阅读说明：本技术 一种基于快速fmcw雷达的弱运动物体监测方法 (Weak moving object monitoring method based on rapid FMCW radar ) 是由 房晓辉 韩海峰 于 2019-10-10 设计创作，主要内容包括：本发明公开了一种基于快速FMCW雷达的弱运动物体监测方法,具体包括如下内容：通过安装在道路上方的快速FMCW雷达产生频率调制的发射信号；发射信号和快速FMCW雷达接收到的回波经过混频得到差拍信号s,经过二次混频变成正交基带信号；经过ADC采样的数字信号数据输出给后端处理电路板,进行计算处理,在对得到的差拍信号进行二维FFT中的多普勒FFT之前,将具有微弱信号的多普勒回波从静止杂波中分离,利用相干相位差法抑制包括干扰在内的平稳回波,再进行多普勒FFT,得到速度信息。(The invention discloses a method for monitoring a weakly moving object based on a rapid FMCW radar, which specifically comprises the following steps: generating a frequency modulated transmission signal by a fast FMCW radar installed above a road; the transmitting signal and the echo received by the rapid FMCW radar are subjected to frequency mixing to obtain a beat signal s, and the beat signal s is converted into an orthogonal baseband signal through secondary frequency mixing; and outputting the digital signal data sampled by the ADC to a back-end processing circuit board, performing calculation processing, separating Doppler echoes with weak signals from static clutter before performing Doppler FFT in two-dimensional FFT on the obtained beat signals, suppressing stable echoes including interference by using a coherent phase difference method, and performing Doppler FFT to obtain speed information.)

1. A method for monitoring a weakly moving object based on a rapid FMCW radar is characterized by comprising the following steps: generating a steep slope linear frequency modulated transmission signal by a rapid FMCW radar installed above a road; the transmitting signal and the echo received by the rapid FMCW radar are subjected to frequency mixing to obtain a beat signal s, and the beat signal s is converted into an orthogonal baseband signal through secondary frequency mixing; and outputting the digital signal data sampled by the ADC to a back-end processing circuit board, performing calculation processing, separating Doppler echoes with weak signals from static clutter before performing Doppler FFT in two-dimensional FFT on the obtained beat signals, suppressing stable echoes including interference by using a coherent phase difference method, and performing Doppler FFT to obtain speed information.

2. A method for monitoring a weakly moving object based on a fast FMCW radar as set forth in claim 1, wherein the mth range unit X (m, k) is:

the front half part of the formula (3) is clutter, and the rear half part is a moving target;

the method for suppressing the stable echo including the interference by using the coherent phase difference method mainly comprises the following two steps: coherent integration and coherent subtraction; in the first step of coherent integration, the X (m, K) integrals after distance FFT of K sweeps in one frame are summed and then divided by K to find the average:

the second step, coherent subtraction, according to equations (3) and (4), obtains:

the above formula shows that only the phase change of the moving object is contained; and then performing Doppler FFT on the gamma (m, k) to obtain speed information.

3. The method for monitoring a weakly moving object based on a fast FMCW radar as set forth in claim 1 or 2, wherein the back-end processing circuit board, performing the calculation process further includes: the distance is determined, the beat frequency of the beat signal in different frequency sweeps is the same, but the phase changesA relatively sensitive response to a small change in distance can be obtained by the following equation:

wherein

the beat signal s can be expressed as a function of parameters k, l:

where k represents the frequency of the sweep and l represents the beat signal of a single sweep passing f_sThe number of the sampling points obtained by sampling,

4. A method as claimed in claim 1, wherein the frequency sweep period of the fast FMCW radar linear modulation waveform is between 1us and 100us, the transmission sequence in frames contains 50 to 1000 frequency sweeps, and the bandwidth of a single frequency sweep is in GHz.

5. The fast FMCW radar-based weak moving object monitoring method as set forth in claim 1, wherein the fast FMCW radar front end is composed of an open-loop VCO circuit, a transmitting antenna, a receiving antenna, and a mixer; the frequency mixer mixes the transmitted wave and the received wave to generate beat signals, the beat signals are output to the rear-end ADC, the rear-end DSP chip receives data of the ADC, after distance FFT is conducted on the beat signals, speed information with weak signals is separated from clutter, then speed FFT operation is conducted, and finally speed distance information of the weak moving object can be accurately obtained.

Technical Field

The invention relates to a method for monitoring a weakly moving object based on a rapid FMCW radar, in particular to a method for monitoring the distance and the speed of a weakly moving target from clutter of a static object, and belongs to the technical field of radar monitoring.

