TAS enhanced beam tracking method based on maneuver cognition for phased array radar

Abstract

The present invention proposes a TAS enhanced beam tracking method for phased array radar based on maneuver recognition, which belongs to the technical field of rotating phased array radar maneuvering target tracking. Rotating phased array radars usually use TAS beams to track targets. The beam width is usually narrow. For highly maneuverable targets, it is easy to lose tracking when the target maneuvers. Usually, for short-range targets, multiple beams are used to form a large wave gate to track the target, but this will cause a waste of time resources. The present invention recognizes the target maneuvering state, including maneuvering detection and maneuvering direction prediction. After detecting the target maneuver, the maneuvering direction of the target is predicted based on the tracking residual, and an enhanced beam is added in the maneuvering direction. Compared with traditional algorithms, tracking time resources are saved by more than 50%.

Description

Technical Field

[0001] The invention belongs to the technical field of radar data processing. Background Art

[0002] Phased array radars use TAS (track and search) mode to track targets, offering two major advantages over TWS (track while search) mode used by mechanically scanned radars: a) adjustable tracking parameters and data rates; and b) high accuracy through sum-and-difference angle measurement. Highly maneuverable incoming targets can suddenly maneuver out of the phased array radar's TAS tracking beam, causing loss of tracking. Therefore, real-time detection of target maneuvers and increasing the tracking data rate upon detection of the start of a maneuver are crucial for maintaining stable target tracking.

[0003] The rotating phased array radar uses one array to replace four arrays, covering 360 degrees of airspace in azimuth through mechanical high-speed rotation. Under the same power consumption and cost, its performance exceeds that of the fixed array phased array radar.

[0004] Increasing the data rate is an effective method for tracking maneuvering targets. However, the data rate of rotating phased array radars cannot be "infinitely variable." The tracking period can only be an integer multiple of the antenna's rotational speed. For highly maneuverable targets, when the tracking period is equal to the antenna's rotational speed period, the only way to maintain tracking is by expanding the tracking gate. When a phased array radar tracks a target using sampling and differential angle measurement, its tracking gate is the beam width. If the target is determined to be maneuvering, multiple beams, typically four or nine beams, are used to form a larger tracking gate. Expanding the gate consumes four or nine times the time resources. Therefore, expanding the gate if the target is not maneuvering results in a waste of time resources. However, if the target is maneuvering and the gate is not expanded, tracking the target will be lost.

[0005] Therefore, it is crucial to determine target maneuvers in real time and, based on these determinations, increase the tracking data rate or expand the threshold. Traditional maneuvering methods are primarily designed for multi-model tracking. Once a target maneuver is determined, the tracking model is changed. Target maneuver detection algorithms can be categorized by the measurement quantity used, including chi-square tests and likelihood ratio tests. Hypothesis testing algorithms for radar maneuver detection draw on techniques from other engineering applications. Radar target tracking has its own unique characteristics. Hypothesis testing quantities primarily utilize sample amplitude information. In radar tracking, the measurement quantity is the target residual. The target parameter is the target's measured position minus the predicted position. The residual contains both amplitude and sign information. The aforementioned algorithms only use the amplitude of the residual to detect target maneuvers. When the target maneuvers dramatically, the amplitude information can pass the threshold. However, when the target maneuvers are small, using amplitude information can lead to misjudgment. Summary of the Invention

[0006] Rotating phased array radars typically track targets using TAS beams. These beams are typically narrow, making it easy for them to lose tracking when maneuvering. For short-range targets, multiple beams are typically used to form a large gate for tracking, but this results in a significant waste of time and resources. This paper proposes a TAS-enhanced beam tracking method for phased array radars based on maneuver awareness. First, the target maneuvers are detected. After the target maneuvers are detected, the target's maneuvering direction is predicted based on the vectors of the residuals in the x, y, and z directions. An auxiliary beam is then added in this direction. Compared to traditional algorithms, this method saves over 50% of tracking time and resources.

[0007] The TAS enhanced beam tracking method for phased array radar based on maneuverability cognition proposed in the present invention includes the following steps:

[0008] Step 1. Real-time detection of whether the target motion state changes

[0009] Step 11. Model the target motion process using a uniform velocity model:

[0010] X(k+1)=F(k)X(k)+Γ(k)v(k)

[0011] Where X(k) is the state vector, F(k) is the state transfer matrix, Γ(k) is the system process noise input matrix, v(k) is the system process noise, which obeys the zero-mean Gaussian distribution and its covariance matrix is Q(k);

[0012] Step 12. Model the measurement process:

[0013] Z(k)=H(k)X(k)+W(k)

[0014] Where H(k) is the measurement matrix, W(k) is the measurement noise, which is assumed to be zero-mean Gaussian white noise, and its covariance matrix is R(k), and Z(k) is the measurement state of the target at time k (x k ,y k , z k ),in:

[0015]

[0016] Where θ is the pitch angle of the target at time k in the radar polar coordinate system, The azimuth of the target at time k in the radar polar coordinate system, r is the distance between the target and the radar at time k in the radar polar coordinate system;

[0017] Step 13. Use the Kalman filter to calculate the tracking innovation γ(k+1) and the covariance S(k+1) of the innovation:

[0018]

[0019] S(k+1)=H(k+1)P(k+1|k)H H (k+1)+R(k+1);

[0020] In the formula P(k+1|k)=F(k)P(k|k)F H (k)+Γ(k)Q(k)Γ H (k), where P(k|k) is the covariance matrix of the target state and P(k+1|k) is the prediction covariance matrix.

