CN105652257A - Method for detecting ship target through double-station single-array-element high-frequency ground wave radar - Google Patents

Abstract

A method for detecting a ship target through double-station single-array-element high-frequency ground wave radar comprises the steps that R-D data of a single channel of the high-frequency ground wave radar within a certain time quantum is utilized for constructing R-D-T three-dimensional data, and a dynamic planning method is utilized for obtaining an undirected flight path detection result of two stations based on the constructed R-D-T three-dimensional data; a flight path matched and correlated with a first flight path of the first station is found in the second station, then flight paths, correlated with left flight paths of the first station, of the second station are found, and finally correlated flight paths within the covering range of the two stations are obtained. Only one receiving array element is needed in each ground wave radar station, and a detecting and tracking integrated method can be utilized for obtaining the undirected flight path of the target and the matching and correlating result of the two stations. The needs for the size of a high-frequency ground wave radar receiving station can be greatly reduced through a ground wave radar system of the single-array-element receiving stations, the defect that traditional array type ground wave radar needs a large-size radar field is overcome, and application of the target detecting ground wave radar is remarkably improved.

Description

A kind of method utilizing dual station list array element high-frequency ground wave radar detection ship target

Technical field

The present invention relates to a kind of dual station list array element high-frequency ground wave radar ship object detection method, it is specifically related to a kind of method utilizing dual station list array element high-frequency ground wave radar detection ship target based on detecting and tracking integration.

Background technology

High-frequency ground wave radar works in high frequency (3��30MHz) frequency range, utilizes frequency electromagnetic waves coastal surface diffraction propagation characteristic, it is possible to detecting the marine moving target beyond sighting distance, maximum probe distance can reach 400km. At present, being all large-scale array type system for marine target detection ground wave radar, radar scanner battle array several hundred meters easily, even goes up km (Heronetal., 2012; Savidgeetal., 2011), it is necessary to the radar place, seashore of big area, and place, region, shoreline is rare, because which limit the widespread use of array high-frequency ground wave radar.

In recent years, high-frequency ground wave radar for target detection developed towards system compact direction. The CODAR ground wave radar net of Roger Si university of U.S. application and construction does ship target detect (Dobsonetal., 2013), domestic utilizes dual station ground wave radar to carry out ship target detection (Wen Bin, 2011). Above-mentioned two researchs adopt first list to stand target side to obtaining a mark, then does Track In Track, is finally that the mode of dual station track association carries out. In this kind of dual station object detection method, each single station can provide the true flight path with real position, utilizes the major objective of dual station to be improve the detection accuracy of overall flight path on basis, single station. But its method is owing to being the thinking adopting Detect before Track, and the poor orientation of monostatic radar can affect Track In Track effect to detection accuracy, finally affects the overall detection accuracy of dual station. Pay the proud son of heaven (2011) and Zhou Junyu (2013) to have studied and only utilize the dual station ground wave radar multi-object tracking method apart from Doppler message, first utilize and only have the object point of Distance geometry speed to form possible flight path, and then reject false track ghost flight path. The above-mentioned non-angular dual station object detection method researched and proposed, no matter utilizing Dan Zhanzai to merge still directly utilizes dual station point mark information to form flight path, it in essence or the thinking of Detect before Track only utilizes the spatial information of target and does not have the attribute information such as echo strength of target in detection.

As can be seen from distance-doppler (R-D) two-dimensional spectrum of actual high-frequency ground wave radar, ship target shows as appearance mark in R-D composes, this appearance mark represents moving target variation characteristic on Distance geometry speed two dimension degree within certain accumulation time, target more more than simple peak point motion and attribute information can be provided, can better distinguish than a mark and identify target. Therefore, the basis of R-D two dimension increases three-dimensional structure that time dimension forms R-D-T again, just target detect can be expanded in time peacekeeping attribute dimension, make full use of target face region and tie up on locus and echo strength attribute all continuous distribution and the characteristic of clutter noise stochastic distribution in time, realize target flight path and the integrative detection of some mark, obtain the undirected flight path result of monostatic radar. And then the undirected flight path result of detection utilizing two single stations associates, forming possible flight path result, the characteristic utilizing target to do translational motion within the short period of time rejects false track, finally obtains real multi-target traces result. Relevant reference is as follows:

[1] literary composition is refined. and (2011). bistatic high frequency radar flight path fusion method research (Master's thesis, Harbin Institute of Technology).

[2] proud son of heaven is paid, Zhou Gongjian, Tian Wenlong, etc. non-angular dual station ground wave radar combination locating and tracking and filtering algorithm [J]. systems engineering and electronic technology, 2011,33 (3): 552-556.

