阅读说明：本技术 基于极化不变量和极化分解的海面角反射器干扰鉴别方法 (Sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition ) 是由 陈伯孝 郎思呈 于 2021-07-29 设计创作，主要内容包括：本发明提供的一种基于极化不变量和极化分解的海面角反射器干扰鉴别方法,参考目标舰船和角反射器的物理性质和海面实际环境差异,从能量方面利用目标舰船和角反射器的行列式模值和功率矩阵迹的差异,从实际海面情况影响上利用本征极化方向角,从能量随机性上利用强度熵,最终构造特征向量训练分类器,以提高分类器的分类准确性；之后利用分类器区分目标舰船和角反射器。因此本发明可以有效解决现有极化鉴别方法中存在的仿真模型设计与实际情况不匹配的情况,从而使得极化角度鉴别角反射器具有更高的精度和可行性。(The sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition provided by the invention refers to the physical properties of a target ship and a corner reflector and the difference of the actual sea surface environment, utilizes the difference of determinant mode values and power matrix traces of the target ship and the corner reflector in the aspect of energy, utilizes an intrinsic polarization direction angle in the aspect of influence of the actual sea surface condition and utilizes intensity entropy in the aspect of energy randomness, and finally constructs a feature vector training classifier so as to improve the classification accuracy of the classifier; the target vessel and the corner reflector are then distinguished using a classifier. Therefore, the method can effectively solve the problem that the design of a simulation model in the existing polarization identification method is not matched with the actual situation, so that the polarization angle identification corner reflector has higher precision and feasibility.)

1. A sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition is characterized by comprising the following steps:

step 1: receiving a polarization four-channel radar echo signal containing polarization information each time, which is returned after a radar system transmits signals for multiple times;

the polarized four channels are channels formed by transmitting and receiving signals by a horizontal polarized array and a vertical polarized array in the radar system;

step 2: adding the radar echo signal received each time as a radar echo signal of one period into a preset sample set;

and step 3: respectively calculating the polarization invariant and the polarization decomposition quantity of the peak value of each sample in the preset sample set;

and 4, step 4: forming a characteristic vector by the polarization invariant and the polarization decomposition quantity of each sample;

and 5: inputting the feature vector of each sample into a preset classifier, and training the classifier to obtain a trained classifier;

step 6: and classifying the returned echo signals by using a trained classifier so as to distinguish the corner reflector from the target.

2. The sea surface corner reflector interference discrimination method according to claim 1, wherein the polarized four channels include a horizontal transmit-receive channel HH, a horizontal transmit-vertical receive channel HV, a vertical transmit-horizontal receive channel VH, and a vertical transmit-vertical receive channel VV;

the horizontal transceiving channel HH is a channel which is transmitted by a horizontal polarization array and is formed by receiving signals of the horizontal polarization array; a horizontal transmitting and vertical receiving channel HV is a channel transmitted by the horizontal polarization array and formed by the vertical polarization array; a vertical transmitting and horizontal receiving channel VH is a channel which is transmitted by a vertical polarization array and is formed by receiving signals of the vertical polarization array; and the vertical transmitting and receiving channel VV is a channel which is transmitted by a vertical polarization array and is formed by a receiving vertical polarization array.

3. The sea surface corner reflector disturbance identification method according to claim 1, wherein the step 3 comprises:

step 31: for each sample in the set of samples, determining a peak value for the sample;

the peak value is an echo signal returned by the position where the corner reflector is located or an echo signal returned by the position where the target is located;

step 32: calculating determinant mode values at each sample peak;

step 33: calculating a power matrix trace at each sample peak;

step 34: calculating an intrinsic polarization direction angle at each sample peak;

step 35: for each sample, the determinant modulus value, the power matrix trace and the intrinsic polarization direction angle of the sample form a polarization invariant of the sample;

step 36: the polarization split amount of each sample is calculated.

4. The sea surface corner reflector disturbance identification method of claim 3,

the determinant mode value is expressed as:

|Δ|＝|det(S)|＝|s_HH+s_VV-s_HVs_VH|

the power matrix trace is represented as:

the intrinsic polarization orientation angle is expressed as:

wherein the content of the first and second substances,S_HHrepresenting the signal component of the horizontal transmit-receive channel, S_VVRepresenting the signal component of the vertical transmit-vertical receive path, S_HVRepresenting the signal component of the horizontal transmit-vertical receive path, S_VHSignal components representing the vertical transmit horizontal receive path are labeled with a conjugate.

