application can greatly improve the performance of weak target multi-frame detection front tracking, and provides an efficient approach for space, air and ground platform weak target detection.

Description

Technical Field

[0001] This invention belongs to the field of radar technology, specifically relating to a pre-detection tracking method, which can provide an efficient means of inter-frame noncoherent accumulation of targets for weak target detection. Background Technology

[0002] Based on the long-range detection requirements of early warning radars and the small radar cross section (RCS) of stealth targets, pre-detection tracking technology utilizes a multi-frame joint detection strategy, fully leveraging the advantages of short-term coherent accumulation within frames and long-term non-coherent accumulation between frames, providing a feasible approach for detecting weak targets. However, with the improvement of the stealth capabilities and maneuverability of new-generation aerospace targets, the effectiveness of existing early warning radars needs further enhancement.

[0003] The value function is a crucial factor affecting the performance of pre-detection tracking algorithms. Existing value function selection methods mainly fall into two categories: amplitude accumulation algorithms and likelihood function accumulation algorithms. The former utilizes the correlation between target motion frames to non-coherently accumulate target amplitude information across multiple frames; the latter uses a probability density function to construct a likelihood ratio detection quantity to improve the efficiency of multi-frame energy accumulation. While these value function selection methods are simple, they do not deeply explore the differences between potential target and background interference in each frame's echo, thus limiting their ability to improve the detection signal-to-noise ratio.

[0004] In summary, it is necessary to study novel value functions suitable for scenarios with lower signal-to-clutter ratios to further improve the detection performance of existing pre-detection tracking algorithms. Summary of the Invention

[0005] The purpose of this invention is to overcome the above-mentioned shortcomings and provide a pre-detection tracking method based on a cascaded incremental value function. First, using multi-channel data, a coarse estimate of the radial velocity value of a potential target is made based on a three-direction vector method. Then, a fine search for the radial velocity value of the potential target is performed using a three-direction vector iterative method. Based on this, the intra-frame cascaded incremental value function is constructed. Finally, the true target trajectory is obtained by backtracking through multiple frames of data. This invention can significantly improve the pre-detection tracking performance of weak targets across multiple frames, providing an efficient approach for weak target detection on air-space-ground platforms.

[0006] To achieve the above-mentioned objectives, the present invention provides the following technical solution:

[0007] A pre-detection tracking method based on cascaded incremental value functions includes the following steps:

[0008] S1 uses the three-guide vector method to search for the radial velocity of the potential target and obtains a coarse estimate of the radial velocity of the potential target based on the three guide vectors;

[0009] S2 iteratively updates the three guidance vectors and performs a fine search for the radial velocity value of the potential target based on the coarse estimate of the radial velocity of the potential target, thus obtaining the fine search result of the radial velocity of the potential target.

[0010] S3 constructs an intra-frame concatenated incremental value function based on the refined search results for the radial velocity of the potential target;

[0011] S4 constructs a detection-before-tracking algorithm based on dynamic programming based on the intra-frame concatenated incremental value function, and obtains the real target trajectory by backtracking multiple frames of data based on the dynamic programming-based detection-before-tracking algorithm.

[0012] Furthermore, in step S1, the coarse estimate of the potential target's radial velocity... for:

[0013] ;

[0014] in, This represents the clutter suppression weight vector based on the subspace projection algorithm. , Let be the orthogonal projection matrix of the clutter. For three steering vectors based on three independent variables, , This is a mixed echo signal of targets, clutter, and noise in the scene. This indicates the conjugate transpose. This represents the operation of finding the absolute value.

[0015] Furthermore, the three guiding vectors are defined as follows:

[0016]

[0017]

[0018] ;

[0019] The three independent variables are:

[0020] ;

[0021] in, , , The maximum unambiguous velocity in the airspace. The virtual part, For the spacing between array elements, The wavelength of radar electromagnetic waves, The number of array elements. Indicates transpose. For platform speed.

