Orthogonal CW radar weak motion optimization detection method

Abstract

The invention discloses an orthogonal CW radar weak motion optimization detection method, and the method comprises the steps: carrying out the displacement pre-motion, collecting an echo signal generated by the pre-motion, and carrying out the fitting to form a reference ellipse, which is used for extracting a mismatch parameter caused by the channel mismatch of an orthogonal CW radar echo signal; the equivalent mathematical expectation of the radar echo signal generated by the reflection of the weak motion target is calculated to realize the compression of the noise, and then the extracted mismatch parameter is used to carry out channel calibration on the echo signal after the noise is compressed, so that the effect of accurately detecting the weak motion is achieved. According to the method, amplitude mismatch and phase mismatch parameters of echo signals are obtained through channel mismatch parameter pre-extraction, mismatch compensation is carried out on the signals, each radar system only needs to carry out channel mismatch parameter prediction once, noise is suppressed through noise compression based on mathematical expectation and by means of probability statistical characteristics of the echo signals, and the noise suppression efficiency is improved. The lower limit of the weak displacement motion amplitude detected by the radar is lower, and the detection precision is higher.

Description

Technical Field

[0001] The present invention relates to a technology in the field of radar motion detection, in particular to an orthogonal continuous wave (CW) radar weak motion optimization detection method. Background Art

[0002] Orthogonal CW radar is sensitive to motion and is widely used in non-contact displacement extraction. However, the hardware of the orthogonal CW radar system is not ideal, which causes the echo signal to have amplitude and phase distortion (mismatch) caused by channel mismatch, such as Figure 1 As shown. At the same time, when the target displacement motion amplitude is much smaller than the radar wavelength, the echo signal phase fluctuation is weak and the signal-to-noise ratio is poor. These two problems deteriorate the accuracy of orthogonal CW radar demodulation of weak motion. To address the channel mismatch problem, there is an ellipse calibration method, which extracts the amplitude and phase mismatch parameters by fitting the echo signal data points on an ellipse. However, in weak motion scenarios, the arc of the fitted ellipse is extremely short, the signal-to-noise ratio is poor, and accurate ellipse information cannot be obtained, resulting in the extraction of incorrect mismatch parameters. In addition, the ellipse calibration calculation process is complex and not suitable for embedded implementation of orthogonal CW radar. To address the problem of low signal-to-noise ratio, the additive noise of the echo signal is often compressed and suppressed by a filter, such as using convolution for denoising, or separating the signal and noise by the subspace method. These methods have limited denoising and require a large amount of data calculation, and are not suitable for use in portable radars that rely on embedded programming. Summary of the invention

[0003] In view of the above-mentioned deficiencies in the prior art in weak motion detection, the present invention proposes an orthogonal CW radar weak motion optimization detection method, which obtains the amplitude mismatch and phase mismatch parameters of the echo signal through channel mismatch parameter pre-extraction and performs mismatch compensation on the signal. Each radar system only needs to perform channel mismatch parameter prediction once, and through noise compression based on mathematical expectation, the probability statistical characteristics of the echo signal are used to suppress the noise, so that the lower limit of the amplitude of weak displacement motion detectable by the radar is lower and the detection accuracy is higher.

[0004] The present invention is achieved through the following technical solutions:

[0005] The invention relates to an orthogonal CW radar weak motion optimization detection method. After displacement pre-motion and acquisition of echo signals generated by the pre-motion to form a reference ellipse for fitting to extract mismatch parameters caused by channel mismatch of the orthogonal CW radar echo signal, the equivalent mathematical expectation of the radar echo signal generated by the reflection of the weak motion target is calculated to realize noise compression, and then the extracted mismatch parameters are used to perform channel calibration on the echo signal after noise compression, thereby achieving the effect of accurately detecting weak motion.

[0006] The pre-motion means that before testing weak motion, a long displacement motion of a strongly reflecting target (such as an angular reflection) is tested first, and the orthogonal echo signal generated by the target motion is recorded. Each radar system only needs to perform one pre-motion test.