5. The sea surface corner reflector disturbance identification method of claim 3, wherein the step 36 comprises:

calculating a coherence matrix at each sample peak;

for each sample, solving the Shannon entropy at the peak value by using the coherent matrix at the peak value of the sample;

the shannon entropy at each sample peak is decomposed into intensity entropies.

6. The sea surface corner reflector disturbance identification method of claim 5,

the coherence matrix is represented as:

the shannon entropy is expressed as:

the intensity entropy is expressed as:

wherein S is_HHRepresenting the signal component of the horizontal transmit-receive channel, S_VVRepresenting the signal component of the vertical transmit-vertical receive path, S_HVRepresenting the signal component of the horizontal transmit-vertical receive path, S_VHRepresenting the signal components of the vertical transmit-horizontal receive path, the superscript H representing the conjugate transpose,λ_iis the i-th eigenvalue, λ, of the coherence matrix T_kDenotes the same as_iThe kth eigenvalue of the coherence matrix T is represented, tr (T) represents the trace of the coherence matrix T.

7. The sea surface corner reflector disturbance identification method of claim 5, wherein the step 4 comprises:

and forming a characteristic vector by the determinant module value, the power matrix trace, the intrinsic polarization direction angle and the intensity entropy.

8. The sea surface corner reflector disturbance identification method according to claim 1, wherein after the step 5, the sea surface corner reflector disturbance identification method further comprises:

removing items formed by target interference caused by corner reflectors in the feature vectors to obtain the feature vectors after removal;

and retraining the classifier by using the feature vectors after the elimination.

Technical Field

The invention belongs to the technical field of radars, and particularly relates to a sea surface corner reflector interference identification method based on polarization invariants and polarization decomposition.

Background

The corner reflector interference is widely applied to the missile resisting guidance head of the ship due to the characteristics of large scattering sectional area, simple deployment, difficult distinction between radar echo and a target ship and the like, and the problem of how to effectively resist the corner reflector interference by the missile guidance head is a well-known problem at present. Currently, because the polarization characteristic of the corner reflector as a passive target is difficult to distinguish from a real target, the corner reflector countermeasure method using only polarization information is few, and most countermeasure methods use polarization information for auxiliary authentication.

The prior art records a ship and corner reflector identification method based on polarization decomposition, wherein echo signals received by a linear polarization base are projected to a circular polarization base by using Krogager decomposition, and feature vectors obtained by decomposition are input into an SVM for classification and identification, so that better identification rate can be obtained when the interference of corner reflectors in different postures is resisted.

In the prior art, the research on anti-corner reflector interference based on polarization decomposition is also recorded, the stability and the sensitivity of each posture of a target are determined by using Krogager decomposition under a circular polarization base and using front and rear frequency point information, a double-threshold discrimination target is set, and the recognition rate is good.

In the prior art, the Krogager decomposition is required to be used in the identification process of the anti-corner reflector, the target scattering characteristic is required to be non-time-varying by the decomposition method, and in practical situations, the target scattering characteristic is often influenced by sea clutter and changes along with the change of the radar carrier frequency along with the irradiation direction of the electromagnetic wave.

Disclosure of Invention

In order to solve the problems in the prior art, the invention provides a sea surface corner reflector interference identification method based on polarization invariants and polarization decomposition. The technical problem to be solved by the invention is realized by the following technical scheme:

the invention provides a sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition, which comprises the following steps:

step 1: receiving a polarization four-channel radar echo signal containing polarization information each time, which is returned after a radar system transmits signals for multiple times;

the four polarization channels are channels formed by horizontal polarization and vertical polarization array transceiving signals in the radar system; the polarized four channels comprise a horizontal transceiving channel HH, a horizontal transmitting and vertical receiving channel HV, a vertical transmitting and horizontal receiving channel VH and a vertical transmitting and vertical receiving channel VV; the horizontal transceiving channel HH is a channel which is transmitted by the horizontal polarization array and is formed by receiving signals of the horizontal polarization array; a horizontal transmitting and vertical receiving channel HV is a channel transmitted by the horizontal polarization array and formed by the vertical polarization array; a vertical transmitting and horizontal receiving channel VH is a channel which is transmitted by a vertical polarization array and is formed by receiving signals of the vertical polarization array; the vertical transmitting and vertical receiving channel VV is a channel which is transmitted by a vertical polarization array and is formed by a receiving vertical polarization array.

