CN105487060A - Two-channel four-slope modulation multi-target extraction method - Google Patents

Abstract

The invention discloses a two-channel four-slope modulation multi-target extraction method, which comprises the steps: S1) controlling, by a radar signal processing unit, a frequency generator to generate modulating signals, which are radiated, received and processed to obtain echo signal frequency spectrum; S2) extracting targets from the echo signal frequency spectrum through a constant false alarm method; S3) generating a four-slope modulation waveform, removing false targets in the targets to obtain real targets, and obtaining actual distance and actual velocity of the real targets; and S4) carrying out target matching and data fusion on results detected in different periods, and carrying out tracking filtering on multiple targets through Kalman filtering. The method effectively solves the problems of multi-target identification and detection as well as multi-target tracking, and has the advantages of high distance measurement and velocity measurement accuracy, good stability, low cost, simple processing and easy realization.

Description

A kind of multiple goal extracting method of binary channels four Slope Modulation

Technical field

The present invention relates to multiple goal extracting method, particularly a kind of multiple goal extracting method of binary channels four Slope Modulation.

Background technology

Along with the development of science and technology, Radar Technology is increasingly extensive in the application of civil area, as traffic speed radar, magnitude of traffic flow police radar, automobile collision avoidance radar, collision avoidance radar peculiar to vessel, security radar equipment etc.Radar detection relative to video detect have round-the-clock, not by advantages such as weather environment affect, there are good market outlook.At present, domestic use is the continuous wave Doppler radar based on Doppler effect more widely, when Doppler effect refers to and have relative motion between emissive source and recipient, the signal frequency received will change, the movement velocity of target can be calculated with the difference on the frequency of Received signal strength by measuring to transmit, traditional radar Doppler can only measure the motion velocity information of object, can not measure the distance of object, positional information, greatly limit range of application.

In patent, Wang Zhongming, Du Jing of Hua Ru Science and Technology Ltd. of Shenzhen have applied for patent-a kind of traffic speed-measuring method and device (publication No. CN102798862A), the patent describe a kind of method that the maximal rate of vehicles all in the scope of testing the speed is measured, the method is only applicable to the situation of single goal, when there being many cars in the scope of testing the speed, the corresponding relation of vehicle and speed cannot be distinguished.The dragon of Chengdu Zhong Yuanxin Electronic Science and Technology Co., Ltd. has rather applied for patent-a kind of continuous wave velocity radar (publication No. CN104251992A), the patent describe a kind of Modulation Continuous Wave Radar, this radar is mainly used in road traffic and tests the speed field, the true velocity of each vehicular traffic can be detected in real time, be only applicable to the situation of single goal.Xu Chengqi has applied for patent-detection radar (publication No. CN103366559A) of making a dash across the red light, the patent provides a kind of radar based on FMCW principle to make a dash across the red light capturing system, when vehicle enters or sail out of radar shadow, radar triggers the traveling picture that high definition camera captures vehicles peccancy.Toyota Motor Corporation rep. has applied for patent-Anti-collision control device (publication No. CN104816697A), the patent describes a kind of method using imageing sensor and radar sensor to detect vehicle accessory object.Tang Ling, the Wang Zhengxin of Shenzhen Graduate School of Peking University, Huang Lijun, Xu Feng, Meng Fanyu, Li Yanhong, Wang Xinan, emerging apply for patent-a kind of Continuous Wave with frequency modulation automobile anti-collision radar system (publication No. CN102788980A), this patent is by carrying out spectrum analysis to radar intermediate frequency signal, measurement result is converted to Distance geometry velocity information, maximum operating range 150 meters, range resolution 1 meter, velocity resolution 1 meter per second.Chen Jianshe, Wang Fei, He Dexi of Anhui Landun Photoelectron Co., Ltd. have applied for patent-a kind of speed-measuring radar system (publication No. CN102445689A) with function of distinguishing vehicle direction, the patent provides a kind of high-performance, low cost, easy to use, the speed-measuring radar system with function of distinguishing vehicle direction that can distinguish direction of traffic.Yang Fu, Chen Weibiao, He Yan of Shanghai Optics and Precision Mechanics institute, Chinese Academy of Sciences have applied for patent-high precision speed-measuring range laser radar system and the distance-finding method that tests the speed (publication No. CN101236253B), the patent provides a kind of method of measurement target Distance geometry speed, but multi-target processing ability is not good, and laser is easily by such environmental effects.Shang Haihong has applied for patent-mono-radar multilane Intelligent speed-measuring method and system (publication No. CN104424804A) thereof depending on the Yang Licheng of telecom technology co., ltd, wangkai, Lv Xiaoming, Yang Zhiwei, Pan Jinyi, the patent provides a kind of single radar multilane Intelligent speed-measuring method and system thereof, use a radar to detect multiple track and the camera triggering corresponding track is taken pictures.Microwave Velocity distance measuring equipment (publication No. CN103630888A) when the Li Heping of CAS Electronics Research Institute has applied for the high-precision real of patent-based on symmetric triangular LFMCW radar, this patent achieves the high-acruracy survey to velocity to moving target and distance, but not good enough for multiobject situation the method treatment effect.