[3] DobsonC, HolensteinK, SmithM, etal.MonostaticvesseldetectionstatisticsfromtheCODARSeaS onde [C] .Oceans-SanDiego, 2013IEEE, 2013:1-4. (single base CODARSeaSonde enlightening ripple radar ship detection statistics)

Summary of the invention

It is an object of the invention to provide a kind of method utilizing dual station list array element high-frequency ground wave radar detection ship target, the method can make full use of in ground wave radar R-D ship target face region and tie up on locus and echo strength attribute all continuous distribution and the characteristic of clutter noise stochastic distribution in time, realize target flight path and the integrative detection putting mark, obtain the undirected flight path result of monostatic radar, then the characteristic utilizing target to do translational motion within the short period of time realizes the accurate correlation of the undirected flight path of dual station, it is achieved multi-target traces detects.

In order to realize above-mentioned purpose, the technical scheme of the present invention is as follows:

A kind of method utilizing dual station list array element high-frequency ground wave radar detection ship target, comprise and first utilize the high-frequency ground wave radar of two websites to the ship objective emission pulse signal on sea, echoed signal is obtained after postponing through the regular hour, this echoed signal is utilized to obtain distance-doppler's data (R-D data), it is characterised in that also to comprise the following steps:

Step 1: high-frequency ground wave radar single pass R-D data construct R-D-T tri-dimension data utilizing certain time section T (T is less than one hour), is designated as Y (r, d, t), wherein r=1 ..., M, M is the range unit lattice number in R-D data, d=1 ... N, N are the doppler's unit lattice number in R-D data, t=1,, A, A are natural number, time period T being divided into A part, thus obtains A moment, t represents one of them moment;

Step 2: based on the R-D-T tri-dimension data Y (r, d, t) built, utilize dynamic programming method, obtain the undirected flight path result of detection of two websites, be respectively T₁(R_i,V_i, t) and T₂(R_j,V_j, t); Wherein, i=1,2...m, m are the flight path quantity that obtains of station 1, and j=1,2...n, n are the flight path quantity that station 2 obtains, and the flight path quantity that two websites obtain is the same, i.e. m=n; T=1 ..., A; Ri represents the distance of t target ship and website 12;

Step 3: find the Article 1 flight path T with website 1 in station 2₁(R₁,V₁, t) mate the flight path of association mutually:

For each flight path T of website 2₂(R_j,V_j, t), j=1,2...n, successively with the Article 1 flight path T in website 1₁(R₁,V₁, t) carry out following calculating:

Step 3.1: according to the distance R of each moment target ship and website 1 and website 2_1tAnd R_2t, utilize triangle law of cosines, calculate the orientation �� of each moment target relative to website 1_1t, t=1,2...A;

Calculation formula is:

It is [R according to each target ship obtained relative to the polar coordinates position of website 1_1t,��_1t], t=1,2...A; It is transformed into latitude and longitude coordinates, it is designated as (Lon_t,Lat_t);

Step 3.2: according to multiple latitude and longitude coordinates of this flight path obtained, utilizes the curve L of this flight path of least square fitting₁; Then each moment source location (i.e. latitude and longitude coordinates) is calculated relative to matched curve L₁Distance be r_t, unit is km, t=1,2...A, then total distance value of A point is:

$F\left(j\right)=\underset{t=1}{\overset{A}{\Σ}}{r}_t---\left(2\right)$

Step 3.3: determine the flight path being associated in website 2 with the Article 1 flight path of website 1:

For other n-1 bar flight path T of website 2₂(R_j,V_j, t), then obtain n-1 result in the same way; And the flight path T can being associated with website 1 Article 1 flight path₂(R_q,V_q, t), q �� 1,2...n, it should meet:

$F\left(q\right)=\min F\left(j\right)=min\left(\underset{t=1}{\overset{A}{\Σ}}{r}_t\right)---\left(3\right)$

Namely with matching after curve L₁Error minimum;

Step 4: then to station 1 remaining m-1 bar flight path, look for the flight path of website 2 associated with it respectively, finally obtain the flight path that is associated in two website coverages.

Compared with prior art, the innovation part of the present invention embodies in the following areas:

Present method only needs each ground wave radar station only to have one to receive array element, utilizes the method for detecting and tracking integration just can obtain the undirected flight path of target and the coupling association results of dual station. The ground wave radar system of single array element receiving station can greatly reduce the size requirements to high-frequency ground wave radar receiving station, overcome the defect that traditional array formula ground wave radar system needs large size radar place, significantly improve the widespread use of target detection ground wave radar system.