Step 2: adding the radar echo signal received each time as a radar echo signal of one period into a preset sample set;

and step 3: respectively calculating the polarization invariant and the polarization decomposition quantity of the peak value of each sample in a preset sample set;

and 4, step 4: forming a characteristic vector by the polarization invariant and the polarization decomposition quantity of each sample;

and 5: inputting the feature vector of each sample into a preset classifier, and training the classifier to obtain the trained classifier;

step 6: and classifying the returned echo signals by using a trained classifier so as to distinguish the corner reflector from the target.

Optionally, step 3 includes:

step 31: for each sample in the set of samples, determining a peak value for the sample;

wherein the peak value is an echo signal returned by the position of the corner reflector or an echo signal returned by the position of the target;

step 32: calculating determinant mode values at each sample peak;

step 33: calculating a power matrix trace at each sample peak;

step 34: calculating an intrinsic polarization direction angle at each sample peak;

step 35: for each sample, the determinant modulus value, the power matrix trace and the intrinsic polarization direction angle of the sample form a polarization invariant of the sample;

step 36: the polarization split amount of each sample is calculated.

Wherein the determinant mode value is expressed as:

|Δ|＝|det(S)|＝|s_HH+s_VV-s_HVs_VH|

the power matrix trace is represented as:

the intrinsic polarization orientation angle is expressed as:

wherein the content of the first and second substances,S_HHrepresenting the signal component of the horizontal transmit-receive channel, S_VVRepresenting the signal component of the vertical transmit-vertical receive path, S_HVRepresenting the signal component of the horizontal transmit-vertical receive path, S_VHSignal components representing the vertical transmit horizontal receive path are labeled with a conjugate.

Optionally, step 36 includes:

calculating a coherence matrix at each sample peak;

for each sample, solving the Shannon entropy at the peak value by using the coherent matrix at the peak value of the sample;

the shannon entropy at each sample peak is decomposed into intensity entropies.

The coherence matrix is represented as:

shannon entropy is expressed as:

the intensity entropy is expressed as:

wherein S is_HHRepresenting the signal component of the horizontal transmit-receive channel, S_VVRepresenting the signal component of the vertical transmit-vertical receive path, S_HVRepresenting the signal component of the horizontal transmit-vertical receive path, S_VHRepresenting the signal components of the vertical transmit-horizontal receive path, the superscript H representing the conjugate transpose,λ_iis the i-th eigenvalue, λ, of the coherence matrix T_kDenotes the same as_iThe kth eigenvalue of the coherence matrix T is represented, tr (T) represents the trace of the coherence matrix T.

Optionally, step 4 includes:

and forming the determinant module value, the power matrix trace, the eigen polarization direction angle and the intensity entropy into a feature vector.

Optionally, after step 5, the sea surface corner reflector disturbance identification method further includes:

removing items formed by target interference caused by the corner reflector in the feature vector to obtain the feature vector after removal;

and retraining the classifier by using the feature vectors after the elimination.

The sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition provided by the invention comprises the steps of considering the physical properties of a target ship and a corner reflector and the difference of the actual sea surface environment, utilizing the difference of determinant mode values and power matrix traces of the target ship and the corner reflector in the aspect of energy, utilizing an intrinsic polarization direction angle in the aspect of influence of the actual sea surface condition, utilizing intensity entropy in the aspect of energy randomness, and finally constructing a feature vector training classifier so as to improve the classification accuracy of the classifier; the target vessel and the corner reflector are then distinguished using a classifier. Therefore, the method can effectively solve the problem that the design of a simulation model in the existing polarization identification method is not matched with the actual situation, so that the polarization angle identification corner reflector has higher precision and feasibility.

The present invention will be described in further detail with reference to the accompanying drawings and examples.

Drawings

Fig. 1 is a schematic flow chart of a sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition according to an embodiment of the present invention;

fig. 2 is a time-frequency diagram of a single period of measured data in a sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition according to an embodiment of the present invention;

fig. 3 is a determinant module value comparison diagram of a corner reflector and a target ship in measured data in a sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition according to an embodiment of the present invention;

fig. 4 is a power matrix trace comparison diagram of a corner reflector and a target ship in measured data in a sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition according to an embodiment of the present invention;

fig. 5 is an intrinsic polarization direction angle comparison diagram of a corner reflector and a target ship in measured data in the sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition according to an embodiment of the present invention;

fig. 6 is a comparison graph of intensity entropy between a corner reflector and a target ship in measured data in a sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition according to an embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to specific examples, but the embodiments of the present invention are not limited thereto.