[0021] Step 14. Determine whether the target is maneuvering:

[0022] If the following equation is satisfied:

[0023]

[0024] It is considered that the target motion has changed and is no longer uniform linear motion, where γ i is the i-th component of the new information γ(k+1), K is the maneuver detection coefficient, and K=3.5 is used in the application;

[0025] Step 2. Predict the target's maneuvering direction:

[0026] Convert the residual γ(k+1) calculated in step 1 to the polar coordinate system:

[0027]

[0028] Then the residual vector is The target maneuver direction is

[0029] Step 3: Determining the beam position of the enhanced beam includes:

[0030] Step 31: Calculate the position of the main beam scan

[0031] Use the following formula to predict the state of the target at the next moment:

[0032]

[0033] In the formula is the filtered value of the current target state, F(k) is the state transfer matrix, Convert to polar coordinate system,

[0034]

[0035] The beam position for the next scan is obtained as Where (x, y, z) is The three coordinate components of .

[0036] Step 32: Calculate the position of the enhanced beam position

[0037] Assuming that the radar beam width is 2α, in the maneuvering direction predicted in S2, a beam identical to the main beam is used to capture the maneuvering target. The position of the enhanced beam is calculated as:

[0038]

[0039] Step 4. Correlate the points within the main beam and the enhancement beam:

[0040] When tracking a target, the radar uses the main beam and enhanced beam positions calculated in step 3 to fire two TAS beams in sequence. Each additional beam fired means twice as much time is consumed. Although a wide gate effect can be achieved by expanding a single beam, the measurement accuracy of the wide gate generated by the single beam expansion will decrease, that is, the measurement accuracy is inversely proportional to the beam width. First, a point is selected in the main beam and associated with the target track. If there is no point in the main beam, a point is selected in the enhanced beam. If there is no point in the enhanced beam, the loss-of-track capture task is initiated, that is, multiple beams are fired around the predicted point to search for the target in a larger range. If association is successful, steps 1 to 4 are repeated.

[0041] The present invention calculates the target's maneuvering direction based on the vectors of the residuals in the three directions of x, y, and z. First, the target maneuvers are detected in real time. After the target maneuvers are detected, an auxiliary enhancement beam is added in the maneuvering direction. Compared with traditional algorithms, tracking time resources are saved by more than 50%. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Attachment Figure 1 It is a flow chart of the TAS enhanced beam tracking method based on maneuver cognition of phased array radar. DETAILED DESCRIPTION

[0043] The present invention will be further explained below with reference to the accompanying drawings and embodiments.

[0044] The embodiments of the present invention include:

[0045] Step 1. Real-time detection of whether the target motion state changes

[0046] Using the classic Kalman filter model, the target is tracked and the motion and measurement processes are modeled:

[0047] Step 11. Model the target motion process using a uniform motion model:

[0048] X(k+1)=F(k)X(k)+Γ(k)v(k)

[0049] Where X(k) is the state vector, F(k) is the state transfer matrix, Γ(k) is the system process noise input matrix, and v(k) is the system process noise, which obeys a zero-mean Gaussian distribution and has a covariance matrix of Q(k). The motion process state vector, state transfer matrix, and noise are initialized using measurement data.

[0050] Step 12. Model the measurement process:

[0051] Z(k)=H(k)X(k)+W(k)

[0052] Where H(k) is the measurement matrix, W(k) is the measurement noise, which is assumed to be zero-mean Gaussian white noise, and its covariance matrix is R(k), and Z(k) is the measurement state of the target at time k (x k ,y k , z k ), since the radar measures the target in the polar coordinate system, and the Kalman filter tracks the target in the rectangular coordinate system, a conversion is required, where:

[0053]

[0054] Where θ is the pitch angle of the target at time k in the radar polar coordinate system, The azimuth of the target at time k in the radar polar coordinate system, r is the distance between the target and the radar at time k in the radar polar coordinate system;

[0055] Step 13. Use the Kalman filter to calculate the tracking innovation γ(k+1) and the covariance S(k+1) of the innovation:

[0056]

[0057] S(k+1)=H(k+1)P(k+1|k)H H (k+1)+R(k+1);

[0058] In the formula P(k+1|k)=F(k)P(k|k)F H (k)+Γ(k)Q(k)Γ H (k), where P(k|k) is the covariance matrix of the target state, and P(k+1|k) is the prediction covariance matrix. The covariance of the new information includes the process variance and the measurement variance. If no maneuver occurs, the covariance of the information will become larger, which can be used to determine whether the target maneuvers occur.