Modulation Continuous Wave Radar has that emissive power is low, range resolution is high, antijamming capability is strong, volume is little, lightweight, advantages of simple structure and simple, is widely applied in numerous occasion.Traditional Modulation Continuous Wave Radar generally adopts sawtooth wave or the modulation of symmetric triangular ripple, by transmit and the difference on the frequency of echoed signal resolves target velocity and distance.

Sawtooth wave exists fuzzy in speed and distance, can not the speed of Measurement accuracy moving target and distance.

When there is n echo signal in receiver echoed signal, ascent stage and descending branch echoed signal respectively have n root spectral line with the difference frequency signal transmitted after Fast Fourier Transform (FFT), mate total n between two to ascent stage and descending branch data ²group data, n ²n group is only had for real goal, all the other n in group data ²-n group is false target, and therefore symmetric triangular ripple frequency modulated(FM) radar exists spectral pair problem under multi-target condition, is only applicable to the situation of single goal.

Summary of the invention

The object of this invention is to provide a kind of multiple goal extracting method of binary channels four Slope Modulation, efficiently solve multi-targets recognition, detection problem, multiple target tracking problem, and there is range finding, rate accuracy is high, good stability, low cost, process is simple, be easy to the advantages such as realization, the application demand of numerous occasion can be met, good market prospects.

In order to realize above object, the present invention is achieved by the following technical solutions:

A multiple goal extracting method for binary channels four Slope Modulation, be characterized in, the method comprises:

S1, radar signal processing unit controlled frequency generator produce modulation signal through overshoot, reception and process obtain echoed signal frequency spectrum;

S2, extracts target by CFAR Methods from echoed signal frequency spectrum;

S3, produce one or four Slope Modulation waveforms, the false target removed in described target draws real goal, and draws actual range and the actual speed of described real goal;

S4, carries out object matching, data fusion to the result of different cycles detection, utilizes Kalman filtering to carry out tracking filter to multiple goal.

Described step S1 comprises:

S1.1, radar signal processing unit controlled frequency generator produces modulation signal, and after power amplifier amplifies, output to emitting antenna, emitting antenna will transmit and radiate;

S1.2, described transmitting is reflected back two receiving antennas by one or more target;

S1.3, echoed signal obtains intermediate-freuqncy signal after low noise amplifier amplification, mixing, low-pass filtering;

S1.4, radar signal processing unit control analog to digital converter carries out sampling to intermediate-freuqncy signal and obtains signal spectrum.

Described step S3 comprises:

S3.1, produce one or four Slope Modulation waveforms, four described Slope Modulation waveforms are made up of two trapezoidal waveforms, and each trapezoidal waveform is formed by ascent stage, horizontal segment, descending branch; The modulation band-width of a trapezoidal waveform is B1, the modulation band-width of another trapezoidal waveform is B2, the modulating time of trapezoidal waveform and slope K 1, K2, K3, K4, and modulation band-width B1, B2 all can be controlled according to current goal quantity, target type by radar signal processing unit in real time;

S3.2, the through type (1) when measuring single goal, calculates radial distance R and the radial velocity V of the relative radar of target;

$R=\frac{cT}{8B}(f_r^{+}+f_r^{-}),v=\frac{c}{4f}(f_r^{-}-f_r^{+})---\left(1\right)$

Wherein, T is modulating time, and B is modulation band-width, f _r ⁺for ascent stage echoed signal and the beat frequency transmitted, f _r ^-for descending branch echoed signal and the beat frequency transmitted;

When measuring n target, whether n > 1, overlap two trapezoidal ascent stages, the lower section of liter and the Distance geometry velocity amplitude of horizontal segment with the Distance geometry velocity amplitude of target, is if so, then real goal, is then virtual target if not;

S3.3, utilizes the phase difference calculating target of binary channels echoed signal and the angle of radar, and obtains actual range and the actual speed of target.