As shown in fig. 1, the method for identifying interference of sea surface corner reflector based on polarization invariant and polarization decomposition provided by the invention comprises:

step 1: receiving a polarization four-channel radar echo signal containing polarization information each time, which is returned after a radar system transmits signals for multiple times;

the four polarization channels are channels formed by horizontal polarization and vertical polarization array transceiving signals in the radar system; the polarized four channels comprise a horizontal transceiving channel HH, a horizontal transmitting and vertical receiving channel HV, a vertical transmitting and horizontal receiving channel VH and a vertical transmitting and vertical receiving channel VV; the horizontal transceiving channel HH is a channel which is transmitted by the horizontal polarization array and is formed by receiving signals of the horizontal polarization array; a horizontal transmitting and vertical receiving channel HV is a channel transmitted by the horizontal polarization array and formed by the vertical polarization array; a vertical transmitting and horizontal receiving channel VH is a channel which is transmitted by a vertical polarization array and is formed by receiving signals of the vertical polarization array; the vertical transmitting and vertical receiving channel VV is a channel which is transmitted by a vertical polarization array and is formed by a receiving vertical polarization array.

Step 2: adding the radar echo signal received each time as a radar echo signal of one period into a preset sample set;

it can be understood that the preset sample set is an empty set, and after the radar echo signals of multiple cycles are added, the sample set includes multiple samples, and each sample is a radar echo signal of one cycle.

And step 3: respectively calculating the polarization invariant and the polarization decomposition quantity of the peak value of each sample in a preset sample set;

as an optional embodiment of the present invention, step 3 includes:

step 31: for each sample in the set of samples, determining a peak value for the sample;

wherein the peak value is an echo signal returned by the position of the corner reflector or an echo signal returned by the position of the target;

step 32: calculating determinant mode values at each sample peak;

wherein the determinant mode value is expressed as:

|Δ|＝|det(S)|＝|s_HH+s_VV-s_HVs_VH|

considering that the cross polarization of a naturally scattering target generally satisfies reciprocity, the above equation is simplified as:

step 33: calculating a power matrix trace at each sample peak;

the power matrix trace is represented as:

step 34: calculating an intrinsic polarization direction angle at each sample peak;

the intrinsic polarization orientation angle is expressed as:

step 35: for each sample, the determinant modulus value, the power matrix trace and the intrinsic polarization direction angle of the sample form a polarization invariant of the sample;

wherein the content of the first and second substances,S_HHrepresenting the signal component of the horizontal transmit-receive channel, S_VVRepresenting the signal component of the vertical transmit-vertical receive path, S_HVRepresenting the signal component of the horizontal transmit-vertical receive path, S_VHSignal components representing the vertical transmit horizontal receive path are labeled with a conjugate.

Step 36: the polarization split amount of each sample is calculated.

As an alternative embodiment of the present invention, step 36 includes:

step a: calculating a coherence matrix at each sample peak;

step b: for each sample, solving the Shannon entropy at the peak value by using the coherent matrix at the peak value of the sample;

step c: the shannon entropy at each sample peak is decomposed into intensity entropies.

The coherence matrix is represented as:

shannon entropy is expressed as:

decomposing the shannon entropy into the intensity entropy to directly give an intensity entropy result:

the intensity entropy is expressed as:

wherein S is_HHRepresenting the signal component of the horizontal transmit-receive channel, S_VVRepresenting the signal component of the vertical transmit-vertical receive path, S_HVRepresenting the signal component of the horizontal transmit-vertical receive path, S_VHRepresenting the signal components of the vertical transmit-horizontal receive path, the superscript H representing the conjugate transpose,λ_iis the i-th eigenvalue, λ, of the coherence matrix T_kDenotes the same as_iThe kth eigenvalue of the coherence matrix T, i and k are only for discrimination, tr (T) represents the trace of the coherence matrix T.