The detecting and tracking doing target in the R-D-T of the distance-doppler-time of the ground wave radar data construct of the single array element of utilization is close-coupled processing, make full use of target face region and tieing up the feature of all continuous distribution on locus and echo strength attribute in time, it is to increase the detection accuracy of monostatic radar flight path. An association of dual station flight path also takes full advantage of the characteristic that target moves along a straight line in carving in short-term, it is achieved the coupling association of the undirected flight path of dual station, it is achieved that the detection of dual station ground wave radar multiple goal.

The present invention adopts and carries out dual station ground wave radar target detection based on detecting and tracking integrative detection thinking, avoid first put mark detect again Track In Track detection traditional detection method in need the process carrying out a mark and track association, decrease the tracking problem choosing the flight path mistake caused due to error thresholds, compensate for the position and movement properties that only utilize target and do not utilize the defect that target strength attribute causes, improve the flight path detection performance of target and the accuracy of the association results of dual station flight path, improve multi-target traces detection performance on the whole.

Accompanying drawing explanation

Fig. 1 is the basic procedure schematic diagram of the present invention.

Fig. 2 is the actual distribution figure of two within the scope of radar detection flight path.

Fig. 3 is two flight paths to the distance distribution plan in time of two websites.

The schematic diagram of two websites in Fig. 4 position.

The doubtful flight path distribution plan of Fig. 5 two undirected Track formings of website.

The distance error distribution plan of the doubtful flight path of Fig. 6 and its matched curve

Embodiment

Below in conjunction with accompanying drawing, the method for the present invention is described further:

As shown in Figure 1, a kind of method utilizing dual station list array element high-frequency ground wave radar detection ship target, comprise and first utilize the high-frequency ground wave radar of two websites (such as Fig. 4) to the ship objective emission pulse signal on sea, echoed signal is obtained after postponing through the regular hour, this echoed signal is utilized to obtain distance-doppler's data (R-D data), it is characterised in that also to comprise the following steps:

Step 1: high-frequency ground wave radar single pass R-D data construct R-D-T tri-dimension data utilizing certain time section T (T is less than one hour), is designated as Y (r, d, t), wherein r=1 ..., M, M is the range unit lattice number in R-D data, d=1 ... N, N are the doppler's unit lattice number in R-D data, t=1,, A, A are natural number, time period T being divided into A part, thus obtains A moment, t represents one of them moment;

Step 2: based on the R-D-T tri-dimension data Y (r, d, t) built, utilize dynamic programming method, obtain the undirected flight path result of detection of two websites, be respectively T₁(R_i,V_i, t) and T₂(R_j,V_j, t); Wherein, i=1,2...m, m are the flight path quantity that obtains of station 1, and j=1,2...n, n are the flight path quantity that station 2 obtains, and the flight path quantity that two websites obtain is the same, i.e. m=n; T=1 ..., A; Ri represents the distance of t target ship and website 12;

Fig. 2 is the actual distribution figure of two within the scope of radar detection flight path.

Fig. 3 be two flight paths to two websites distance over time, wherein T1 represents that flight path 1, T2 represents flight path 2.

Step 3: find the Article 1 flight path T with website 1 in station 2₁(R₁,V₁, t) mate the flight path of association mutually:

For each flight path T of website 2₂(R_j,V_j, t), j=1,2...n, successively with the Article 1 flight path T in website 1₁(R₁,V₁, t) carry out following calculating:

Step 3.1: according to the distance R of each moment target ship and website 1 and website 2_1tAnd R_2t, utilize triangle law of cosines, calculate the orientation �� of each moment target relative to website 1_1t, t=1,2...A;

Calculation formula is:

It is [R according to each target ship obtained relative to the polar coordinates position of website 1_1t,��_1t], t=1,2...A; It is transformed into latitude and longitude coordinates, it is designated as (Lon_t,Lat_t);

The undirected flight path combination at two stations being traveled through one time with this, the doubtful flight path obtained is shown in Fig. 5. Wherein T1-1 is the doubtful flight path representing that the undirected flight path in 1 the 1st article, station is formed with 2 the 1st articles, station track association, T1-2 is the doubtful flight path representing that the undirected flight path in 1 the 1st article, station is formed with 2 the 2nd articles, station track association, T2-1 is the doubtful flight path representing that the undirected flight path in 1 the 2nd article, station is formed with 2 the 1st articles, station track association, and T2-2 is the doubtful flight path representing that the undirected flight path in 1 the 2nd article, station is formed with 2 the 2nd articles, station track association.

