A multi-radar joint positioning method based on azimuth angle and doppler velocity

Abstract

The application discloses a multi-radar joint positioning method based on azimuth angle and Doppler velocity, and mainly solves the problems of poor positioning precision caused by large sensor measurement error and insufficient utilization of multi-source data. The method is aimed at a sea surface target moving at a constant speed, and uses target direction finding results of 1-2 fixed passive radars and target Doppler velocity measurement values of 1 fixed ground wave active radar. According to different numbers of radars, different filter initial values and covariance matrices are constructed, corresponding measurement equations and different filter starting times are determined, and filtering is performed on the target state according to the constant speed motion model. The filtering output is used as the positioning result of the target. The method uses the measurement of passive radars and ground wave radars of two different systems, reduces the influence of large single measurement error on the positioning result, shortens the positioning time, and improves the positioning precision.

Description

Technical Field

[0001] This invention belongs to the field of radar data processing technology, and specifically relates to a multi-source sensor joint positioning method. Background Technology

[0002] Existing passive radar has a long detection range and can intercept radiation source signals from over-the-horizon targets at sea, obtaining the target's azimuth through direction finding. However, direction finding errors are relatively large, resulting in poor accuracy for passive radar-based cooperative localization, which further deteriorates with increasing target distance. Ground-wave radar, a sensor with a different system than passive radar, also possesses over-the-horizon target detection capabilities, but its range and angular resolution are poor, while its Doppler velocity measurement is better. With the evolving needs of military strike and defense, situational awareness capabilities based on a single type of sensor are insufficient for diverse mission scenarios, unable to provide multi-dimensional situational data, and unable to provide higher-quality target data. Therefore, research is needed on technologies based on the combined application of multi-source sensors, particularly concerning target localization. Summary of the Invention

[0003] To address the issues of poor positioning accuracy and insufficient utilization of multi-source data caused by large measurement errors in existing sensors, this invention proposes a multi-radar joint positioning method based on azimuth and Doppler velocity. For a target moving at a constant speed on the sea surface, the method utilizes the target azimuth measurement of 1 to 2 fixed passive radars and the target Doppler velocity measurement of 1 fixed ground wave radar. Depending on the number of radars, different initial filter values ​​and covariance matrices are constructed, the corresponding measurement equations and different filter start times are determined, and the UKF is used to filter the target state. The filtered output is then used as the target positioning result.

[0004] The technical solution of this invention includes:

[0005] Step 1: For a target moving at a constant speed on the sea surface, m fixed passive radars intercept the target's radiation source with a period of Δ. T The signal was used to determine the target's azimuth. Where m is 1 or 2, there is another fixed ground wave active radar that operates according to the radiation source signal period Δ T Measure the target Doppler velocity The errors all follow a Gaussian distribution, and the standard deviations are respectively

[0006] Step 2: Calculate the initial value of the target state and its covariance P0=diag([(3*v x，0 *Δ T ) 2 , (3*v x，0 ) 2 , (3*v y，0 *Δ T )2 , (3*v y，0 ) 2 ]), where diag(a) represents a matrix with vector a as its diagonal elements: in the scenario where m=2, in It is used at time 1 The position estimate obtained by triangulation. It is used at time 2. The obtained position estimate; in the scenario where m=1, Where α is the passive radar's utilization of the first N measurements The estimated direction of target motion at time t=N, r N And |v| are the values ​​obtained by ground-wave radar from the first N measurements. The estimated target distance r at time t=N N And the magnitude of the velocity |v|, N≥3;

[0007] Step 3: with For measurement, the target state vector x is calculated according to the uniform motion model. t =[x t ;v x，t ;y t ;v y，t ] T Perform filtering, where x t It is the x-axis position, v x t is the x-axis velocity component, y t It is the y-axis position, v y，t This refers to the y-axis velocity component, with the filtering starting at time T, i.e., t = T, T+1, ...: When m = 2, the filtering starting time T ≥ 3, and the measurement vector is... Nonlinear terms in the measurement equation:

[0008]

[0009] Where (x) 1 y 1 ) and (x 2 y 2 (x) represent the positions of the two passive radars. 3 y 3 () indicates the location of the ground wave radar and measures the noise. Let be a Gaussian noise vector with a mean of 0 and a variance matrix of . When m = 1, the filter start time T ≥ N + 1, and the measurement vector is... Nonlinear terms in the measurement equation:

[0010]

[0011] Measurement noise It is a Gaussian vector with a mean of 0 and a variance matrix of .

[0012] Step 4: Filter the results As the final output.

[0013] This invention utilizes measurements from two different sensor systems: passive radar and ground wave radar. By leveraging multi-source data, the positioning accuracy is improved compared to a single sensor system. Through joint data processing, the impact of large single measurement errors on positioning accuracy is reduced. Calculating optimal initial filter values ​​shortens the time required to obtain effective positioning results. Attached Figure Description

[0014] Figure 1 This is the implementation process of the example.

[0015] Figure 2 This is the distance error curve of the positioning result in the embodiment.

[0016] Figure 3 This is the velocity error curve of the positioning result in the example. Detailed Implementation

[0017] The method of the present invention will be described in further detail below with reference to the accompanying drawings and preferred embodiments.