Background

In the road target monitoring technology, only a radar monitoring system is not influenced by environmental factors such as weather, light, haze and the like. Therefore, the radar system is widely used. Among radar systems, Frequency Modulated Continuous Wave (FMCW) radar is preferred because it has lower hardware requirements and a simpler design than pulse wave radar.

In a conventional FMCW radar, for a stationary object, the beat frequency fb of the beat signal depends on the delay time between the radar transmission and the echo, determining the distance of the target from the radar. However, for moving objects there must be a doppler effect and influence the beat frequency fb, so a single slope (only rising edge or only falling edge) modulated sawtooth FMCW, due to the large time width, the single sweep period often shows 1 to 10ms, a so-called range-velocity coupling phenomenon exists and cannot be eliminated. Thus, the beat frequency on a single sweep does not accurately monitor distance and velocity.

In order to obtain a unique value for the distance between the target and the radar and the target speed, in conventional FMCW radars modulation is usually performed by means of alternating rising and falling frequency sweeps, i.e. frequency linear modulation with triangular waves. Therefore, two straight lines with opposite slopes are determined in the R-v plane, and the unique values of the target distance and the target speed can be determined. However, in multi-target monitoring, because radar parameters are the same, a plurality of groups of parallel straight lines are generated when the frequency sweep of the transmitted signals of a plurality of targets is in a frequency band in which the frequency sweep is increased and decreased, for example, when two targets are detected, four intersection points exist, and then two of the intersection points are false targets. As the number of targets increases, the number of false targets increases by a factor.

In order to overcome the multivalue property, another straight line with scanning frequency is often added, and three straight lines intersect at one point, so that a real target is determined. However, as the number of targets increases, it is often necessary to use other sweep frequencies to more easily resolve multiple values. This increases the design difficulty, as well as the hardware requirements.

And in the traditional FMCW radar system, the range shift and Doppler shift are often in the kHz range, which is also the flicker noise range of the active device, and limits the FMCW radar resolution. In order to solve the above problems, the fast FMCW is adopted, that is, the FMCW wave with the increased sweep frequency of the sawtooth modulation is in the order of GHz bandwidth and us period, which are difficult to be achieved by the closed loop system formed by the traditional VCO and PLL, so in order to achieve the requirement of the fast FMCW, the open loop radar system is applied, that is, the phase locked loop PLL is not used for generating the linear chirp signal, but the DAC signal is directly used for controlling the VCO to output the required linear frequency signal with high slope. The distance-speed coupling problem of the sawtooth wave can be solved, and the multi-value problem of triangular wave modulation is avoided.

In addition, on a real road, there are many strong echoes from stationary objects. In this case, although the clutter is clearly separated from the pedestrian in position in the doppler frequency domain, the side lobe of the clutter may completely mask the pedestrian. To prevent this distortion, a popular solution is to add a window function before the FFT, which provides side lobe suppression for unwanted echoes in the range and doppler frequency domains. However, due to the windowing function, the frequency spectrum tends to overflow on adjacent spectral lines, so that the clutter mainlobe in the doppler frequency domain is widened, thereby covering weak moving objects such as pedestrians.

Disclosure of Invention

The purpose of the invention is as follows: aiming at the problems and the defects in the prior art, the invention provides a method for monitoring a weakly moving object based on a rapid FMCW radar. The fast FMCW radar system has frequency sweep period of FMCW linear modulation waveform of 1-100 us, and the transmission sequence in frame unit includes 50-1000 frequency sweeps, and the bandwidth of each frequency sweep is GHz unit, so that the scanning frequency of the transmitted FMCW signal is far greater than that of the traditional FMCW signal, and the Doppler frequency shift produced by the moving target is negligible relative to the distance frequency shift, and the distance-speed coupling phenomenon produced by FMCW single frequency sweep is eliminated. The invention carries out two-dimensional fast Fourier transform on the obtained beat signal to obtain distance and speed information. And eliminating clutter generated by a static object by using a coherent phase method so as to distinguish a weak moving target.

The technical scheme is as follows: a monitoring method of a weak moving object based on a rapid FMCW radar specifically comprises the following contents:

generating a steep slope linear frequency modulated transmission signal by a rapid FMCW radar installed above a road; the transmitting signal and the received echo are subjected to frequency mixing to obtain a beat signal s, and the beat signal s is converted into an orthogonal baseband signal through secondary frequency mixing; the digital signal data sampled by the ADC is output to a back-end processing circuit board for calculation processing, because the Doppler effect generated by a weak moving object is weak and is easily covered by the reflected wave and noise of a peripheral static object, before Doppler FFT in two-dimensional FFT is carried out on the obtained beat signal, a coherent phase difference method is used for restraining stable echo including interference, Doppler information in the echo with the weak signal is separated from clutter, and therefore a slow-speed moving target covered in the clutter before the Doppler information is reliably monitored, and then Doppler FFT is carried out to obtain speed information.