Described step S4 comprises:

S4.1, arrange a target heap and deposit history target information, each history target in described target heap comprises ID, the distance factor, velocity factor, angular factors and system factor;

S4.2, calculates objective attribute target attribute according to formula (2) to the actual range of the target after resolving and actual speed at every turn:

Wherein, α is the distance factor, and β is velocity factor, and χ is angular factors, and κ is system factor, for objective attribute target attribute.

S4.3, judge objective attribute target attribute whether pile with target in the objective attribute target attribute of certain ID mate, be then if so, same target, and use the calculated value of current target properties to upgrade described ID; If not, then redistribute an ID for described target, and join in target heap;

S4.4, if certain ID in target heap is not upgraded for continuous Z time, then deletes corresponding target information from target heap;

S4.5, utilizes Kalman filtering to carry out recurrence tracking and filtering to multiple goal.

The present invention compared with prior art, has the following advantages:

1, a kind of four slope Modulation Continuous Wave Radar waveforms newly, a kind of double trapezoid modulation waveform is formed by two ascent stages, two descending brancies, two horizontal segments within a modulation period, by controlling ascent stage, the slope of descending branch, modulation band-width and modulating time in real time, effectively false target can be rejected.

2, this method effectively can solve range-velocity coupling problem and multiple goal spectral pair problem, accurately can obtain the distance of each target under multi-target condition, speed and azimuth information.

Accompanying drawing explanation

Fig. 1 is the process flow diagram of the multiple goal extracting method of a kind of binary channels four of the present invention Slope Modulation;

Fig. 2 is the structural drawing of the radar system that the present invention adopts;

Fig. 3 is CFAR disposal system schematic diagram;

Fig. 4 is CFAR treatment effect figure;

Fig. 5 is the schematic diagram of four Slope Modulation waveforms in the present invention;

Fig. 6 is that target multi slope of the present invention crosses figure;

Fig. 7 is phase place angle measurement schematic diagram of the present invention;

Fig. 8 is Kalman filtering design sketch.

Embodiment

Below in conjunction with accompanying drawing, by describing a preferably specific embodiment in detail, the present invention is further elaborated.

As shown in Figure 1, a kind of multiple goal extracting method of binary channels four Slope Modulation, comprises:

S1, radar signal processing unit controlled frequency generator produce modulation signal through overshoot, reception and process obtain echoed signal frequency spectrum;

S2, extracts target by CFAR Methods from echoed signal frequency spectrum;

S3, produce one or four Slope Modulation waveforms, the false target removed in described target draws real goal, and draws actual range and the actual speed of described real goal;

S4, carries out object matching, data fusion to the result of different cycles detection, utilizes Kalman filtering to carry out tracking filter to multiple goal.

As shown in Figure 2, above-mentioned step S1 comprises:

S1.1, radar signal processing unit controlled frequency generator produces modulation signal, and after power amplifier (PA) amplifies, output to emitting antenna, emitting antenna will transmit and radiate;

S1.2, described transmitting is reflected back two receiving antennas (1,2) by one or more target;

S1.3, echoed signal obtains intermediate-freuqncy signal after low noise amplifier (LNA) amplification, frequency mixer 3 mixing, wave filter 4 low-pass filtering;

S1.4, radar signal processing unit control analog to digital converter carries out sampling to intermediate-freuqncy signal and obtains signal spectrum.

Radar detection requires to extract useful target in the background that there is interference, interference comprises atural object, sleet, the noise jamming such as wave, receiver internal noise, artificial active interference and sourceless seism, and the adjacent objects interference etc. mixed in together with useful target, as shown in Figure 3, above-mentioned step S2 uses CFAR disposal system (OS-CFAR) to process radar echo signal, be specially: D is detected unit, front window and rear window are the reference samples of detected cells D, i sample is comprised in each window, 2i in a front window and rear window sample is sorted by order from small to large:

X(1)≤X(2)≤X(3)≤X(4)≤……≤X(2i)

Choose kth sample x (k) as the estimation Z of detecting device to clutter power level, T is regulation coefficient, detecting unit D and threshold value S=TZ is compared the judgement that can obtain current detection unit, namely during D>S, current detection cells D is judged as effective target, otherwise is noise jamming.In the diagram, solid line is echoed signal frequency spectrum, and dotted line is that round dot is the target obtained after CFAR process by the threshold value calculated that sorts.

