CN112162283A - All-section networking traffic radar multi-target detection system - Google Patents
All-section networking traffic radar multi-target detection system Download PDFInfo
- Publication number
- CN112162283A CN112162283A CN202010832404.2A CN202010832404A CN112162283A CN 112162283 A CN112162283 A CN 112162283A CN 202010832404 A CN202010832404 A CN 202010832404A CN 112162283 A CN112162283 A CN 112162283A
- Authority
- CN
- China
- Prior art keywords
- target
- radar
- distance
- long
- moment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 42
- 230000006855 networking Effects 0.000 title claims abstract description 12
- 238000001914 filtration Methods 0.000 claims abstract description 15
- 238000012545 processing Methods 0.000 claims abstract description 14
- 230000001133 acceleration Effects 0.000 claims description 24
- 239000011159 matrix material Substances 0.000 claims description 22
- 238000005259 measurement Methods 0.000 claims description 6
- 230000007704 transition Effects 0.000 claims description 6
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 abstract description 3
- 238000013179 statistical model Methods 0.000 abstract description 3
- 238000009434 installation Methods 0.000 description 17
- 230000004927 fusion Effects 0.000 description 14
- 230000005540 biological transmission Effects 0.000 description 6
- 238000004891 communication Methods 0.000 description 4
- 238000000034 method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000000737 periodic effect Effects 0.000 description 3
- 239000013598 vector Substances 0.000 description 3
- 101100129500 Caenorhabditis elegans max-2 gene Proteins 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/87—Combinations of radar systems, e.g. primary radar and secondary radar
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/66—Radar-tracking systems; Analogous systems
- G01S13/72—Radar-tracking systems; Analogous systems for two-dimensional tracking, e.g. combination of angle and range tracking, track-while-scan radar
- G01S13/723—Radar-tracking systems; Analogous systems for two-dimensional tracking, e.g. combination of angle and range tracking, track-while-scan radar by using numerical data
- G01S13/726—Multiple target tracking
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/91—Radar or analogous systems specially adapted for specific applications for traffic control
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/91—Radar or analogous systems specially adapted for specific applications for traffic control
- G01S13/92—Radar or analogous systems specially adapted for specific applications for traffic control for velocity measurement
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/411—Identification of targets based on measurements of radar reflectivity
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/418—Theoretical aspects
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention provides a multi-target detection system for traffic radar of a whole road section networking, wherein two long-distance radars are respectively responsible for detecting targets in a range of 25-550 m in the forward direction and the backward direction; the short-distance radar is responsible for detecting the target with the distance of 50m and the angle of 360 degrees; the edge processing platform receives target information sent by a long-distance radar and a short-distance radar, correlates the target track of the same target, finally obtains a fused target track, adopts seamless networking, and can realize real-time monitoring on the speed, type, position and other information of vehicles, pedestrians in 8 lanes and emergency lanes within the range of a full road of 1000 m; and performing target tracking by adopting an improved current statistical model and a Kalman filtering algorithm, and performing real-time stable tracking on the running tracks of passing road vehicles and pedestrians.
Description
Technical Field
The invention belongs to the technical field of long-distance traffic millimeter wave radars, and particularly relates to a full-road-section networking traffic radar multi-target detection system.
Background
Smartmicro UMRR-0C is today the highest performance traffic radar, with a beam coverage area of up to 100 degrees, while up to 250 meters (car) or 340 meters (truck) detection range, covering up to 8 lanes. The (X, Y, Z) cartesian or polar distance, azimuth, elevation and velocity vectors can be provided simultaneously, and up to 256 targets can be processed simultaneously.
The Huilyr wide area radar microwave detector adopts a two-dimensional active scanning type array radar microwave detection technology, can detect 200 meters, and can simultaneously detect at least 8 lanes; up to 128 targets can be tracked and detected simultaneously; the system can track and detect the position coordinates (x, y), the speed (Vx, Vy) and the vehicle type classification of each target, has graphical operation software, and displays the tracked condition of each target in the detection area and real-time information such as the instant position, the speed and the length of the vehicle in real time.
The radar detection range of a single system in the existing traffic radar detection system is within 200 m. The coverage area of a single system is small, the detection of a long-distance target needs to be realized by cascading a plurality of systems, and the number of the systems and the installation cost are increased. One of the problems solved by the present invention is to extend the coverage of a single system and reduce the cost.