[0059] Step 14. Determine whether the target is maneuvering:

[0060] The basis for judgment is the confidence level of the normal distribution or chi-square distribution. Many literatures have proved that the square of the new information divided by the information covariance obeys the chi-square distribution. The judgment is made by setting the confidence interval. If the ratio calculated by the current filter does not satisfy the chi-square distribution, it is considered that a maneuver has occurred. If the following formula is satisfied:

[0061]

[0062] It is considered that the target motion has changed and is no longer uniform linear motion, where γ i is the i-th component of the new information γ(k+1), K is the maneuver detection coefficient, and in practical applications, K=3.5;

[0063] Step 2. Predict the target's maneuvering direction:

[0064] Convert the residual γ(k+1) calculated in step 1 to the polar coordinate system:

[0065]

[0066] Then the residual vector is The target maneuver direction is

[0067] Step 3. Determine the beam position of the enhanced beam

[0068] Step 31: Calculate the position of the main beam scan

[0069] Use the following formula to predict the state of the target at the next moment:

[0070]

[0071] In the formula is the filtered value of the current target state, F(k) is the state transfer matrix, Convert to polar coordinate system,

[0072]

[0073] The beam position for the next scan is obtained as Where (x, y, z) is The three coordinate components of .

[0074] Step 32: Calculate the position of the enhanced beam position

[0075] Assuming that the radar beam width is 2α, in the maneuvering direction predicted in step 2, a beam identical to the main beam is used to capture the maneuvering target. The position of the enhanced beam is calculated as:

[0076]

[0077] Step 4. Correlate the points within the main beam and the enhancement beam:

[0078] When tracking a target, the radar uses the main beam and enhanced beam positions calculated in step 3 to sequentially shoot two TAS beams. First, a point is selected in the main beam and associated with the target track. If there is no point in the main beam, a point is selected in the enhanced beam. If there is no point in the enhanced beam, the lost tracking capture task is initiated, that is, multiple beams are shot around the predicted point to search for the target in a larger range. If the association is successful, steps 1-4 are repeated.

Claims

1. A TAS-enhanced beam tracking method for phased array radar based on maneuverability recognition is characterized by: Step 1. Real-time detection of whether the target motion state changes: Step 11. Model the target motion process using a uniform motion model: X(k+1)=F(k)X(k)+Γ(k)v(k) Where X(k) is the state vector, F(k) is the state transfer matrix, Γ(k) is the system process noise input matrix, v(k) is the system process noise, which obeys the zero-mean Gaussian distribution and its covariance matrix is Q(k); Step 12. Model the measurement process: Z(k)=H(k)X(k)+W(k) Where H(k) is the measurement matrix, W(k) is the measurement noise, which is assumed to be zero-mean Gaussian white noise, and its covariance matrix is R(k), and Z(k) is the measurement state of the target at time k (x k ,y k , z k ),in: Where θ is the pitch angle of the target at time k in the radar polar coordinate system, The azimuth of the target at time k in the radar polar coordinate system, r is the distance between the target and the radar at time k in the radar polar coordinate system; Step 13. Use the Kalman filter to calculate the tracking innovation γ(k+1) and the covariance S(k+1) of the innovation: S(k+1)=H(k+1)P(k+1|k)H H (k+1)+R(k+1); In the formula P(k+1|k)=F(k)P(k|k)F H (k)+Γ(k)Q(k)Γ H (k), where P(k|k) is the covariance matrix of the target state and P(k+1|k) is the prediction covariance matrix; Step 14. Determine whether the target is maneuvering: If the following equation is satisfied: It is considered that the target motion has changed and is no longer uniform linear motion, where γ i is the i-th component of the new information γ(k+1), K is the maneuver detection coefficient, K=3.5; Step 2: Predict the target's maneuvering direction: Convert the residual γ(k+1) calculated in step 1 to the polar coordinate system: Then the residual vector is The target maneuver direction is Step 3: Determine the beam position of the enhanced beam: Step 31: Calculate the main beam scanning position and use the following formula to predict the target state at the next moment: In the formula is the filtered value of the current target state, F(k) is the state transfer matrix, Convert to polar coordinate system; The beam position for the next scan is obtained as Where (x, y, z) is The three coordinate components of ; Step 32: Calculate the position of the enhanced beam: Assuming the radar beam width is 2α, then in the maneuvering direction predicted in step 2, a beam identical to the main beam is used to capture the maneuvering target. The position of the enhanced beam is calculated as: Step 4: Associate the points within the main beam and the enhanced beam: When tracking a target, the radar uses the main beam and enhanced beam positions calculated in step 3 to sequentially shoot two TAS beams. First, a point is selected in the main beam and associated with the target track. If there is no point in the main beam, a point is selected in the enhanced beam. If there is no point in the enhanced beam, the lost tracking capture task is initiated, that is, multiple beams are shot around the predicted point to search for the target in a larger range. If association is successful, steps 1 to 4 are repeated.

Images