[0007] The reference ellipse refers to: a phase diagram is drawn with the radar orthogonal echo signals as the horizontal and vertical coordinates. Due to the orthogonal signal amplitude mismatch and phase mismatch caused by channel mismatch, the phase diagram that should be circular appears as an ellipse. The length of the ellipse arc depends on the length of the target movement. The pre-motion displacement is long enough, so its echo signal can be fitted to form a complete ellipse. The complete ellipse can make the extraction of mismatch parameters more accurate.

[0008] The mismatch parameter refers to: due to the imperfect phase shifter, power divider and mixer in the orthogonal radar system hardware, the echo signal channel mismatch is caused, so that the orthogonal echo signals that should have equal amplitudes and a phase difference of 90° do not meet the conditions in terms of amplitude and phase; the adaptation parameter includes the amplitude mismatch parameter A e , that is, the ratio of the amplitudes of the two orthogonal echo signals and the phase mismatch parameter That is, the phase difference between the two echo signals is subtracted by 90°.

[0009] The equivalent mathematical expectation calculation refers to: dividing the collected echo signal into M segments with a data length of S, taking the average of the S data points in each segment to obtain an equivalent data point, and obtaining M data points to replace the original echo signal, where: M·S=L, L is the number of collected echo signal data points; according to the principle of random signal analysis, for a one-dimensional random signal c(n), its mathematical expectation When n is a sufficiently large finite value (for example, n = S), For a certain signal s(n), the mathematical expectation represents the average value of the signal over the entire period of time, that is, E[s(n)] = average[s(n)]. When the sampling rate is much greater than the rate at which s(n) changes over time, it is considered that s(n) is almost unchanged within a segment, so And the mathematical expectation calculation satisfies linearity, E[s(n)+c(n)]=E[s(n)]+E[c(n)].

[0010] The channel calibration refers to: using the reference ellipse to calculate the radar system mismatch parameter A e and The orthogonal echo signals are calibrated to have equal amplitudes and a phase difference of 90°.

[0011] The present invention relates to a system for implementing the above method, comprising: a channel calibration and noise compression module, a DC bias calibration module and a phase demodulation module, wherein: the channel calibration and noise compression module performs channel calibration processing according to the radar orthogonal channel mismatch parameters extracted by pre-motion, and at the same time uses the equivalent mathematical expectation calculation of the echo signal to compress the additive noise, so as to achieve the effect of compensating the amplitude and phase distortion of the orthogonal echo signal and improve its signal-to-noise ratio. The DC bias calibration module uses the orthogonality of the echo signal to remove the DC bias voltage of the orthogonal echo signal and normalize the signal amplitude. The phase demodulation module performs phase solution calculation according to the relationship between the phase information of the echo signal and the weak motion of the target, and finally extracts the weak motion of the target. Technical Effects

[0012] The present invention uses a pre-motion method to accurately extract the mismatch parameters of the radar system with a small amount of data, completes the channel calibration, and uses the statistical characteristics of the determined signal and additive noise in the echo signal to propose an equivalent mathematical expectation calculation to achieve noise suppression, greatly improving the accuracy of radar detection of weak motion. Compared with the prior art, the present invention overcomes the problem of difficult accurate channel calibration and low signal-to-noise ratio when the radar tests weak motion. The present invention conveniently and accurately calibrates the channel while using the random signal characteristics of the additive noise in the echo signal for denoising. In general, the present invention effectively improves the ability and effect of orthogonal CW radar to detect weak motion at close range. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is the block diagram of the existing orthogonal CW radar system;

[0014] Figure 2 It is a flow chart of the present invention;

[0015] Figure 3 This is a schematic diagram for calculating equivalent mathematical expectation;

[0016] Figure 4 Figure 1. Experimental setup for channel mismatch parameter extraction.

[0017] Figure 5 Schematic diagram of experimental results for channel mismatch parameter extraction;

[0018] Figure 6 Diagram of the experimental setup for detecting the radial motion of nanofilm curling;

[0019] Figure 7 Schematic diagram of the experimental results for detecting the radial motion of nanofilm curling. DETAILED DESCRIPTION