[0022] Furthermore, the specific method of step S2 is as follows:

[0023] S2.1 Calculation based on three guide vectors , and ;

[0024] The independent variables in the three guiding vectors are calculated as follows:

[0025] ;

[0026] S2.2 If , , The following conditions must be met:

[0027] [ , , ]=

[0028] make , Then return to step S2.1;

[0029] like , , The following conditions must be met:

[0030] [ , , ]=

[0031] make , Then return to step S2.1;

[0032] like , , Between

[0033] [ , , ]=

[0034] make , Then return to step S2.1;

[0035] S2.3 Repeat steps S2.1 to S2.2 until... ,Will , , Substituting the coarse estimate of the potential target's radial velocity into the result, we obtain the fine search result for the potential target's radial velocity. ;

[0036] in, This represents the clutter suppression weight vector based on the subspace projection algorithm. , Let be the orthogonal projection matrix of the clutter. For three steering vectors based on three independent variables, , This is a mixed echo signal of targets, clutter, and noise in the scene. This indicates the conjugate transpose. This represents the absolute value operation. This represents the operation of finding the maximum value.

[0037] Furthermore, in step S3, the expression for the intra-frame concatenated increment value function is:

[0038] ;

[0039] ;

[0040] ;

[0041] in, For clutter suppression increment, For orthogonal projection increments, , , Represents a constant. The sample covariance matrix is ​​obtained by the maximum likelihood estimation method. The target radial velocity is The guiding vector at that time, , For the fine search results of the radial velocity of potential targets, The virtual part, For the spacing between array elements, The wavelength of radar electromagnetic waves, The number of array elements. Indicates transpose. For platform speed, Let be the orthogonal projection matrix of the clutter. This is a mixed echo signal of targets, clutter, and noise in the scene. This indicates the conjugate transpose.

[0042] Furthermore, the pre-detection tracking algorithm based on dynamic programming is as follows:

[0043] ;

[0044] in, For potential targets Frame motion track, The first potential target Frame motion track, , for Frame observation data, This is a multi-frame cumulative value function. This is the multi-frame cumulative decision threshold obtained from the intra-frame concatenated increment value function.

[0045] Furthermore, when At that time, the inter-frame motion equation of the potential target is expressed as:

[0046] ;

[0047] in, Let be the potential target motion state transition matrix.

[0048] Furthermore, ;

[0049] in, The multi-frame cumulative value function is obtained from the intra-frame concatenated incremental value function. The second-order statistic, This represents the false alarm probability.

[0050] Compared with the prior art, the present invention has at least one of the following advantages:

[0051] (1) This invention combines the three-guided vector method with the cascaded incremental value function, and obtains the real target trajectory by backtracking multiple frames of data, which greatly improves the tracking performance of weak targets before multi-frame detection and provides an efficient way for weak target detection on air-space-ground platforms;

[0052] (2) This invention proposes a refined target radial velocity estimation method based on three guide vectors, which can estimate the target output value corresponding to different guide vectors ( , and The adaptive adjustment of the iterative guidance vector takes into account factors such as search computation and parameter estimation accuracy, and has high practical value.

[0053] (3) The present invention proposes a pre-detection tracking method based on cascaded incremental value functions. Instead of using simple amplitude or likelihood ratio functions to construct extended value functions, it integrates the advantages of clutter suppression increment and orthogonal projection increment. The new detection quantity can significantly improve the performance of multi-frame joint pre-detection tracking algorithm. Attached Figure Description

[0054] Figure 1This is a flowchart of the pre-detection tracking method based on cascaded incremental value functions according to the present invention;

[0055] Figure 2 The detection probabilities corresponding to different value functions in Embodiment 1 of the present invention are as follows: Figure 2 (a) represents the detection probability corresponding to the amplitude value function. Figure 2 (b) The detection probability corresponding to the space-time filter value function. Figure 2 (c) Cascaded incremental value function corresponding to detection probability. Detailed Implementation

[0056] The features and advantages of the present invention will become clearer and more explicit from the following detailed description.

[0057] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments. Although various aspects of embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated otherwise.

[0058] The application scenario of the pre-detection tracking method based on cascaded incremental value functions of this invention is as follows:

[0059] The cascaded incremental value function in this invention can be applied to the field of constant false alarm rate (CFAR) detection for weak radar targets. First, using multi-channel data, a coarse estimate of the target's radial velocity is performed based on the three-guided vector method. Then, the precise velocity of the target is searched using a three-guided vector iterative method. Based on this, the intra-frame cascaded incremental value function is constructed, thereby significantly improving the tracking performance of weak targets before multi-frame detection, such as... Figure 1 As shown, the implementation steps are as follows:

[0060] Step 1: Make a rough estimate of the radial velocity of the potential target.