The existing traffic radar detection system adopts 24G millimeter wave signals, however, the 24GHz frequency band lacks wide bandwidth, and the high-performance requirement cannot be met. The 77G millimeter wave signal in the invention uses a wide bandwidth signal and a higher radio frequency, and has the following advantages compared with a 24G sensor: (1) the SRR frequency band of 77GHz can provide 4GHz bandwidth, so that the distance resolution and precision are improved, and the requirement of high resolution under the condition of short distance is met. (2) The 77GHz sensor improves speed resolution by increasing the radio frequency signal frequency. (3) The 77GHz sensor emits higher radio frequency, and the sensor is smaller in size and easier to install.
The existing traffic radar detection system is not free of blind areas, information such as actual positions and speeds of targets cannot be obtained in the blind areas, and estimated values can only be output through a tracking algorithm. The problem to be solved by the present invention therefore also includes how to reduce the detection dead zone and enlarge the radar beam coverage.
The existing networking radar detection system uses a plurality of sensors, and a repeated coverage area exists among the plurality of radar sensors. When an object appears in a repeated coverage area and is detected by a plurality of sensors, the same object ID is easy to be inconsistent. Therefore, one of the difficulties of the invention comprises the association and fusion of multiple tracks of multiple radar sensors in the networking radar system, thereby ensuring that the target ID is unchanged and the track is unique in a long-distance range.
Disclosure of Invention
In view of this, the invention aims to provide a full-section networking traffic radar multi-target detection system, which can realize real-time stable tracking of the running tracks of vehicles and pedestrians on a passing road.
A full-section networking traffic radar multi-target detection system comprises two long-distance radars, a plurality of short-distance radars and an edge processing platform; the two long-distance radars are respectively responsible for detecting targets in the range of 25 m-550 m in the forward direction and the backward direction; the short-distance radar is responsible for detecting the target with the distance of 50m and the angle of 360 degrees; and the edge processing platform receives target information sent by the long-distance radar and the short-distance radar, associates the target tracks of the same target and finally obtains the fused target tracks.
Preferably, the long-range radar and the short-range radar track the target, specifically:
s81, importing the set target acceleration maximum value amaxAnd a minimum value of aminAnd a detection value;
s82, assuming that the target acceleration accords with the corrected Rayleigh distribution, calculating to obtain the acceleration variance as follows:
wherein
Is the mean value of the accelerations extracted from the measured values, amFor the acceleration extreme, adaptive adjustment is performed by the following formula:
wherein
Is an acceleration threshold; according to the predicted value X of the current k momentk|kAnd calculating to obtain a predicted value based on the moment k +1 of the acceleration model
Wherein F is a state transition matrix of the filter, and U is a control matrix of the filter;
s83, calculating Kalman gain: k is Pk+1|k×H×(H×Pk+1|k×HT+R)-1Where H is the measurement matrix, R is the noise covariance matrix, Pk+1|kFor the filtering covariance matrix to be equal to P at the last momentk|kSum of the filtered result and the system noise variance by the state transition matrix F, wherein the system noise variance and the acceleration variance σ2Is in direct proportion;
s84, calculating a predicted value X of the filteringk+1|k+1＝Xk+1|k+K(Zk+1-Xk+1|k) Wherein Z isk+1The detected value at the moment k + 1;
s85, measuring the value Z at the moment k +1k+1Track prediction value X corresponding to the existing k +1 momentk+1|k+1Performing minimum distance matching, and if the association is successful, updating the target track; updating the covariance matrix Pk+1|k+1＝(I-K×H)Pk+1|kCalculating a filtering predicted value at the next moment by using the filtering predicted value; otherwise, a new track is created for the unmatched targets.
Preferably, after the short-distance radar determines the three-dimensional information and the target speed of the target, clustering the target points through a DBscan algorithm to obtain each target; and extracting a target contour, measuring and calculating RCS corresponding to the target point through a radar equation, storing a one-dimensional range profile of the target point, extracting a range profile waveform entropy characteristic, a central moment characteristic and a target length characteristic, and calculating a receiving signal-to-noise ratio (SNR).