[0020] like Figure 1As shown, the orthogonal CW radar system involved in this embodiment includes: a transmitting antenna, a receiving antenna, a voltage-controlled oscillator (VCO), a power amplifier (PA), a low-noise amplifier (LNA), a down-conversion module, a baseband operational amplifier (Amp) and a data acquisition card (DAQ), wherein: the down-conversion module includes a 90-degree phase shifter (Phase Shifter) and two mixers (Mixer). The VCO generates a continuous sine wave signal of a fixed frequency and is divided into two paths, one of which is sent to the down-conversion module as a local oscillator (LO) signal, and the other is amplified by the power amplifier (PA) and then transmitted by the transmitting antenna. After the signal reaches the target to be measured, it is modulated by the target's weak motion and generates an echo. The receiving antenna receives the echo and sends it to the down-conversion module through a low-noise amplifier (LNA). In the down-conversion module, the echo signal is mixed with the original local oscillator signal and the local oscillator signal after 90° phase shift, and the mismatched orthogonal echo signals I(t) and Q(t) are output. After DAQ data acquisition, I(t) and Q(t) are obtained to obtain digital signals I(n) and Q(n).

[0021] The mismatched orthogonal echo signal includes: Where: A is the signal amplitude, A e and are the amplitude mismatch parameter and phase mismatch parameter of the orthogonal channel, λ is the orthogonal CW radar carrier wavelength, x(t) is the target weak motion, DC I / DC Q is the DC bias, N I (t) / N Q (t) is additive noise, which is usually defined as a zero-mean Gaussian random process, that is, E[N I (t)]=0,E[N Q (t)] = 0, It is residual phase noise and can be ignored in radar close-range detection scenarios.

[0022] The digital signal includes: Where: To meet the equivalent mathematical expectation calculation conditions, the sampling frequency f s Need to meet f s >>2d[x(t)] / λd[t].

[0023] like Figure 2 As shown in the figure, a weak motion optimization detection method based on the above system involved in this embodiment is completed through the channel calibration and noise compression module to extract the radar channel mismatch parameters and the channel calibration and noise reduction of the weak motion echo signal, and then the DC offset calibration is used to compensate the DC in the echo signal. I / DC Q, and normalize the amplitude A, and finally extract the weak motion x(n) contained in the phase through displacement demodulation.

[0024] like Figure 2 As shown, the extraction of radar channel mismatch parameters refers to: using radar to detect a sufficiently long displacement motion m(t) (for example, an amplitude of 0.5λ), and converting the echo data I cal (n) and Q cal (n) Stored in I cal (n) is the horizontal axis, Q cal (n) is the ordinate, the data points are plotted in the two-dimensional phase diagram, a complete reference ellipse is formed by fitting, and the ellipse coefficient is extracted. Finally, the amplitude mismatch parameter A is calculated based on the ellipse equation. e and phase mismatch parameters

[0025] The channel calibration comprises:

[0026] 1) Calculate the equivalent mathematical expectation: Divide the collected data I(n) / Q(n) into M segments, each segment is S long, and make the equivalent mathematical expectation of the mth segment (1≤m≤M), that is, Where m = S / f s , which can be considered as "slow time".

[0027] 2) Calibrate channel mismatch: Use the estimated Ae and to I E (m) / Q E (m) is calibrated, where: I E (m) is the reference value and no further calibration is required, that is, I CE (m) = I E (m); in:

[0028] The DC offset calibration is to use The characteristics of I CE and Q CE Perform circle fitting and calculate the center of the circle (DC I ,DC' Q ) and the radius A of the circle, and finally obtain the normalized orthogonal echo signal and in:

[0029] The displacement demodulation is to extract the phase and calculate the micro-motion trajectory contained in the phase using the inverse tangent function, that is,

[0030] Through specific experiments, Figure 4 The standard slide (Zaber T-NA08A50-KT04Μ) shown in the figure generates two uniform linear motions u(t) and m(t), and the u(t) motion displacement is 0.0125cm (=0.01λ@24GHz), which is used to simulate the traditional method of extracting mismatch parameters. The m(t) motion displacement is 0.625cm (=0.5λ@24GHz), which is the pre-motion echo signal I proposed by the present invention for extracting the channel mismatch parameters of the radar. cal (n) and Q cal (n). The setting parameters are: the radar transmits a continuous sine wave signal with a frequency of f c =24GHz, carrier wavelength λ = 1.25cm, distance between slide and radar d0 = 15cm. The experimental results are as follows Figure 5 As shown, in Figure 5 In (a), since u(t) moves weakly, the fitting forms an error ellipse, and the error mismatch parameters extracted by the error ellipse cannot be used for channel calibration. cal (n) and Q cal (n) forms a complete ellipse and perfectly matches the fitting ellipse. The radar channel mismatch coefficient calculated by the fitting ellipse is A e =1.06,