[0061] First, the radial velocity of the potential target is searched using the three-guide vector method. The three guide vectors are defined as follows:

[0062] (1)

[0063] (2)

[0064] (3)

[0065] In the above formula

[0066] (4)

[0067] here , , The maximum unambiguous velocity in the airspace. Represents the spacing between array elements. The number of array elements, i.e. One channel, Indicates transpose. This indicates the wavelength of radar electromagnetic waves.

[0068] The radial velocity search method for potential targets, i.e., the coarse estimation results of the radial velocity of potential targets, are as follows:

[0069] (5)

[0070] in This represents a mixed echo signal containing targets, clutter, and noise in the scene. Represents the clutter suppression weight vector based on the subspace projection algorithm It is the orthogonal projection matrix of the clutter, representing the orthogonal complement space of the clutter. This indicates the conjugate transpose. This represents the operation of finding the absolute value.

[0071] Step 2 involves a fine search for the radial velocity values ​​of potential targets.

[0072] like , , Between

[0073] [ , , ]= (6)

[0074] make , Substitute into equation (4) in step 1 to... , , Update.

[0075] like , , Between

[0076] [ , , ]= (7)

[0077] make , Substitute into equation (4) in step 1 to... , , Update.

[0078] like , , Between

[0079] [ , , ]= (8)

[0080] make , Substitute into equation (4) in step 1 to... , , Update.

[0081] Repeat step 2 iteratively until the constraints are met. Next , , Substituting into equation (5) in step 1, we obtain the refined search results for the target. .

[0082] Step 3: Construct the intra-frame concatenated increment function.

[0083] Define the expressions for clutter suppression increment and orthogonal projection increment as follows:

[0084] (9)

[0085] (10)

[0086] In the formula , , Represents a constant. The sample covariance matrix is ​​obtained by the maximum likelihood estimation method. Based on equations (9) and (10), the expression for the intra-frame concatenated increment value function is as follows:

[0087] (11)

[0088] Step 4 involves backtracking multiple frames of data to obtain the actual target trajectory.

[0089] The target's inter-frame motion equation is expressed as:

[0090] (12)

[0091] in , For the number of observation frames, Let be the target motion state transition matrix.

[0092] Assuming the target Frame motion trajectory is The detection-before-tracking algorithm based on dynamic programming can be described as follows:

[0093] (13)

[0094] Constant false alarm threshold The setup method is as follows:

[0095] (14)

[0096] in This is the multi-frame cumulative decision threshold for the algorithm. Multi-frame cumulative value function The second-order statistic, for Frame observation data, This represents the false alarm probability. It is obtained based on the cascaded incremental value function within a single frame.

[0097] Example 1:

[0098] In this embodiment, simulation data is used to further illustrate the effects of the present invention:

[0099] The radar platform, target, clutter, and noise parameters for the observation scenario are set as follows:

[0100] Carrier platform speed =120m / s, radar carrier frequency The frequency is 10 GHz, and the antenna uses an element spacing of [missing information]. An 8-channel design with a depth of 0.125m allows for the construction of three steering vectors:

[0101] (1)

[0102] (2)

[0103] (3)

[0104] Then, steps 1 to 4 are performed to obtain the constant false alarm threshold. Based on this, the target signal-to-noise ratio versus detection probability curve is introduced to evaluate the detection performance of the algorithm, and the true radial velocity of the target relative to the radar platform is also considered. * The target echo fluctuation is 1 m / s, and the clutter sample size is 1000. The proportions of uniform clutter and non-uniform clutter are 80% and 20%, respectively, with corresponding clutter-to-noise ratios of 13 dB and 25 dB. Figure 2By comparing the target signal-to-noise ratio and detection probability curves based on different value functions, it can be seen that the detection probability based on the cascaded incremental value function is significantly higher than the detection probabilities corresponding to the amplitude value function and the space-time filtering value function.

[0105] Simulation analysis results: For the detection of weak targets, the pre-detection tracking method based on the cascaded incremental value function of this invention can effectively improve the detection probability of the target under the same false alarm rate.

[0106] The present invention has been described in detail above with reference to specific embodiments and exemplary examples; however, these descriptions should not be construed as limiting the present invention. Those skilled in the art will understand that various equivalent substitutions, modifications, or improvements can be made to the technical solutions and embodiments of the present invention without departing from the spirit and scope of the invention, and all such modifications and improvements fall within the scope of the present invention. The scope of protection of the present invention is defined by the appended claims.