Preferably, after the long-distance radar determines the three-dimensional information and the target speed of the target, the target points are clustered through a DBscan algorithm, and therefore each target is obtained.
The invention has the following beneficial effects:
according to the full-road-section networking traffic radar multi-target detection system, two long-distance radars are respectively responsible for detecting targets in the range of 25-550 m in the forward direction and the backward direction; the short-distance radar is responsible for detecting the target with the distance of 50m and the angle of 360 degrees; the edge processing platform receives target information sent by a long-distance radar and a short-distance radar, correlates the target track of the same target, finally obtains a fused target track, adopts seamless networking, and can realize real-time monitoring on the speed, type, position and other information of vehicles, pedestrians in 8 lanes and emergency lanes within the range of a full road of 1000 m; and performing target tracking by adopting an improved current statistical model and a Kalman filtering algorithm, and performing real-time stable tracking on the running tracks of passing road vehicles and pedestrians.
Drawings
FIG. 1 is a diagram of an in-road installation radar connection scheme;
FIG. 2 is a diagram of a roadside mounted radar connection scheme;
FIG. 3 is a side view of a radar beam;
fig. 4 is a radar signal processing flow chart.
Detailed Description
The invention is described in detail below by way of example with reference to the accompanying drawings.
The detection system provided by the invention consists of 2 long-distance radars P1, a plurality of short-distance radars P2 and 1 edge processing platform P4. The installation form can be selected from in-road installation or roadside installation, the in-road installation is fixed on the portal frame through a special tool for top installation, the road test installation is fixed on the rod through a special tool for side installation, and then the road test installation is interacted with a user through traffic radar comprehensive management and control software P5. The road-side installation scheme is shown in fig. 1, and the road-side installation scheme is shown in fig. 2, wherein the left side is the radar installation position, and the right side is the beam coverage area diagram.
The invention mainly comprises the following modules: the system comprises a long-distance radar module, a blind-filling radar module and an edge calculation node module. The millimeter wave radar of the long-distance radar module and the millimeter wave radar of the blind-filling radar module are connected with the edge computing node through the network port, and the edge computing node can communicate with the traffic radar comprehensive control software through various connection modes such as the network port/serial port and 5G/4G network transmission so as to realize effective summary transmission of multi-source data.
The long-distance radar module comprises two long-distance radars, wherein signals transmitted by the long-distance radars are narrow-beam signals, long-distance detection is realized by utilizing the characteristics of power concentration and long transmission distance of the narrow-beam signals, and the coverage range of each long-distance radar beam is 25 m-550 m from the front to the back. The invention uses two long-distance radars to take charge of the forward direction and the backward direction respectively, so that the detection range of a single radar system reaches 1000 m. In large-scene long-distance application, a plurality of cascaded systems are required to cover the whole road section range, the radar system provided by the invention can reduce the demand of the radar system in the same detection range, and the cost is saved due to the camera, the installation equipment and other facilities matched with the radar system. Because the 1000m range detection is realized by adopting a mode of combining two long-distance radars, a blind area is not avoided at the short-distance position of the radars, and a blind area is covered by using a blind-supplement radar module subsequently. The blind-filling radar module comprises a plurality of short-distance radars, the short-distance radars transmit wide beam signals to realize short-distance wide-range detection, and the coverage range of each radar beam is 0-50 m. The number, the beam width and the installation angle of the blind-repairing radars need to be set according to the specific installation form of the radars, if the blind-repairing radars are installed in a road, four near-field radars can be used and are respectively installed in the front left direction, the front right direction, the rear left direction and the rear right direction, and the blind-zone detection radars can be used for compensating the detection blind zones existing in the long-distance radars. At this time, the beam width of the radar blind-filling radar is required to be greater than 90 degrees so as to realize 360-degree full coverage of the blind area. And the radar beam elevation angle needs to be designed to reach the detection range required by the blind-filling radar. As shown in fig. 3, Max1 is the maximum detection distance of the far-distance radar, Min1 is the minimum detection distance of the far-distance radar, Max2 is the maximum detection distance of the near blind-repairing radar, Min2 is the minimum detection distance of the near blind-repairing radar, β is the beam pitch angle of the blind-repairing radar, and H is the height. Determining the detection range of the blind-filling radar so as to meet the following requirements:
Min2＜Min1＜Max2
thus, β can be calculated by:
when Min2 is 0, the full range of radar beam coverage without blind areas can be achieved.