Above-mentioned step S3 comprises:

S3.1, produce one or four Slope Modulation waveforms (see Fig. 5), four described Slope Modulation waveforms are made up of two trapezoidal waveforms, and each trapezoidal waveform is formed by ascent stage, horizontal segment, descending branch; The modulation band-width of a trapezoidal waveform is B1, the modulation band-width of another trapezoidal waveform is B2, the modulating time of trapezoidal waveform and slope K 1, K2, K3, K4, and modulation band-width B1, B2 all can be controlled according to current goal quantity, target type by radar signal processing unit in real time, be embodied as in Fig. 5 and transmit, dotted line is echoed signal;

S3.2, the through type (1) when measuring single goal, calculates radial distance R and the radial velocity V of the relative radar of target;

$R=\frac{cT}{8B}(f_r^{+}+f_r^{-}),v=\frac{c}{4f}(f_r^{-}-f_r^{+})---\left(1\right)$

Wherein, T is modulating time, and B is modulation band-width, f _r ⁺for ascent stage echoed signal and the beat frequency transmitted, f _r ^-for descending branch echoed signal and the beat frequency transmitted;

When measuring n target, 1, first trapezoidal ascent stage of n > and descending branch data mate total n between two ²group data, second trapezoidal ascent stage and descending branch data are mated between two also has n ²group data, and in the echoed signal of trapezoidal horizontal segment, only there is the velocity information of target, comprise n velocity information in horizontal segment frequency spectrum.As shown in Figure 6, the distance of real goal, speed and modulation band-width, modulation period have nothing to do, distance, the velocity amplitude of ascent stage, descending branch, horizontal segment can overlap, and the Distance geometry speed of false target and modulation band-width, modulation period are relevant, can not overlap completely.Whether, be if so, then real goal, be then virtual target if not if two trapezoidal ascent stages, the lower section of liter and the Distance geometry velocity amplitude of horizontal segment being overlapped with the Distance geometry velocity amplitude of target;

S3.3, utilize the phase difference calculating target of binary channels echoed signal and the angle of radar, and obtain actual range and the actual speed of target, see Fig. 7, dotted line is normal direction, is embodied as target direction, and the two angle is θ, because receiving antenna 1 and the spacing of receiving antenna 2 are d, the echoed signal that two antennas receive produces phase differential because there is a wave path-difference Δ R according to relation in Fig. 7:

λ is phase wave length, can try to achieve phase differential by the result of passage 1 and passage 2 Fast Fourier Transform (FFT) thus determine target direction θ.

Modulation Continuous Wave Radar completes the detection of a target in each cycle, obtains distance, the velocity information of current target, uses Kalman filter estimating target parameter, realize multiple target tracking between the cycle.First to determine that the target that radar detects at every turn is a fresh target or a target detected above before to target following.Suppose after original state t0 time data resolves, have 3 targets, distance, speed are respectively r01, v01, r02, v02, r03, v03, the ID of three targets is set to ID1, ID2, ID3 respectively, also there are 3 targets after t1 time data resolves, distance, speed are respectively r11, v11, r12, v12, r13, v13, how the distance in t1 moment, speed and ID are carried out pairing and become crucial.

Above-mentioned step S4 comprises:

S4.1, arrange a target heap and deposit history target information, each history target in described target heap comprises ID, the distance factor, velocity factor, angular factors and system factor;

S4.2, calculates objective attribute target attribute according to formula (2) to the actual range of the target after resolving and actual speed at every turn:

Wherein, α is the distance factor, and β is velocity factor, and χ is angular factors, and κ is system factor, for objective attribute target attribute, r, v, θ are distance, speed, the angle of target.

S4.3, judge objective attribute target attribute whether pile with target in the objective attribute target attribute of certain ID mate, be then if so, same target, and use the calculated value of current target properties to upgrade described ID; If not, then redistribute an ID for described target, and join in target heap;

S4.4, if certain ID in target heap is not upgraded for continuous Z time, then deletes the target information of correspondence from target heap, thus to achieve under multi-target condition the tracking of target, catches, abandons process.

S4.5, utilizes Kalman filtering to carry out recurrence tracking and filtering, in engineer applied to multiple goal, the data of radargrammetry are usually with noise, the actual value of desired data can not be obtained, measurement data distortion can be made in some cases, even cause data unavailable.Traditional filtering method requires that signal and noise have different frequency distribution, and Kalman filter effectively can reduce the impact of noise, target data is extracted from noise.To the effect of the filtering of certain target range as shown in Figure 8, stain is the range data that radar obtains certain target detection, and solid black lines is the data after Kalman filtering.