[0107] The contents not described in detail in this specification are common knowledge to those skilled in the art.

Claims

1. A pre-detection tracking method based on cascaded incremental value functions, characterized in that, Includes the following steps: S1 uses the three-guide vector method to search for the radial velocity of the potential target and obtains a coarse estimate of the radial velocity of the potential target based on the three guide vectors; S2 iteratively updates the three guidance vectors and performs a fine search for the radial velocity value of the potential target based on the coarse estimate of the radial velocity of the potential target, thus obtaining the fine search result of the radial velocity of the potential target. S3 constructs an intra-frame concatenated incremental value function based on the refined search results for the radial velocity of the potential target; S4 constructs a detection-before-tracking algorithm based on dynamic programming based on the intra-frame concatenated incremental value function, and obtains the real target trajectory by backtracking multiple frames of data based on the dynamic programming-based detection-before-tracking algorithm. The specific method of step S2 is as follows: S2.1 Calculation based on three guide vectors , and ; The independent variables in the three guiding vectors are calculated as follows: ； S2.2 If , , The following conditions must be met: [ , , ]= make , Then return to step S2.1; like , , The following conditions must be met: [ , , ]= make , Then return to step S2.1; like , , Between [ , , ]= make , Then return to step S2.1; S2.3 Repeat steps S2.1 to S2.2 until... ,Will , , Substituting the coarse estimate of the potential target's radial velocity into the result, we obtain the fine search result for the potential target's radial velocity. ; in, This represents the clutter suppression weight vector based on the subspace projection algorithm. , Let be the orthogonal projection matrix of the clutter. For three steering vectors based on three independent variables, , This is a mixed echo signal of targets, clutter, and noise in the scene. This indicates the conjugate transpose. This represents the absolute value operation. This represents the operation of finding the maximum value; , , The maximum unambiguous velocity in the spatial domain; In step S3, the expression for the intra-frame concatenated increment function is: ； ； ； in, For clutter suppression increment, For orthogonal projection increments, , , Represents a constant. The sample covariance matrix is ​​obtained by the maximum likelihood estimation method. The target radial velocity is The guiding vector at that time, , For the fine search results of the radial velocity of potential targets, The virtual part, For the spacing between array elements, The wavelength of radar electromagnetic waves, The number of array elements. Indicates transpose. For platform speed, Let be the orthogonal projection matrix of the clutter. This is a mixed echo signal of targets, clutter, and noise in the scene. This indicates the conjugate transpose.

2. The pre-detection tracking method based on a cascaded incremental value function according to claim 1, characterized in that, In step S1, the coarse estimate of the radial velocity of the potential target for: ； in, This represents the clutter suppression weight vector based on the subspace projection algorithm. , Let be the orthogonal projection matrix of the clutter. For three steering vectors based on three independent variables, , This is a mixed echo signal of targets, clutter, and noise in the scene. This indicates the conjugate transpose. This represents the operation of finding the absolute value.

3. The pre-detection tracking method based on a cascaded incremental value function according to claim 2, characterized in that, The three guiding vectors are defined as follows: ； The three independent variables are: ； in, , , The maximum unambiguous velocity in the airspace. The virtual part, For the spacing between array elements, The wavelength of radar electromagnetic waves, The number of array elements. Indicates transpose. For platform speed.

4. The pre-detection tracking method based on a cascaded incremental value function according to claim 1, characterized in that, The pre-detection tracking algorithm based on dynamic programming is as follows: ； in, For potential targets Frame motion track, The first potential target Frame motion track, , for Frame observation data, This is a multi-frame cumulative value function. This is the multi-frame cumulative decision threshold obtained from the intra-frame concatenated increment value function.

5. The pre-detection tracking method based on a cascaded incremental value function according to claim 4, characterized in that, when At that time, the inter-frame motion equation of the potential target is expressed as: ； in, Let be the potential target motion state transition matrix.

6. The pre-detection tracking method based on a cascaded incremental value function according to claim 4, characterized in that, ； in, The multi-frame cumulative value function is obtained from the intra-frame concatenated incremental value function. The second-order statistic, This represents the probability of a false alarm.

Images