The long-distance radar and the short-distance radar have the following functions of (1) self-checking and periodic self-checking and safety protection measures; (2) the function of a transparent transmission channel is realized; (3) the system has the function of track information fusion; (4) the system has a triggering function, a counting function and an event detection function; (5) have RTC clock function; (6) the method has the function of updating the program on line. The millimeter wave radar can obtain the three-dimensional information of the target and the target speed by performing digital processing on the echo signal. And tracking the target by the tracking technology according to the measurement results of the multiple time points. After the processing is finished, the periodic measurement result is sent to the edge computing node module.
The edge computing node module has the following functions: (1) the edge computing processing platform is used as a junction for communication between the upper computer and the radar, and both a power supply and a communication signal need to be transmitted through the edge computing processing platform. And 8 radars can be connected at most. (2) Has self-checking and periodic self-checking functions and safety protection measures. (3) Has the function of a transparent transmission channel. (4) The system has a track information fusion function, summarizes data of various radars, realizes target fusion of multiple radars and realizes unique output of target ID. (5) The system has a trigger function, a statistic function and an event detection function. (6) Having the RTC clock function. (7) The method has the function of updating the program on line. (8) The function of outputting TTL level is provided. A maximum of 24 IO ports. (9) And target synchronization among multiple radars. (10) Support 5G (mobile, communication, telecommunication) network transmission, and is compatible with 4G (mobile, communication, telecommunication). The module receives data information sent by a long-distance radar and a blind-filling radar, and performs space alignment and time synchronization with the long-distance radar and the blind-filling radar at first; and determining a target track relation among different sensors through track association, and finally obtaining a fused target track by adopting a track fusion algorithm.
The invention realizes the identification and collection of the road traffic state information by effective technical means, and can be applied to various scenes: 1) real-time dynamic tracking of vehicles on the highway; 2) monitoring the traffic state of the urban road section; 3) assisting intelligent traffic management decision analysis; 4) tunnel navigation enhancements, and the like.
The method comprises the following steps: and (4) loading a radar program, and finishing parameter initialization and parameter setting.
Step two: the millimeter wave radar periodically transmits 77G frequency modulation continuous wave signals to the radiation area and receives echo signals reflected by the obstacles.
Step three: the echo signals and the transmitting signals are mixed to form intermediate frequency signals, and the intermediate frequency signals are converted into digital signals through ADC sampling.
Step four: and (3) carrying out digital processing (windowing and Fourier transform) on the single pulse digital signal to obtain a compressed target, and completing the phase correlation of the target in a distance dimension.
Step five: and obtaining target speed information by the pulse signals through two-dimensional Fourier transform, and determining the target through a CFAR detection algorithm.
Step six: the target azimuth angle and pitch angle information can be obtained by carrying out angle measurement on target signals received by the multiple antennas.
Step seven:
blind-filling radar: after the above steps are completed, the three-dimensional information of the target and the target speed can be preliminarily determined. Clustering target points through a DBscan algorithm, and selecting a peak Val maximum value in each class as a clustering point of all the points in the class, thereby obtaining each target. When the target distance of the clustering point is equal to rConst, calculating yMax-yMin ═ ySize and xMax-xMin ═ xSize according to the clustering result so as to extract a target contour, measuring and calculating RCS corresponding to the target point through a radar equation, storing a one-dimensional range profile of the target point dopplerIdx, and extracting a range profile waveform entropy characteristic, a center moment characteristic and a target length characteristic.
Where SNR is the received signal-to-noise ratio, PtFor transmitter power peak, GtFor increasing the transmitterYi, GRFor receiver gain, λ is the signal wavelength, TeIs the normal temperature coefficient, B is the receiver bandwidth, F is the noise coefficient, and σ is the target RCS.
Long-range radar: and clustering the target points through a DBscan algorithm, and simultaneously selecting the maximum value of the peakVal in each class as clustering points of all the points to be output, thereby obtaining each target.