In sum, the multiple goal extracting method of a kind of binary channels four of the present invention Slope Modulation, efficiently solve multi-targets recognition, detection problem, multiple target tracking problem, and there is range finding, rate accuracy is high, good stability, low cost, process is simple, is easy to the advantages such as realization, the application demand of numerous occasion can be met, good market prospects.

Although content of the present invention has done detailed introduction by above preferred embodiment, will be appreciated that above-mentioned description should not be considered to limitation of the present invention.After those skilled in the art have read foregoing, for multiple amendment of the present invention and substitute will be all apparent.Therefore, protection scope of the present invention should be limited to the appended claims.

Claims (4)

1. a multiple goal extracting method for binary channels four Slope Modulation, it is characterized in that, the method comprises:

S1, radar signal processing unit controlled frequency generator produce modulation signal through overshoot, reception and process obtain echoed signal frequency spectrum;

S2, extracts target by CFAR Methods from echoed signal frequency spectrum;

S3, produce one or four Slope Modulation waveforms, the false target removed in described target draws real goal, and draws actual range and the actual speed of described real goal;

S4, carries out object matching, data fusion to the result of different cycles detection, utilizes Kalman filtering to carry out tracking filter to multiple goal.

2. the multiple goal extracting method of binary channels four Slope Modulation as claimed in claim 1, it is characterized in that, described step S1 comprises:

S1.1, radar signal processing unit controlled frequency generator produces modulation signal, and after power amplifier amplifies, output to emitting antenna, emitting antenna will transmit and radiate;

S1.2, described transmitting is reflected back two receiving antennas by one or more target;

S1.3, echoed signal obtains intermediate-freuqncy signal after low noise amplifier amplification, mixing, low-pass filtering;

S1.4, radar signal processing unit control analog to digital converter carries out sampling to intermediate-freuqncy signal and obtains signal spectrum.

3. the multiple goal extracting method of binary channels four Slope Modulation as claimed in claim 1, it is characterized in that, described step S3 comprises:

S3.1, produce one or four Slope Modulation waveforms, four described Slope Modulation waveforms are made up of two trapezoidal waveforms, and each trapezoidal waveform is formed by ascent stage, horizontal segment, descending branch; The modulation band-width of a trapezoidal waveform is B1, the modulation band-width of another trapezoidal waveform is B2, the modulating time of trapezoidal waveform and slope K 1, K2, K3, K4, and modulation band-width B1, B2 all can be controlled according to current goal quantity, target type by radar signal processing unit in real time;

S3.2, the through type (1) when measuring single goal, calculates radial distance R and the radial velocity V of the relative radar of target;

$R=\frac{cT}{8B}(f_r^{+}+f_r^{-}),v=\frac{c}{4f}(f_r^{-}-f_r^{+})---\left(1\right)$

Wherein, T is modulating time, and B is modulation band-width, f _r ⁺for ascent stage echoed signal and the beat frequency transmitted, f _r ^-for descending branch echoed signal and the beat frequency transmitted;

When measuring n target, whether n > 1, overlap two trapezoidal ascent stages, the lower section of liter and the Distance geometry velocity amplitude of horizontal segment with the Distance geometry velocity amplitude of target, is if so, then real goal, is then virtual target if not;

S3.3, utilizes the phase difference calculating target of binary channels echoed signal and the angle of radar, and obtains actual range and the actual speed of target.

4. the multiple goal extracting method of binary channels four Slope Modulation as claimed in claim 1, it is characterized in that, described step S4 comprises:

S4.1, arrange a target heap and deposit history target information, each history target in described target heap comprises ID, the distance factor, velocity factor, angular factors and system factor;

S4.2, calculates objective attribute target attribute according to formula (2) to the actual range of the target after resolving and actual speed at every turn:

Wherein, α is the distance factor, and β is velocity factor, and χ is angular factors, and κ is system factor, for objective attribute target attribute.

S4.3, judge objective attribute target attribute whether pile with target in the objective attribute target attribute of certain ID mate, be then if so, same target, and use the calculated value of current target properties to upgrade described ID; If not, then redistribute an ID for described target, and join in target heap;

S4.4, if certain ID in target heap is not upgraded for continuous Z time, then deletes corresponding target information from target heap;

S4.5, utilizes Kalman filtering to carry out recurrence tracking and filtering to multiple goal.