The back end processing circuit board performs calculation processing: firstly, the distance is determined, the beat frequency (distance frequency shift) of the beat signal in different frequency sweeps is the same, but the phase changes

A relatively sensitive response to a small change in distance can be obtained by the following equation:

wherein

v is the velocity of the monitored object, T_chirpOne chirp (chirp) period, c is the speed of light, f₀The center frequency of the FMCW waveform is swept. The target motion velocity can be represented by the FFT of the phase change between different sweeps, which is called doppler FFT.

The beat signal s can be expressed as a function of parameters k, l:

where k represents the number of sweeps, i.e. chirp number, and l represents the beat signal of a single sweep passing f_sThe number of the sampling points obtained by sampling,

representing an initial phase offset;

let the r-th distance element X (r, k), i.e. the r-th element is obtained by calculation of the distance FFT according to the equation:

the front half part of the formula (3) is clutter, and the rear half part is a moving target; that is, the phase of the moving target will change periodically with the frequency of the sweep, and the phase of the clutter will not change;

the coherent phase difference method is used for suppressing the stable echo including the interference so as to reliably monitor the slow moving target which is hidden in the clutter, and the method mainly comprises the following two steps: coherent integration and coherent subtraction; in the first step of coherent integration, the X (m, K) integrals after distance FFT of K sweeps in one frame are summed and then divided by K to find the average:

generally, the value of K is 50 to 1000, the phase change of the moving target on the frequency sweep is counteracted, and the phase of the clutter

Will remain; the second step, coherent subtraction, according to equations (3) and (4), obtains:

the above formula shows that only the phase change of the moving object is contained; and then performing Doppler FFT on the gamma (m, k) to obtain weak movement velocity information.

Drawings

FIG. 1 is a schematic diagram of a method for identifying distance and velocity of a moving object in the prior art;

FIG. 2 is a block diagram of a fast FMCW radar system front end;

FIG. 3 is a schematic diagram of a method of identifying a weakly moving object in the method of the present invention;

1. radar front end, 2, pedestrian, 3, stationary object, 4, sampling frequency, 5, mixer, 6, VCO, 7, transmitting antenna, 8, receiving antenna, 9, frequency in quadrature mixing, 10, ADC sampling 16, waveform of transmitting wave on f-t coordinate, 17, waveform in time domain of beat signal after ADC sampling, 18, array matrix module, 19, coherent averaging module, 20, coherent subtraction module.

Detailed Description

The present invention is further illustrated by the following examples, which are intended to be purely exemplary and are not intended to limit the scope of the invention, as various equivalent modifications of the invention will occur to those skilled in the art upon reading the present disclosure and fall within the scope of the appended claims.

A monitoring method of a weak moving object based on a rapid FMCW radar, wherein a radar front end is composed of an open-loop VCO circuit, a transmitting antenna, a receiving antenna and a mixer; the frequency mixer mixes the transmitted wave and the received wave to generate a beat signal, the beat signal is subjected to orthogonal frequency mixing to reduce the sampling frequency and is output to the rear-end ADC, the rear-end DSP chip receives data of the ADC, because the Doppler effect generated by a weak moving object is weak, the Doppler effect is easily covered by the reflected wave and noise of a peripheral static object, after distance FFT is carried out on the beat signal, the speed information with the weak signal is separated from clutter, then speed FFT operation is carried out, and finally the speed distance information of the weak moving object can be accurately obtained. The invention separates the Doppler echo with weak signal from the static clutter before the Doppler FFT in the two-dimensional FFT, and utilizes the coherent phase difference method to restrain the stable echo including the interference, thereby reliably monitoring the slow moving target such as the pedestrian and the like which is hidden in the clutter before. The method specifically comprises the following steps:

generating a frequency modulated transmission signal by a fast FMCW radar installed above the road, the transmission signal frequency and time being in a sweep period T_chirpInner linear relation, B bandwidth, f₀Is the center frequency, T_chirpFor a sweep period，S＝B/T_chirpIs the sweep frequency; the echo received by the fast FMCW radar is equivalent to the delay of t_dThe transmission signal of (1); the beat signal s is obtained through frequency mixing, and becomes a quadrature baseband signal through secondary frequency mixing. As shown in fig. 1, the digital signal data sampled by the ADC is output to the back-end processing circuit board for calculation processing.