Step eight: the blind-filling radar and the long-range radar track the target by utilizing multi-frame data and adopting an improved current statistical model and a Kalman filtering algorithm, and specifically comprises the following steps:
s82, selecting a motion model: assuming that the target acceleration conforms to the corrected Rayleigh distribution, the acceleration variance is calculated as:
wherein
Mean value of acceleration, a, extracted for the detected value ZmFor the acceleration extreme, the adaptive adjustment can be performed by the following formula:
wherein
Is an acceleration threshold. Calculating to obtain a predicted value X of the acceleration modelk+1|k＝F×Xk|k+U×
Wherein F is a state transition matrix of the filter, and U is a control matrix of the filter;
s83, initializing a target detection point, and calculating Kalman gain: k is Pk+1|k×H×(H×Pk+1|k×HT+R)-1Where H is the measurement matrix and R is the noise covarianceVariance matrix, variance of acceleration σ2Is in direct proportion; p is a filtering covariance matrix;
s84, calculating a predicted value X of the filteringk+1|k+1＝Xk+1|k+K(Zk+1-Xk+1|k) Wherein Z isk+1Is a detection value. Updating the covariance matrix Pk+1|k+1＝(I-K×H)Pk+1|kAnd calculating the filtering predicted value at the next moment by using the filtering predicted value.
S85, correlating the data and measuring the value Z at the k +1 momentk+1Track prediction value X corresponding to the existing k +1 momentk+1|k+1Performing minimum distance matching, and if the association is successful, updating the target track; otherwise, a new track is created for the unmatched targets.
Step nine: and sending information such as the target track and the distance, the speed, the angle and the like of the current target to the edge computing node module. The blind-repairing radar also needs to synchronously upload a target profile corresponding to the target track and the RCS to the edge node calculation node module.
S91, the edge computing node receives target data from a long-distance radar and a short-distance radar, radar coordinate system transformation is carried out through the installation angle to realize target space alignment, and time synchronization is carried out through the frame number frameIdx;
and S92, dividing the long-distance radar and the blind-filling radar into a target forward radar and a target backward radar according to the target course. Firstly, target association is carried out in a waveform repeated coverage area, target track coarse association is carried out through target starting point and ending point information (position coordinates, speed, acceleration and course), the track closest to the target track coarse association is selected as an association pair, and the track posture and the shape similarity between two tracks successfully associated with the target coarse association are judged through wavelet frame and multi-scale transformation. And judging the target track which needs to meet the following wavelet coefficient vector judgment criteria for the local detail features of the track as successful association:
where n is the length of the wavelet coefficient sequence, WTX1And WTX1Respectively represent each rulerThe wavelet coefficient vector in degree is the threshold value of Gaussian distribution under a given check level.
Representing target trajectory observation noise variances of different sensors;
s93, after the two tracks of the same target are associated, track fusion can be carried out by adopting a method based on minimum fusion error mean square error matrix track so as to stabilize the tracks.
In the formula Pf(k | k) represents the mean square error matrix of the fusion error at time k, F1And F2State transition matrices, D, for sensor 1 and sensor 2, respectively1And D2And the weight coefficient is the weight coefficient of the flight path fusion. And (3) fusing the flight paths meeting the conditions, wherein the fused flight paths are represented as follows:
Xf＝D1X1+D2X2
according to the spatial distribution of the radar, the track fusion sequence is that the track fusion of the forward radar and the blind-filling radar is carried out firstly, then the track fusion of the blind-filling radar is carried out, and finally the track fusion of the blind-filling radar and the backward long-distance radar is carried out;
s94, after the steps are completed, the object ID uniqueness can be realized in the detection range of 1000m of a single system. After the targets are roughly associated, the target feature information xSize, ySize, RCS, a range profile entropy feature, a central moment feature and a target length feature are used as input parameters, target classification results in different systems are obtained through SVM training, and then the algorithms in the steps 2 and 3 are used for carrying out track association and fusion so as to realize the uniqueness of the ID of the targets in the range of the whole road section;
and S95, after the above processing is completed, the edge computing node module performs trigger detection, statistical detection and event detection, and finally transmits the data to the traffic radar comprehensive control software P5 and interacts with the user through P5.