Step 3.2: according to multiple latitude and longitude coordinates of this flight path obtained, utilizes the curve L of this flight path of least square fitting₁; Then each moment source location (i.e. latitude and longitude coordinates) is calculated relative to matched curve L₁Distance be r_t, unit is km, t=1,2...A, then total distance value of A point is:

$F\left(j\right)=\underset{t=1}{\overset{A}{\Σ}}{r}_t---\left(2\right)$

Step 3.3: determine the flight path being associated in website 2 with the Article 1 flight path of website 1:

For other n-1 bar flight path T of website 2₂(R_j,V_j, t), then obtain n-1 result in the same way; And the flight path T can being associated with website 1 Article 1 flight path₂(R_q,V_q, t), q �� 1,2...n, it should meet:

$F\left(q\right)=\min F\left(j\right)=min\left(\underset{t=1}{\overset{A}{\Σ}}{r}_t\right)---\left(3\right)$

Namely with matching after curve L₁Error minimum;

Step 4: then to station 1 remaining m-1 bar flight path, look for the flight path of website 2 associated with it respectively, finally obtain the flight path that is associated in two website coverages.

Fig. 6 gives the distance error distribution plan of doubtful flight path matched curve, and as can be seen from the figure, the error of stand the 1st article of undirected flight path of 1 and the 1st article of undirected flight path at station 2 is minimum, therefore illustrates that T1-1 is coupling association. Equally, the error of stand the 2nd article of undirected flight path of 1 and the 2nd article of undirected flight path at station 2 is minimum, and therefore T2-2 is also coupling flight path.

Claims (1)

1. one kind utilizes the method for dual station list array element high-frequency ground wave radar detection ship target, comprise and first utilize the high-frequency ground wave radar of two websites to the ship objective emission pulse signal on sea, echoed signal is obtained after postponing through the regular hour, this echoed signal is utilized to obtain distance-doppler's data (R-D data), it is characterised in that also to comprise the following steps:

Step 1: high-frequency ground wave radar single pass R-D data construct R-D-T tri-dimension data utilizing certain time section T (T is less than one hour), is designated as Y (r, d, t), wherein r=1 ..., M, M is the range unit lattice number in R-D data, d=1 ... N, N are the doppler's unit lattice number in R-D data, t=1,, A, A are natural number, time period T being divided into A part, thus obtains A moment, t represents one of them moment;

Step 2: based on the R-D-T tri-dimension data Y (r, d, t) built, utilize dynamic programming method, obtain the undirected flight path result of detection of two websites, be respectively T₁(R_i,V_i, t) and T₂(R_j,V_j, t); Wherein, i=1,2...m, m are the flight path quantity that obtains of station 1, and j=1,2...n, n are the flight path quantity that station 2 obtains, and the flight path quantity that two websites obtain is the same, i.e. m=n; T=1 ..., A; Ri represents the distance of t target ship and website 12;

Step 3: find the Article 1 flight path T with website 1 in station 2₁(R₁,V₁, t) mate the flight path of association mutually:

For each flight path T of website 2₂(R_j,V_j, t), j=1,2...n, successively with the Article 1 flight path T in website 1₁(R₁,V₁, t) carry out following calculating:

Step 3.1: according to the distance R of each moment target ship and website 1 and website 2_1tAnd R_2t, utilize triangle law of cosines, calculate the orientation �� of each moment target relative to website 1_1t, t=1,2...A;

Calculation formula is:

D is the distance (1) of website 1 website 2

It is [R according to each target ship obtained relative to the polar coordinates position of website 1_1t,��_1t], t=1,2...A; It is transformed into latitude and longitude coordinates, it is designated as (Lon_t,Lat_t);

Step 3.2: according to multiple latitude and longitude coordinates of this flight path obtained, utilizes the curve L of this flight path of least square fitting₁; Then each moment source location (i.e. latitude and longitude coordinates) is calculated relative to matched curve L₁Distance be r_t, unit is km, t=1,2...A, then total distance value of A point is:

$F\left(j\right)=\underset{t=1}{\overset{A}{\Σ}}{r}_t---\left(2\right)$

Step 3.3: determine the flight path being associated in website 2 with the Article 1 flight path of website 1:

For other n-1 bar flight path T of website 2₂(R_j,V_j, t), then obtain n-1 result in the same way; And the flight path T can being associated with website 1 Article 1 flight path₂(R_q,V_q, t), q �� 1,2...n, it should meet:

$F\left(q\right)=\min F\left(j\right)=\min \left(\underset{t=1}{\overset{A}{\Σ}}{r}_t\right)---\left(3\right)$

Namely with matching after curve L₁Error minimum;

Step 4: then to station 1 remaining m-1 bar flight path, look for the flight path of website 2 associated with it respectively, finally obtain the flight path that is associated in two website coverages.