[0031] The radial displacement of the nanofilm is detected to test the detection accuracy of the method for weak target motion. The experimental scene is as follows: Figure 6 As shown in (a), the radar setting parameters are: the radar transmits a continuous sine wave signal with a frequency of f c =24GHz, carrier wavelength λ = 1.25cm. DAQ sets sampling frequency f s =10KHz, data segment length S = 10. Figure 6 As shown in (b), the parameters of the nanofilm are: before curling, the length and width of the film are both l = 150 μm, and the film performs a circular curling motion with a fixed diameter of d = 52.9 μm. The curling displacement is captured by a high-speed camera. The duration of the movement is less than 0.2 seconds. The radar is facing the film surface, and the distance from the film is d0 = 8 cm. The experimental results are as follows Figure 7 As shown, we can see that Figure 7 (a) and (c) are the spectrum and signal-to-noise ratio of the original I / Q signal and the demodulated result using the original signal. Figure 7 (b) is the spectrum and signal-to-noise ratio of the I / Q signal after channel calibration and noise compression. Channel calibration does not affect the signal-to-noise ratio of the signal, so noise compression suppresses the additive noise by 2.3 dB and 2.1 dB respectively. Figure 7 (d) is the signal demodulation result after noise compression only, and Figure 7 (c) By comparison, we can see that the root mean square error (RMSE) is reduced by 5.1 μm. After further calibration, the demodulation results are as follows: Figure 7 As shown in (e), the root mean square error is further reduced by 1.5μm. Compared with the traditional method, channel calibration and noise compression technology reduce the overall demodulation error of the radar for weak motion by 6.6μm, accounting for about 15.4% of the traditional method. Weak displacement motion becomes more obvious, improving the accuracy of radar detection of displacement motion.

[0032] Compared with the existing technology, this method improves the orthogonal CW radar channel mismatch and noise, overcomes the problem that the amplitude and phase mismatch of the echo signal are difficult to calibrate when the radar detects weak motion, and improves the signal-to-noise ratio of the echo signal, ultimately improving the orthogonal radar's detection capability and detection accuracy for weak motion.

[0033] The above-mentioned specific implementation can be partially adjusted in different ways by those skilled in the art without departing from the principle and purpose of the present invention. The protection scope of the present invention shall be based on the claims and shall not be limited by the above-mentioned specific implementation. Each implementation scheme within its scope shall be subject to the constraints of the present invention.

Claims

1. An orthogonal CW radar weak motion optimization detection method, characterized in that: Through displacement pre-motion and acquisition of echo signals generated by pre-motion, a reference ellipse is formed by fitting to extract the mismatch parameters caused by the channel mismatch of the orthogonal CW radar echo signal. Then, the equivalent mathematical expectation of the radar echo signal generated by the reflection of the weak moving target is calculated to achieve noise compression. Then, the extracted mismatch parameters are used to perform channel calibration on the echo signal after noise compression, thereby achieving the effect of accurately detecting weak motion.

2. The orthogonal CW radar weak motion optimization detection method according to claim 1 is characterized in that: The pre-motion means that before testing weak motion, a long displacement motion of a strongly reflective target is tested first, and the orthogonal echo signal generated by the target motion is recorded. Each radar system only needs to perform one pre-motion test.

3. The orthogonal CW radar weak motion optimization detection method according to claim 1 is characterized in that: The reference ellipse refers to: a phase diagram is drawn with the radar orthogonal echo signals as the horizontal and vertical coordinates respectively. Due to the orthogonal signal amplitude mismatch and phase mismatch caused by channel mismatch, the phase diagram that should be circular appears as an ellipse. The arc length of the ellipse depends on the length of the target movement. The pre-motion displacement is long enough, so its echo signal can be fitted to form a complete ellipse. The complete ellipse can make the extraction of mismatch parameters more accurate.