[0018] Reference Figure 1 The processing flow in the present invention, in a preferred embodiment, is as follows:

[0019] Step 1: For a target moving at a constant speed on the sea surface, m fixed passive radars intercept the target's radiation source with a period of Δ. T The signal was used to determine the target's azimuth. Where m is 1 or 2, there is another fixed ground wave active radar that operates according to the radiation source signal period Δ T Measure the target Doppler velocity The errors all follow a Gaussian distribution, and the standard deviations are respectively

[0020] Step 2: Calculate the initial value of the target state and its covariance P0=diag([(3*v x，0 *Δ T ) 2 , (3*v x，0 ) 2 , (3*v y，0 *Δ T ) 2 , (3*v y，0 ) 2]), where diag(a) represents a matrix with vector a as its diagonal elements; when m=2, the initial value of the target state and its covariance are calculated using the triangulation results, i.e., in It is used at time 1 The position estimate obtained by triangulation. It is used at time 2. The obtained location estimate, i.e.

[0021]

[0022] i = 1, 2, where (x 1 y 1 ) and (x 2 y 2 These represent the positions of the two passive radars; when m=1, by estimating the target's direction of motion, target distance, and target velocity, the initial value of the target state and its covariance are calculated, i.e., Where α is the passive radar's utilization of the first N measurements The estimated direction of target motion at time t=N, r N And |v| are the values ​​obtained by ground-wave radar from the first N measurements. The estimated target distance r at time t=N N And the magnitude of the velocity |v|, N≥3;

[0023] Step 3: Based on measurement data Use UKF to analyze the target state vector X t Perform filtering, X t =[x t ;v x，t ;y t ;v y，t ] T x t It is the x-axis position, v x，t It is the x-axis velocity component, y t It is the y-axis position, v y，t It is the y-axis velocity component, and the filtering starts at time T, i.e., t = T, T+1, ...; the state transition matrix F is...

[0024]

[0025] When m = 2, the filter start time T ≥ 3, and the measurement vector is... The measurement matrix is

[0026]

[0027] Where (x) 1 y 1 ) and (x2 y 2 (x) represent the positions of the two passive radars. 3 y 3 () indicates the location of the ground wave radar and measures the noise. Let be a Gaussian noise vector with a mean of 0 and a variance matrix of . When m = 1, the filter start time T ≥ N + 1, and the measurement vector is... The measurement matrix is

[0028]

[0029] Measurement noise It is a Gaussian vector with a mean of 0 and a variance matrix of .

[0030] Step 4: Filter the results As the final output.

[0031] The beneficial effects of the method of the present invention can be further illustrated by the following simulation.

[0032] 1. Simulation conditions:

[0033] Condition 1: There are 2 fixed passive radars and 1 fixed ground wave radar. The position of passive radar 1 is [-15000, 25000], the position of passive radar 2 is [15000, -25000], and the position of the ground wave radar is [30000, -50000], in meters. The standard deviation of the direction finding error of the passive radar is 1°, and the standard deviation of the Doppler velocity measurement error of the ground wave radar is 10 m / s. The sampling interval is 10 s. The initial position of the target is [140000, 80000], the heading is -150° (0° is the positive X-axis direction, increasing clockwise), and the speed is 35 knots. UKF parameters: dimension 4, α = 0.01, β = 2, λ = -3.9996.

[0034] 2. Simulation content:

[0035] Under condition 1, the distance positioning error and velocity error of the target were simulated using the method of this invention. 200 measurements were taken, and 1000 Monte Carlo simulations were performed at each measurement time. The distance error results are as follows: Figure 2 As shown, the speed error is as follows Figure 3 As shown.

[0036] 3. Simulation Analysis:

[0037] from Figure 2It can be seen that the distance positioning error decreases with the increase of the number of measurements. The method of this invention is superior to the single passive radar direction finding and filtering method. After 20 measurements, the distance positioning error is reduced to 3km to 4km. Compared with the single passive radar direction finding and filtering method, the accuracy is improved by 1km to 2km. When the positioning error is 5km, the time is shortened by 10 to 15 measurements, totaling 100 to 150 seconds.

[0038] from Figure 3 It can be seen that the velocity estimation error decreases with the increase of the number of measurements. The method of this invention is superior to the single passive radar direction finding and filtering method. Taking the X component as an example, after 50 measurements, the velocity error is reduced to about 2 m / s. Compared with the single passive radar direction finding and filtering method, the accuracy is improved by about 9 m / s. When the velocity error reaches 5 km, the time is shortened by about 60 measurements, totaling 600 seconds.

Claims

1. A multi-radar joint positioning method based on azimuth angle and Doppler velocity, characterized in that: Step 1: For a target moving at a constant speed on the sea surface, m fixed passive radars intercept the target's radiation source with a period of... The signal was used to determine the target's azimuth. ,in m The value is 1 or 2, and there is another fixed ground wave active radar that operates according to the period of the radiation source signal. Measure the target Doppler velocity The errors all follow a Gaussian distribution, and the standard deviations are respectively ; Step 2: Calculate the initial value of the target state and its covariance Where diag(a) represents a matrix with vector a as its diagonal. In the scenario where m=2, ,in It is used at time 1 The position estimate obtained by triangulation. It is used at time 2. The obtained position estimate; in the scenario where m=1, ,in Passive radar utilizes the first N measurements The estimated direction of target motion at time t=N and Ground wave radar utilizes the first N measurements Estimated target distance at time t=N and speed magnitude , ; Step 3: with Measurement, based on the uniform motion model, of the target state vector Perform filtering, where It is the position on the x-axis. It is the x-axis velocity component. It is the position on the y-axis. It is the y-axis velocity component, and the filtering starts at time T, i.e. When m=2, the filter start time The measurement vector is Nonlinear terms in the measurement equation: ； in and These are the locations of the two passive radars. This refers to the location of the ground wave radar and the measurement of noise. Let be a Gaussian noise vector with a mean of 0 and a variance matrix of . ; When m=1, the filter start time Measurement vector Nonlinear terms in the measurement equation: ； Measurement noise Gaussian vectors with a mean of 0 and a variance matrix of . ; Step 4: Filter the results As the final output.

Images