In summary, the above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (4)
1. A full-section networking traffic radar multi-target detection system is characterized by comprising two long-distance radars, a plurality of short-distance radars and an edge processing platform; the two long-distance radars are respectively responsible for detecting targets in the range of 25 m-550 m in the forward direction and the backward direction; the short-distance radar is responsible for detecting the target with the distance of 50m and the angle of 360 degrees; and the edge processing platform receives target information sent by the long-distance radar and the short-distance radar, associates the target tracks of the same target and finally obtains the fused target tracks.
2. The system of claim 1, wherein the long-range radar and the short-range radar track the target, and specifically comprises:
s81, importing the set target acceleration maximum value amaxAnd a minimum value of aminAnd detecting the measured value;
s82, assuming that the target acceleration accords with the corrected Rayleigh distribution, calculating to obtain the acceleration variance as follows:
wherein
Is the mean value of the accelerations extracted from the measured values, amFor the acceleration extreme, adaptive adjustment is performed by the following formula:
wherein
Is an acceleration threshold; according to the predicted value X of the current k momentk|kAnd calculating to obtain a predicted value based on the moment k +1 of the acceleration model
Wherein F is a state transition matrix of the filter, and U is a control matrix of the filter;
s83, calculating Kalman gain: k is Pk+1|k×H×(H×Pk+1|k×HT+R)-1Where H is the measurement matrix, R is the noise covariance matrix, Pk+1|kFor the filtering covariance matrix to be equal to P at the last momentk|kSum of the filtered result and the system noise variance by the state transition matrix F, wherein the system noise variance and the acceleration variance σ2Is in direct proportion;
s84, calculating a predicted value X of the filteringk+1|k+1＝Xk+1|k+K(Zk+1-Xk+1|k) Wherein Z isk+1The detected value at the moment k + 1;
s85, measuring the value Z at the moment k +1k+1Track prediction value X corresponding to the existing k +1 momentk+1|k+1Performing minimum distance matching, and if the association is successful, updating the target track; updating the covariance matrix Pk+1|k+1＝(I-K×H)Pk+1|kCalculating a filtering predicted value at the next moment by using the filtering predicted value; otherwise, a new track is created for the unmatched targets.
3. The system as claimed in claim 1, wherein after the short-range radar determines the three-dimensional information and the target speed of the target, the target points are clustered through a DBscan algorithm, so as to obtain each target; and extracting a target contour, measuring and calculating RCS corresponding to the target point through a radar equation, storing a one-dimensional range profile of the target point, extracting a range profile waveform entropy characteristic, a central moment characteristic and a target length characteristic, and calculating a receiving signal-to-noise ratio (SNR).
4. The system as claimed in claim 1, wherein after the long-distance radar determines the three-dimensional information of the target and the target speed, the target points are clustered by the DBscan algorithm, thereby obtaining each target.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010832404.2A CN112162283B (en) | 2020-08-18 | Multi-target detection system of full-road-section networking traffic radar |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010832404.2A CN112162283B (en) | 2020-08-18 | Multi-target detection system of full-road-section networking traffic radar |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112162283A true CN112162283A (en) | 2021-01-01 |
| CN112162283B CN112162283B (en) | 2024-07-02 |
Family
ID=
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113611112A (en) * | 2021-07-29 | 2021-11-05 | 中国第一汽车股份有限公司 | Target association method, device, equipment and storage medium |
| CN114035171A (en) * | 2021-11-11 | 2022-02-11 | 西安迅尔电子有限责任公司 | Real-time algorithm for processing target track of security radar of phased array system nuclear power station |