4. The orthogonal CW radar weak motion optimization detection method according to claim 1 is characterized in that: The mismatch parameter refers to: due to the imperfect phase shifter, power divider and mixer in the orthogonal radar system hardware, the echo signal channel mismatch is caused, so that the orthogonal echo signals that should have equal amplitudes and a phase difference of 90° do not meet the conditions in terms of amplitude and phase; the adaptation parameter includes the amplitude mismatch parameter A e , that is, the ratio of the amplitudes of the two orthogonal echo signals and the phase mismatch parameter That is, the phase difference between the two echo signals is subtracted by 90°.

5. The orthogonal CW radar weak motion optimization detection method according to claim 1 is characterized in that: The equivalent mathematical expectation calculation refers to: dividing the collected echo signal into M segments with a data length of S, taking the average of the S data points in each segment to obtain an equivalent data point, and obtaining M data points to replace the original echo signal, where: M·S=L, L is the number of collected echo signal data points; according to the principle of random signal analysis, for a one-dimensional random signal c(n), its mathematical expectation When n is a sufficiently large finite value (for example, n = S), For a certain signal s(n), the mathematical expectation represents the average value of the signal over the entire period of time, that is, E[s(n)] = average[s(n)]. When the sampling rate is much greater than the rate at which s(n) changes over time, it is considered that s(n) is almost unchanged within a segment, so And the mathematical expectation calculation satisfies linearity, E[s(n)+c(n)]=E[s(n)]+E[c(n)].

6. The orthogonal CW radar weak motion optimization detection method according to claim 1 is characterized in that: The channel calibration refers to: using the reference ellipse to calculate the radar system mismatch parameter A e and The orthogonal echo signals are calibrated to have equal amplitudes and a phase difference of 90°.

7. An orthogonal CW radar weak motion optimization detection system implementing the method described in any one of claims 1 to 6, characterized in that: include: The module comprises a channel calibration and noise compression module, a DC bias calibration module and a phase demodulation module, wherein: the channel calibration and noise compression module performs channel calibration processing according to the radar orthogonal channel mismatch parameters extracted by pre-motion, and uses the equivalent mathematical expectation calculation of the echo signal to compress the additive noise, so as to achieve the effect of compensating the amplitude and phase distortion of the orthogonal echo signal and improve its signal-to-noise ratio; the DC bias calibration module uses the orthogonality of the echo signal to remove the DC bias voltage of the orthogonal echo signal and normalize the signal amplitude; the phase demodulation module performs phase deconvolution calculation according to the relationship between the phase information of the echo signal and the target weak motion, and finally extracts the target weak motion.

8. The orthogonal CW radar weak motion optimization detection system according to claim 7 is characterized in that: The orthogonal CW radar system includes: a transmitting antenna, a receiving antenna, a voltage controlled oscillator (VCO), a power amplifier (PA), a low noise amplifier (LNA), a down-conversion module, a baseband operational amplifier (Amp) and a data acquisition card (DAQ), wherein: the down-conversion module includes a 90-degree phase shifter (Phase Shifter) and two mixers (Mixer), VCO generates a continuous sine wave signal with a fixed frequency and is divided into two paths. One path is sent to the down-conversion module as the local oscillator (LO) signal, and the other path is amplified by the power amplifier (PA) and then transmitted by the transmitting antenna. After the signal reaches the target to be measured, it is modulated by the target's weak motion and generates an echo. The receiving antenna receives the echo and sends it to the down-conversion module through a low noise amplifier (LNA). In the down-conversion module, the echo signal is mixed with the original local oscillator signal and the local oscillator signal after 90° phase shift, and the mismatched orthogonal echo signals I(t) and Q(t) are output. After DAQ data acquisition, I(t) and Q(t) are obtained to obtain digital signals I(n) and Q(n); The mismatched orthogonal echo signal includes: Where: A is the signal amplitude, A e and are the amplitude mismatch parameter and phase mismatch parameter of the orthogonal channel, λ is the orthogonal CW radar carrier wavelength, x(t) is the target weak motion, DC I / DC Q is the DC bias, N I (t) / N Q (t) is additive noise, which is usually defined as a zero-mean Gaussian random process, that is, E[N I (t)]=0,E[N Q (t)] = 0, is the residual phase noise; The digital signal includes: Where: To meet the equivalent mathematical expectation calculation conditions, the sampling frequency f s Need to meet f s >>2d[x(t)] / λd[t].

Images