Firstly, the distance is determined, and for the fast FMCW, the frequency shift f caused by the distance is determined if the sweep frequency S is large_rDoppler shift f much greater than velocity_vThen beat frequency f_bMainly determined by the range shift. Also after performing the FFT, in the frequency domain, due to the sweep period T_chirpSmaller, with a spectral resolution of 1/T_chirp，f_vWhen the frequency spectrum cannot be larger than the frequency spectrum resolution, the beat frequency f cannot be distinguished by mapping on the frequency spectrum_bDistance features are represented singly, the Doppler shift f_vCan be completely ignored. The FFT at this time is called a range FFT. When the peak corresponds to the frequency position, it is considered as f_r. Thus, the distance-velocity coupling problem is not considered, and the distance frequency shift information is directly obtained through the distance FFT, so that the distance frequency shift information is obtained

The distance is obtained where S is the sweep frequency, i.e. the slope of chirp in the FMCW waveform, d is the distance between the object and the radar, and c is the speed of light, i.e. the speed of the electromagnetic wave. This is also where fast FMCW is superior to conventional FMCW.

As mentioned above, the beat frequency (i.e., the distance frequency shift) of the beat signal is the same in different frequency sweeps, but the phase changes

A relatively sensitive response to a small change in distance can be obtained by the following equation:

wherein

v is the velocity of the monitored object, T_chirpFor a chirp period, c is the speed of light, f₀The center frequency of the FMCW waveform. The target motion velocity can be represented by the FFT of the phase change between different sweeps, which is called doppler FFT.

In summary, the beat signal s can be expressed as a function with k and l as parameters:

where k represents the frequency of the sweep and l represents the beat signal of a single sweep passing f_sThe number of sampling points obtained by sampling.

Indicating the initial phase offset.

The mathematical expression for fast FMCW two-dimensional FFT can be described as first performing an M-point fourier transform on s (l, k) single frequency sweep:

where L is a sampling point in one sweep, k represents a sweep frequency number, M is a number of points from the FFT, hereinafter referred to as a distance unit number, and M is a number of the distance unit, and M is 1 to M.

In a frame range, K chirp exists, so that K points exist in the same distance unit, and N-point Fourier transform is performed on the K points in the same unit distance of s (l, K):

where N is the number of doppler FFT points, which are called doppler cells, and N is the number of doppler cells.

Based on the above-mentioned fast FMCW radar to monitor the road target, the strong noise generated by the stationary object may submerge the weak moving target in its main lobe or multiple side lobes, so a suitable window is generally added in the time domain. However, the time domain windowing, although suppressing the side lobes, also causes the spectrum in the frequency domain to spread over adjacent spectral lines, so that the clutter mainlobe in the doppler frequency domain becomes wider, masking the moving target.

In order to overcome the problem, the invention provides a method for monitoring a weak moving object by correspondingly improving the two-dimensional FFT, namely, before Doppler FFT processing, a motion component is saved, and a zero Doppler component is suppressed. Let the r-th distance unit X (r, k), which is obtained by the calculation of equation (3), be simplified as:

wherein A is_cThe amplitude of the clutter is the amplitude of the clutter,the phase of the clutter is the phase of the clutter,

for the phase change caused by the doppler effect of the kth frequency sweep, the first half of equation (3) is clutter and the second half is a moving target. That is, the phase of the moving object will change periodically with the scan frequency, and the phase of the clutter will not change.

A method of monitoring weak motion based on coherent phase differences is proposed here. The method mainly comprises two steps: coherent integration and coherent subtraction. In the first step of coherent integration, we sum the distance FFT's X (m, K) integrals of K sweeps in a frame, then divide by K to find the average.

Generally, K is 50 to 1000, and due to the above-mentioned periodic phase change, the phase change of the moving object in the sweep frequency is cancelled out, and the noise remains. The second step, coherent subtraction, according to equations (3) and (4), obtains:

the above equation shows that only the phase change of the moving object is contained. And then performing Doppler FFT on the gamma (m, k) to obtain speed information.

FIG. 2 shows the whole process of the present invention, the transmitting signal transmits a frame of FMCW signals with K frequency sweeps, after frequency mixing, a beat signal is obtained, and f is used_sThe sampling rate of the sweep frequency is L.

As shown in fig. 3, M-point fast fourier transform is performed on each of the K frequency sweeps, and the M frequency sweeps are stored in rows of the array matrix module 18. The coherent averaging module 19 performs the process of equation 9 to obtain an average value. The coherence subtraction module 20 is a process of equation (5).

After passing through the modules 19 and 20, only the phase of the moving target changes, and the purpose of suppressing clutter is achieved. At this time, doppler FFT is performed on the result of equation (5), and the spectrum information of the phase can be obtained. The speed information is obtained by conversion of the formula (1).