| CN114049793A (en) * | 2021-10-22 | 2022-02-15 | 三维通信股份有限公司 | Vehicle avoidance system and vehicle avoidance method |
| CN114445661A (en) * | 2022-01-24 | 2022-05-06 | 电子科技大学 | Embedded image identification method based on edge calculation |
| CN114578338A (en) * | 2022-05-05 | 2022-06-03 | 南京隼眼电子科技有限公司 | Radar device |
| CN115472022A (en) * | 2022-09-06 | 2022-12-13 | 同盾科技有限公司 | Fusion speed measuring method, fusion speed measuring device, storage medium and electronic equipment |
| CN117129982A (en) * | 2023-08-28 | 2023-11-28 | 河北德冠隆电子科技有限公司 | Linear scanning angle accurate adjustable data dynamic fusion perception radar |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202491798U (en) * | 2012-03-28 | 2012-10-17 | 大连华泰中天商用设施有限公司 | Anti-collision device for automobile |
| CN104020468A (en) * | 2014-06-20 | 2014-09-03 | 南京恩瑞特实业有限公司 | Close range target updating cycle calculating method based on secondary radar |
| WO2014160027A1 (en) * | 2013-03-13 | 2014-10-02 | Image Sensing Systems, Inc. | Roadway sensing systems |
| CN105329237A (en) * | 2014-08-08 | 2016-02-17 | 现代自动车株式会社 | Method and apparatus for avoiding a vehicle collision with low power consumption based on conversed radar sensors |
| CN106004659A (en) * | 2016-08-03 | 2016-10-12 | 安徽工程大学 | Surrounding environment perception system for vehicles and control method of surrounding environment perception system |
| CN106932771A (en) * | 2017-03-30 | 2017-07-07 | 成都紫瑞青云航空宇航技术有限公司 | A kind of radar simulation targetpath tracking and system |
| CN107458361A (en) * | 2017-08-10 | 2017-12-12 | 南京超靖界智能科技有限公司 | Vehicle safety accessory system and its control method |
| CN107728138A (en) * | 2017-09-15 | 2018-02-23 | 电子科技大学 | A kind of maneuvering target tracking method based on current statistical model |
| CN108646237A (en) * | 2018-05-09 | 2018-10-12 | 西安电子科技大学 | Radar maneuvering target tracking optimization method based on current statistical model |
| CN110304054A (en) * | 2019-07-03 | 2019-10-08 | 国唐汽车有限公司 | A kind of automatic emergency brake method based on Multi-sensor Fusion |
| CN110515073A (en) * | 2019-08-19 | 2019-11-29 | 南京慧尔视智能科技有限公司 | The trans-regional networking multiple target tracking recognition methods of more radars and device |
| CN110542898A (en) * | 2019-06-18 | 2019-12-06 | 同济大学 | Radar group-based vehicle behavior continuous tracking detection system and method |
| CN210101616U (en) * | 2019-06-19 | 2020-02-21 | 新石器慧通（北京）科技有限公司 | Anti-collision system and vehicle |
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202491798U (en) * | 2012-03-28 | 2012-10-17 | 大连华泰中天商用设施有限公司 | Anti-collision device for automobile |
| WO2014160027A1 (en) * | 2013-03-13 | 2014-10-02 | Image Sensing Systems, Inc. | Roadway sensing systems |
| CN104020468A (en) * | 2014-06-20 | 2014-09-03 | 南京恩瑞特实业有限公司 | Close range target updating cycle calculating method based on secondary radar |
| CN105329237A (en) * | 2014-08-08 | 2016-02-17 | 现代自动车株式会社 | Method and apparatus for avoiding a vehicle collision with low power consumption based on conversed radar sensors |
| CN106004659A (en) * | 2016-08-03 | 2016-10-12 | 安徽工程大学 | Surrounding environment perception system for vehicles and control method of surrounding environment perception system |
| CN106932771A (en) * | 2017-03-30 | 2017-07-07 | 成都紫瑞青云航空宇航技术有限公司 | A kind of radar simulation targetpath tracking and system |
| CN107458361A (en) * | 2017-08-10 | 2017-12-12 | 南京超靖界智能科技有限公司 | Vehicle safety accessory system and its control method |
| CN107728138A (en) * | 2017-09-15 | 2018-02-23 | 电子科技大学 | A kind of maneuvering target tracking method based on current statistical model |
| CN108646237A (en) * | 2018-05-09 | 2018-10-12 | 西安电子科技大学 | Radar maneuvering target tracking optimization method based on current statistical model |
| CN110542898A (en) * | 2019-06-18 | 2019-12-06 | 同济大学 | Radar group-based vehicle behavior continuous tracking detection system and method |
| CN210101616U (en) * | 2019-06-19 | 2020-02-21 | 新石器慧通（北京）科技有限公司 | Anti-collision system and vehicle |
| CN110304054A (en) * | 2019-07-03 | 2019-10-08 | 国唐汽车有限公司 | A kind of automatic emergency brake method based on Multi-sensor Fusion |