And 4, step 4: forming a characteristic vector by the polarization invariant and the polarization decomposition quantity of each sample;

in the step, determinant mode values, power matrix traces, eigen-polarization direction angles and intensity entropies can be combined into characteristic vectors, so that the characteristic vectors T of the target ship and the corner reflector are constructed as [ delta, P, alpha, SE ]_I]。

And 5: inputting the feature vector of each sample into a preset classifier, and training the classifier to obtain the trained classifier;

step 6: and classifying the returned echo signals by using a trained classifier so as to distinguish the corner reflector from the target.

Wherein the classifier outputs a classification label for distinguishing the target from the corner reflector.

Compared with the existing polarization corner reflector distinguishing method, the polarization corner reflector distinguishing method provided by the embodiment effectively solves the problem that the simulation condition is not matched with the actual measurement environment, integrates the polarization differences generated by different environments and physical states of the corner reflector and the target ship, and is caused by the deployment measures of the corner reflector and the inherent properties of the corner reflector, so that the method is not easily influenced by the actual measurement environment and radar measurement errors, and has wider applicability and good recognition rate.

The sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition provided by the invention comprises the steps of considering the physical properties of a target ship and a corner reflector and the difference of the actual sea surface environment, utilizing the difference of determinant mode values and power matrix traces of the target ship and the corner reflector in the aspect of energy, utilizing an intrinsic polarization direction angle in the aspect of influence of the actual sea surface condition, utilizing intensity entropy in the aspect of energy randomness, and finally constructing a feature vector training classifier so as to improve the classification accuracy of the classifier; the target vessel and the corner reflector are then distinguished using a classifier. Therefore, the method can effectively solve the problem that the design of a simulation model in the existing polarization identification method is not matched with the actual situation, so that the polarization angle identification corner reflector has higher precision and feasibility.

As an optional implementation manner of the present invention, the present invention may further retrain the trained classifier, and the training process is as follows:

step a: removing items formed by target interference caused by the corner reflector in the feature vector to obtain the feature vector after removal;

step b: and retraining the classifier by using the feature vectors after the elimination.

It can be understood that the feature vectors are removed, so that the interference to the target caused by the side lobe of the corner reflector can be avoided, and the classification accuracy of the trained classifier is further improved.

In order to verify the effectiveness and the practicability of the sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition provided by the present application, the present embodiment is further explained by a certain horizontal-vertical polarization-based radar measured data processing experiment:

please refer to fig. 2, fig. 2 is a time-frequency schematic diagram of a single period of measured data in the sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition according to an embodiment of the present invention, it can be seen that peaks exist at 10.12km, 10.23km, and 10.34km, where the first two peaks are corner reflectors and the third peak is a target ship, and it can be observed that the amplitude of an echo signal of a corner reflector is higher than that of a target by more than 15 dB. The sea surface corner reflector disturbance discrimination method based on polarization invariant and polarization decomposition provided herein was adopted for these three peaks, and the results are shown in fig. 3, fig. 4, fig. 5, fig. 6, and table 1.

Fig. 3 is a determinant module value comparison diagram of a corner reflector and a target ship in measured data in a sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition according to an embodiment of the present invention; fig. 4 is a power matrix trace comparison diagram of a corner reflector and a target ship in measured data in the sea surface corner reflector interference identification method based on polarization invariants and polarization decomposition according to the embodiment of the present invention, and it can be seen that echo energy of the corner reflector is significantly greater than target echo; fig. 5 is a comparison graph of an eigen polarization direction angle between a corner reflector and a target ship in measured data in the sea surface corner reflector interference identification method based on polarization invariants and polarization decomposition according to an embodiment of the present invention, which shows that the eigen polarization direction angle of the target is more stable, because a boat carrying the corner reflector is more susceptible to wave fluctuation, which causes instability of the eigen polarization direction angle; fig. 6 is a comparison graph of intensity entropy between a corner reflector and a target ship in measured data in the sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition provided by the embodiment of the present invention, where the intensity entropy reflects the random degree of total polarization energy of the target, and it can be seen that because the corner reflector is composed of multiple scattering points, the distribution of polarization energy is more dispersed, and thus the intensity entropy is higher than that of the target ship.

Referring to table 1, table 1 describes the results of identifying the corner reflector interference and the target in the sea surface corner reflector interference identification method based on polarization invariant and polarization decomposition provided by the embodiment of the present invention, and it can be seen that the method provided by the present invention has high identification performance and can effectively distinguish the corner reflector and the target.

TABLE 1 comparison of identification results

The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.