| CN110515073A (en) * | 2019-08-19 | 2019-11-29 | 南京慧尔视智能科技有限公司 | The trans-regional networking multiple target tracking recognition methods of more radars and device |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113611112A (en) * | 2021-07-29 | 2021-11-05 | 中国第一汽车股份有限公司 | Target association method, device, equipment and storage medium |
| CN114049793A (en) * | 2021-10-22 | 2022-02-15 | 三维通信股份有限公司 | Vehicle avoidance system and vehicle avoidance method |
| CN114035171A (en) * | 2021-11-11 | 2022-02-11 | 西安迅尔电子有限责任公司 | Real-time algorithm for processing target track of security radar of phased array system nuclear power station |
| CN114445661A (en) * | 2022-01-24 | 2022-05-06 | 电子科技大学 | Embedded image identification method based on edge calculation |
| CN114445661B (en) * | 2022-01-24 | 2023-08-18 | 电子科技大学 | Embedded image recognition method based on edge calculation |
| CN114578338A (en) * | 2022-05-05 | 2022-06-03 | 南京隼眼电子科技有限公司 | Radar device |
| CN115472022A (en) * | 2022-09-06 | 2022-12-13 | 同盾科技有限公司 | Fusion speed measuring method, fusion speed measuring device, storage medium and electronic equipment |
| CN115472022B (en) * | 2022-09-06 | 2024-03-22 | 同盾科技有限公司 | Fusion speed measuring method, fusion speed measuring device, storage medium and electronic equipment |
| CN117129982A (en) * | 2023-08-28 | 2023-11-28 | 河北德冠隆电子科技有限公司 | Linear scanning angle accurate adjustable data dynamic fusion perception radar |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110208793B (en) | Auxiliary driving system, method, terminal and medium based on millimeter wave radar | |
| CN110456343B (en) | Instant positioning method and system based on FMCW millimeter wave radar | |
| US11479262B2 (en) | Geographically disparate sensor fusion for enhanced target detection and identification in autonomous vehicles | |
| CN112098990B (en) | Method for detecting and tracking medium and high speed vehicle by vehicle-mounted high-resolution millimeter wave radar | |
| CN109358322B (en) | Forward target detection radar and method | |
| Cui et al. | 3D detection and tracking for on-road vehicles with a monovision camera and dual low-cost 4D mmWave radars | |
| CN110596731A (en) | Active obstacle detection system and method for metro vehicle | |
| CN213303095U (en) | Road test device based on artificial intelligence | |
| CN107783123A (en) | Pilotless automobile complex environment anticollision MMW RADAR SIGNAL USING processing system and method | |
| CN112859033A (en) | Target detection method, device and related equipment | |
| CN107783132A (en) | Autonomous driving vehicle anticollision millimetre-wave radar system and signal processing method | |
| CN112162283B (en) | Multi-target detection system of full-road-section networking traffic radar | |
| CN112162283A (en) | All-section networking traffic radar multi-target detection system | |
| Segawa et al. | Preceding vehicle detection using stereo images and non-scanning millimeter-wave radar | |
| da Silva et al. | Novel post-Doppler STAP with a priori knowledge information for traffic monitoring applications: basic idea and first results | |
| Bialer et al. | Performance evaluation of wide aperture radar for automotive applications | |
| CN112590781A (en) | Vehicle anti-collision warning method and device based on millimeter wave phased array radar | |
| Barros Cardoso da Silva et al. | Novel post-Doppler STAP with a priori knowledge information for traffic monitoring applications: basic idea and first results | |
| Zheng | (Retracted) Laser radar-based intelligent vehicle target recognition and detection system using image detection technology | |
| CN114236488B (en) | Target classification method, target classification device, terminal device and storage medium | |
| CN113554876B (en) | Multi-millimeter wave radar domain control system | |
| KR102364288B1 (en) | An intelligent radar sensor for a smart crossroad system | |
| US20230305135A1 (en) | Electronic device, method for controlling electronic device, and program | |
| Streck et al. | Comparison of two different radar concepts for pedestrian protection on bus stops | |
| CN116500621B (en) | Radar blind area early warning method based on double-subframe obstacle recognition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |