Two-dimensional vibration measurement method based on single millimeter wave radar

Through a single low-cost millimeter wave radar and signal reflector, combined with Fourier transform and beamforming technology, the low efficiency and high cost of two-dimensional vibration measurement in the light difference environment in the existing technology is solved, and high accuracy is achieved. All-weather two-dimensional vibration monitoring, suitable for large-scale deployment.

CN118915005APending Publication Date: 2024-11-08SHANDONG UNIV
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410977460.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The existing two-dimensional vibration measurement technology is poor in environments with poor light, has high equipment costs, is complex in deployment, and is difficult to achieve high-precision micro vibration measurement.

Method used

Using a single low-cost millimeter wave radar, the designed millimeter wave signal reflector and radar are deployed, and the vibration signals are separated using Fourier transform and beamforming technology to construct geometric models, combining the visual and non-visit signals for two-dimensional vibrations. Calculate, using virtual radar to provide second-dimensional vibration observations.

Benefits of technology

It realizes all-weather, low-cost, non-invasive two-dimensional vibration monitoring, provides more comprehensive health information, reduces equipment costs and deployment complexity, and is suitable for large-scale deployment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118915005A_ABST
    Figure CN118915005A_ABST
Patent Text Reader

Abstract

The invention belongs to the technical field of radar measurement, and particularly relates to a two-dimensional vibration measurement method based on a single millimeter-wave radar, which comprises the following steps of: firstly, deploying a designed millimeter-wave signal reflector by using a reflecting surface in an environment, then separating a line-of-sight signal and a non-line-of-sight signal from a captured millimeter-wave signal, and extracting a vibration signal from the line-of-sight signal and the non-line-of-sight signal; distance and angle parameters are calculated by using an advanced signal processing method, and finally the two-dimensional vibration of the target is calculated by using the designed geometric model. The method has the advantages that the reflecting surface in the environment is used for constructing a virtual radar to provide second-dimensional vibration observation, the cost for deploying additional radars is saved, and the problem of synchronization between the radars is avoided. In addition, the millimeter wave has the characteristics of strong penetrability, no need of illumination and non-invasiveness, so that the device can be applied to outdoor foggy days and factories with weak light. And finally, the millimeter wave radar is low in price, and the designed signal reflector is simple to manufacture, passive and low in cost, so that the system is suitable for large-scale deployment and use.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application belongs to the field of radar measurement technology, and specifically relates to a two-dimensional vibration measurement method based on a single millimeter-wave radar. Background Art

[0002] Two-dimensional vibration measurement is one of the important technologies for structural health monitoring and machine failure analysis. Traditional two-dimensional vibration measurement mainly uses multi-axis accelerometers, laser vibrometers, or high-resolution cameras. Accelerometers need to be fixed on the monitored machine and require additional wiring to provide power and transmit data, which is complex to deploy and expensive. Although laser and camera-based measurement methods can achieve non-contact vibration measurement, they cannot work in environments with poor lighting conditions. Therefore, in recent years, measurement schemes based on radio frequency signals have been proposed. Radio frequency signals can provide non-invasive vibration measurement in environments with poor lighting, but the measurement accuracy of radio frequency identification and ultra-wideband is limited by their signal wavelength. In addition, the above laser, camera, and radio frequency-based solutions usually require multiple devices to provide multi-dimensional vibration measurement of the target, which not only increases the cost, but also brings about the problem of equipment synchronization. In practical applications, the machinery and structures to be monitored may exist in dimly lit factories and outdoor foggy environments, so the measurement equipment needs to provide effective two-dimensional vibration measurement in various environments. Secondly, due to the large demand for structural health monitoring and the wide range of deployment, low-cost and easy-to-deploy sensors should be considered. Finally, since the vibration amplitude of a large number of devices is small, the measurement equipment needs to have high measurement accuracy. The short-wavelength millimeter wave has the characteristics of being independent of light, strong penetration, and high measurement accuracy. Low-cost millimeter wave radar can provide vibration measurement in foggy and poor light environments, and is suitable for large-scale deployment. Summary of the invention

[0003] Based on this, the present invention designs a two-dimensional vibration measurement device and method based on a single millimeter wave radar, which uses low-cost millimeter wave radar equipment to achieve accurate two-dimensional vibration measurement in common environments. Its technical solution is:

[0004] A two-dimensional vibration measurement method based on a single millimeter wave radar comprises the following steps:

[0005] S1. Deploy the designed millimeter wave signal reflector;

[0006] S2. The millimeter wave radar sends a frequency modulated continuous wave. After the signal propagates to the object, it is reflected back to the radar. The transmitted signal and the received signal are mixed to obtain an intermediate frequency signal. The intermediate frequency signal is Fourier transformed and beamformed to calculate the distance and angle of the target in the scene, and to separate the vibration signal.

[0007] S3. Extract vibration from line-of-sight and non-line-of-sight signals;

[0008] S4. Accurately estimate the distance and angle of the target;

[0009] S5. convert the vibration signals in the line-of-sight and non-line-of-sight directions into a set of orthogonal coordinate axes, construct a geometric model using the geometric relationship between the radar, the object, and the reflecting surface, and use the distance angle parameters calculated in S4 to accurately calculate the two-dimensional vibration;

[0010] S6. Perform Fourier transform on the vibration signals on the orthogonal coordinate axes x and y to obtain the vibration frequencies of the target in the x-axis and y-axis directions; perform peak detection on the vibration signals on the orthogonal coordinate axes x and y, and calculate the peak-to-peak value as the vibration displacement on the x-axis and y-axis.

[0011] Preferably, in step S1: the millimeter wave signal reflector includes two corner reflectors, and the angle between the two corner reflectors is θ; a reflecting surface is set near the target to be measured, and the signal reflector is deployed on the target, so that the centerline of one corner reflector faces the radar, and the centerline of the other corner reflector faces the environmental reflecting surface, thereby enhancing the intensity of both the line-of-sight signal and the non-line-of-sight signal; the radar faces the reflecting surface, and the high antenna gain in the direction of the radar centerline is used to further compensate for the severe attenuation of multiple reflections of the non-line-of-sight signal.

[0012] Preferably, in step S2, two objects can be parsed from the radar signal, one of which corresponds to the line-of-sight direction, which is the path for the signal to reach the object directly; and the other corresponds to the non-line-of-sight direction, which is the path for the signal to reach the object after being reflected by the reflecting surface.

[0013] Preferably, in step S3, due to vibration, the complex signal forms an arc in the in-phase orthogonal domain, a circle is fitted using the Taubin method and the Levenberg-Marquardt method, and then the center of the circle is subtracted from the complex signal to remove the static signal component to obtain a purified vibration signal.

[0014] Preferably, in step S4, Chirp-Z transform and multi-signal classification algorithm are performed on the intermediate frequency signal to accurately estimate the distance and arrival angle of the line-of-sight signal and the non-line-of-sight signal to the radar, and the distance angle parameters are sent to the geometric model to calculate the two-dimensional vibration on the orthogonal coordinate axes, and the two-dimensional vibration index is obtained.

[0015] Preferably, in step S5: in addition to the line-of-sight path between the radar and the vibration target, another non-line-of-sight path is constructed using the reflective surface in the environment. This non-line-of-sight path can be regarded as coming from a virtual radar, which is used to provide vibration observation in a second dimension; and a geometric model is established using the geometric relationship between the radar, the object, and the reflective surface, that is,

[0016] T is the vibration target, R pis the location of the radar, W is the reflection surface, C is the reflection point, R v It is a virtual radar constructed by the reflecting surface. Due to the reflection principle, the triangle R is constructed. p TR v , and |CR p |=|CR v |;

[0017] The coordinates of the vibrating target are (x(t), y(t)), which change with time due to vibration; the signal from R p Go directly to T and then return to R p is the line-of-sight signal, half of its length is recorded as r l (t); signal from R p First go to C, then reflect to T and return to R along the same route p is a non-line-of-sight signal, half of its length is recorded as r n (t), by r n1 (t) and r n2 (t) consists of two parts;

[0018] Real Radar R p and Virtual Radar R v The distance between them is denoted as r d , in triangle R p TR v In the above equation, y(t) can be expressed as:

[0019]

[0020] in Represents triangle R p TR v The area of ​​triangle R p TR v The half perimeter of According to the Pythagorean theorem, x(t) can be expressed as:

[0021]

[0022] Preferably, for x(t) and y(t) in (x 0 ,y 0 ) and perform Taylor expansion to obtain the two-dimensional vibration displacements Δx(t) and Δy(t) as follows:

[0023]

[0024] where r l0 and r n0 r l (t) and r n The initial value of (t) is obtained by Fourier transforming the intermediate frequency signal; Δrl (t) and Δr n (t) respectively represent the r caused by target vibration l (t) and r n The change of (t) can be determined by the phase change Δφ between the line-of-sight signal and the non-line-of-sight signal. l (t) and Δφ n (t) is calculated as:

[0025]

[0026] The partial derivatives used to calculate Δx(t) and Δy(t) in the formula are expressed as follows:

[0027]

[0028]

[0029] Among them, y 0 and p 0 Corresponding to the initial values ​​of y(t) and p, that is,

[0030]

[0031] The above formula shows that the four partial derivatives are given by and r d calculated; and Through radar ranging, in triangle R p TR v Applying the cosine theorem, we get r d It is expressed as follows:

[0032]

[0033] Where ∠R p TR v Unknown, therefore, in triangle ∠R p Using the cosine theorem to express ∠R in TC p TR v as follows:

[0034]

[0035] in and is still unknown, but together they constitute Therefore, we can solve the following equations to calculate and

[0036]

[0037] In the above equations, ∠TRp C can be calculated by radar angle measurement.

[0038] Preferably, the millimeter wave signal reflector includes two corner reflectors, and the combined angle θ of the two corner reflectors is 45°; each corner reflector is composed of three mutually perpendicular intersecting surfaces, which can make the incident signal reflect back to its source in the reverse direction after multiple reflections, and the corner reflector can achieve higher signal enhancement within the range of ±30° near its center line.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. The present invention can realize all-weather two-dimensional vibration monitoring of target machines and structures, can provide more comprehensive information than one-dimensional vibration monitoring, and better characterize the health status of machines and structures.

[0041] 2. We use the reflective surfaces in the environment to construct a virtual radar to provide the second dimension of vibration observation, saving the cost of deploying additional radars and avoiding the problem of synchronization between radars.

[0042] 3. The millimeter wave has strong penetrability, no need for illumination, and is non-invasive, so the present invention can be applied outdoors in foggy days and factories with weak light. Finally, since the millimeter wave radar is inexpensive, the designed signal reflector is simple to make, passive, and low-cost, making the present invention suitable for large-scale deployment. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 The figure is a flow chart of two-dimensional vibration measurement based on a single millimeter-wave radar.

[0044] Figure 2 The schematic diagram of two-dimensional vibration measurement based on a single millimeter-wave radar.

[0045] Figure 3 Geometric model for measuring two-dimensional vibrations.

[0046] Figure 4 It is a millimeter wave signal reflector used for two-dimensional vibration measurement. DETAILED DESCRIPTION

[0047] In order to facilitate understanding of the present application, the present application will be described more fully below with reference to the relevant drawings. Embodiments of the present application are provided in the drawings. However, the present application can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present application more thorough and comprehensive.

[0048] Figure 1-4 As shown, the technical steps adopted by the present invention to achieve the above-mentioned purpose are as follows:

[0049] S1: First, deploy the designed millimeter-wave signal reflector. The signal reflector designed by the present invention is composed of two corner reflectors combined at an angle of 45°, and can achieve the strongest reverse reflection near the centerline direction of the two corner reflectors. The corner reflector we use is a retroreflector formed by the intersection of three mutually perpendicular metal surfaces. The incident signal will eventually return to its source in the opposite direction of its incident direction under the continuous reflection of the three surfaces, avoiding the signal from diverging in other directions, thereby increasing the intensity of the reflected signal. Using the wall near the target or an artificially placed iron plate as a reflecting surface, the signal reflector is deployed on the target, so that the centerline of one corner reflector faces the radar, and the centerline of the other corner reflector faces the environmental reflecting surface, thereby enhancing the intensity of the line-of-sight signal and the non-line-of-sight signal at the same time. We let the radar face the reflecting surface, and use the high antenna gain in the centerline direction of the radar to further compensate for the severe attenuation of multiple reflections of the non-line-of-sight signal.

[0050] S2: The millimeter wave radar sends a frequency modulated continuous wave. After the signal propagates to the object, it is reflected back to the radar. The transmitted signal and the received signal are mixed to obtain an intermediate frequency signal. The intermediate frequency signal is transformed by Fourier and beamformed to calculate the distance and angle of the target in the scene, and the vibration signal is separated. According to the scene in S1, two objects can be analyzed from the radar signal, one of which corresponds to the line-of-sight direction, which is the path for the signal to reach the object directly; the other corresponds to the non-line-of-sight direction, which is the path for the signal to reach the object after being reflected by the reflective surface.

[0051] S3: Extract vibration from line-of-sight and non-line-of-sight signals. Due to vibration, the complex signal forms an arc in the in-phase orthogonal domain. The Taubin method and the Levenberg-Marquardt method are used to fit a circle, and then the center of the circle is subtracted from the complex signal to remove the static signal component and obtain the purified vibration signal. After deploying the M-retroreflector on the target, the millimeter wave is sent and received to detect vibration. First, the received millimeter wave signal is range-Fourier transformed and beamformed. Specifically, a Hamming window is applied to each chirp before the range-Fourier transform. For beamforming, we use 8 transmitting and receiving antennas in the horizontal direction of the radar to scan at 1° intervals from -90° to 90°. After range-Fourier transform and beamforming, the two areas with the highest signal intensity in the range-angle heat map correspond to the line-of-sight signal and the non-line-of-sight signal, respectively. Therefore, we detect these two peak range-angle bins and extract their signal phase sequences over time.

[0052] Since the received signal is superimposed with the static components of the surrounding stationary objects, we preprocess the original signal phase to remove the static signal, otherwise the extracted signal phase will be lower than the true value. Specifically, we use the Taubin and Levenberg-Marquardt method to fit the arc formed by the complex signal in the in-phase orthogonal domain, and then subtract the center of the fitted arc from the complex signal to remove the static component.

[0053] S4: Accurately estimate the distance and angle of the target. The range resolution of the range-Fourier transform in S2 is limited by the radar bandwidth, and the angle resolution of the beamforming is limited by the number of radar antennas, resulting in the target position being only a coarse-grained approximate result, which is insufficient to support accurate two-dimensional vibration measurement. Therefore, we further use the Chirp-Z transform and multi-signal classification algorithm to perform fine-grained parameter estimation to obtain the accurate distance and arrival angle of line-of-sight and non-line-of-sight signals.

[0054] S5: In S3, we obtain two one-dimensional vibration signals of the target object in the line-of-sight and non-line-of-sight directions. In order to represent the two-dimensional vibration with vibration signals on a set of orthogonal axes, we established a geometric model, using the geometric relationship between the radar, the object, and the reflecting surface, and using Taylor's formula to convert the vibration signals in the line-of-sight and non-line-of-sight directions to a set of orthogonal coordinate axes (x-axis, y-axis), and use the distance angle parameters calculated in S4 to accurately calculate the two-dimensional vibration.

[0055] S6: Perform Fourier transform on the vibration signals on the orthogonal coordinate axes x and y to obtain the vibration frequencies of the target in the x-axis and y-axis directions. Perform peak detection on the vibration signals on the orthogonal coordinate axes x and y, and calculate the peak-to-peak value as the vibration displacement on the x-axis and y-axis. By combining the vibration displacement at each moment on the x-axis and y-axis, the two-dimensional vibration trajectory of the target is restored.

[0056] Figure 2 The schematic diagram of measuring two-dimensional vibration based on a single millimeter-wave radar. There are many reflective surfaces in the environment, such as walls, furniture, and signs. Therefore, in addition to the line-of-sight path between the radar and the vibrating target, we use the reflective surfaces in the environment to construct another non-line-of-sight path. This non-line-of-sight path can be regarded as coming from a virtual radar to provide vibration observation in the second dimension. By integrating the line-of-sight signal and the non-line-of-sight signal, the two-dimensional vibration of the target can be calculated, that is, the vibration along the orthogonal coordinate axes (x-axis, y-axis).

[0057] Figure 3 is the geometric model diagram for measuring two-dimensional vibration. T is the vibration target, R p is the location of the radar, W is the reflection surface, C is the reflection point, R vIt is a virtual radar constructed by a reflective surface. Due to the reflection principle, there is |CR p |=|CR v |. The coordinates of the vibrating target are (x(t), y(t)) and change with time due to vibration. The signal is from R p Go directly to T and then return to R p is the line-of-sight signal, half of its length is recorded as r l (t); signal from R p First go to C, then reflect to T and return to R along the same route p is a non-line-of-sight signal, half of its length is recorded as r n (t), by r n1 (t) and r n2 (t)It consists of two parts.

[0058] Figure 4 It is a millimeter wave signal reflector for two-dimensional vibration measurement. It is composed of two corner reflectors with a combined angle θ of 45°. Each corner reflector is composed of three mutually perpendicular intersecting surfaces, which can make the incident signal reflect back to its source along its reverse direction after multiple reflections, avoiding the signal from scattering in other directions. The corner reflector achieves high signal enhancement within the range of ±30° near its center line (dashed line in the figure). Because the top view of the designed millimeter wave signal reflector looks like a capital letter "M", it is also called M-retroreflector.

[0059] like Figure 4 As shown, we connect two corner reflectors, each of which supports the enhancement of line-of-sight signals and non-line-of-sight signals. The two corner reflectors can be connected at different angles, i.e. Figure 4 Specifically, the corner reflector achieves higher signal enhancement within the range of ±30° near its center line, such as Figure 4 As shown by the dashed line in the middle. If θ = 0°, the incident non-line-of-sight signal is required to be perpendicular to the reflecting surface, which is not feasible due to the scene arrangement. If θ = 90°, both corner reflectors face the real radar, so the non-line-of-sight signal cannot be enhanced. Considering the above two boundary cases, we set θ to their average value, i.e. 45°, which allows the first and second corner reflectors to effectively reverse-reflect line-of-sight signals and non-line-of-sight signals from different directions.

[0060] Millimeter wave vibration sensing principle:

[0061] The present invention uses the ability of millimeter wave radar to sense tiny displacements to measure the vibration of the target. The tiny displacement of the target will be reflected in the phase change of the intermediate frequency signal. In more detail, the transmitting antenna of the millimeter wave radar sends a frequency modulated continuous signal wave. A chirp can be expressed as Its frequency f(t) changes from f0 It starts to increase linearly with the slope k over time, that is, f(t) = f 0 +kt. After the signal propagates to the vibrating target, it is reflected back to the receiving antenna of the millimeter-wave radar. The time delay of the received signal relative to the transmitted signal is Where c is the speed of light, d(t) represents the distance between the radar and the target, so the received signal can be expressed as s r (t) = αs t (t-τ), where α represents the signal attenuation coefficient. The received signal and the transmitted signal are then mixed by a mixer to obtain an intermediate frequency signal. Where * represents a conjugate operation. Since the frequency of the intermediate frequency signal is proportional to the distance between the radar and the target, the distance of the target relative to the radar can be calculated by performing Fourier transform on the intermediate frequency signal. Using multiple antennas on the millimeter wave radar, the angle of the target relative to the radar can be calculated through beamforming technology, thereby achieving target positioning. Subsequently, the signal phase of the distance-angle bin where the target is located in each chirp is extracted, that is, Where λ is the signal wavelength, d 0 is the initial distance from the radar to the target, and the displacement of the target relative to the initial position at each moment is obtained. Since the wavelength of millimeter waves is short, their phase is more sensitive to tiny vibrations, so a clear vibration pattern can be obtained from them.

[0062] The principle of a single millimeter-wave radar sensing two-dimensional vibration:

[0063] The above method can actually only obtain one-dimensional vibration of the target. Specifically, the displacement of the vibrating target along the line-of-sight direction will result in a larger Δd(t), while the equal displacement of the vibrating target in other directions, especially perpendicular to the line-of-sight direction, will only cause a very small Δd(t), which will then appear as a very small phase change in the intermediate frequency signal, making it difficult to detect. Therefore, a simple solution is to deploy two radars to monitor the vibration of the target in two dimensions. However, this method increases the system cost and requires strict synchronization between radars, further increasing the complexity of deployment. Therefore, the present invention utilizes reflectors around the target (e.g., walls, metal plates) to construct a virtual radar, such as Figure 2 As shown, the virtual radar actually behaves as a non-line-of-sight path of the radar signal. In this way, we can simultaneously obtain observations of the target vibration in two dimensions from the line-of-sight path and non-line-of-sight path of a single radar, reducing equipment costs and deployment difficulty.

[0064] 2D vibration measurement model:

[0065] The geometric model used for 2D vibration measurement is as follows: Figure 3 The real-time coordinates of the vibrating target T are (x(t), y(t)), which change with time due to vibration and can be further expressed as x(t) = x0 +Δx(t),y(t)=y 0 +Δy(t), where (x 0 ,y 0 ) is the initial position coordinate of the vibration target, Δx(t) and Δy(t) are its displacement changes relative to the initial position, which is the two-dimensional vibration we want to calculate.

[0066] Since the signal phases of the line-of-sight path and the non-line-of-sight path obtained above are not directly related to the coordinates of T, we further use the path lengths of the line-of-sight signal and the non-line-of-sight signal to represent x(t) and y(t). Figure 2 As shown, due to the reflection principle, there is |CR p |=|CR v |, and then the line-of-sight path and half of the non-line-of-sight path of the radar signal propagation are denoted as |TR p |,|TR v |, further expressed as r l (t) and r n (t). At the same time, the real radar R p and Virtual Radar R v The distance between them is denoted as r d , is a constant after the scene is determined. p TR v In the above equation, y(t) can be expressed as:

[0067]

[0068] in Represents triangle R p TR v The area of ​​triangle R p TR v The half perimeter of

[0069]

[0070] According to the Pythagorean theorem, x(t) can be expressed as:

[0071]

[0072] Next, for x(t) and y(t) in (x 0 ,y 0 ) and perform Taylor expansion to obtain the desired two-dimensional vibration displacements Δx(t) and Δy(t) as follows:

[0073]

[0074] where r l0 and r n0 rl (t) and r n The initial value of (t) can be obtained by Fourier transforming the intermediate frequency signal. l (t) and Δr n (t) respectively represent the r caused by target vibration l (t) and r n The change of (t) can be determined by the phase change Δφ between the line-of-sight signal and the non-line-of-sight signal. l (t) and Δφ n (t) is calculated as:

[0075]

[0076] The partial derivatives used to calculate Δx(t) and Δy(t) in the formula are expressed as follows:

[0077]

[0078]

[0079] Among them, y 0 and p 0 Corresponding to the initial values ​​of y(t) and p, that is,

[0080]

[0081] The above formula shows that the four partial derivatives are given by and r d Calculated. and Obtained by radar ranging, however, r d cannot be directly measured from the Fourier transform result, so we further calculate the p TR v Applying the cosine theorem, we get r d It is expressed as follows:

[0082]

[0083] Where ∠R p TR v Unknown, therefore, in triangle ∠R p Using the cosine theorem to express ∠R in TC p TR v as follows:

[0084]

[0085] in and is still unknown, but together they constitute Therefore, we can solve the following equations to calculate and

[0086]

[0087] In the above equations, ∠TR p C can be calculated by radar angle measurement.

[0088] Distance parameter estimation:

[0089] Range-Fourier transform is a classic method used by FMCW radar for distance measurement, but its range resolution is limited by the signal bandwidth, which is essentially due to the limited spectrum resolution. The Fourier transform can be regarded as the z-transform of a discrete time series, which is uniformly sampled on a unit circle with a fixed frequency interval. In FMCW signals, the signal bandwidth is determined by the frequency change slope k and the chirp duration. For commercial radars with a bandwidth of 1-4 GHz, the range resolution of 3.75-15 cm is not enough to achieve accurate vibration measurement.

[0090] In order to achieve fine-grained distance estimation, we use Chirp-Z transform. The Fourier transform is limited to the unit circle, while the Chirp-Z transform allows sampling of a spiral line, thereby refining the local spectrum. Specifically, the sampling point of the Chirp-Z transform is represented by z k =AW -k (k=0,…,M-1), where A is a complex constant representing the starting point of the spiral, and M is the number of sampling points. Controls the change of the spiral line, where W 0 represents the stretch rate of the helix, Determines the sampling resolution of the Chirp-Z transform, so we choose a smaller However, using a small This will increase the amount of calculation. Therefore, in the implementation, we select only part of the spiral arc for sampling based on the coarse-grained distance of the distance-Fourier transform, and then set a smaller This corresponds to a distance interval of 5 mm, thus achieving millimeter-level resolution in distance measurement.

[0091] Angle parameter estimation:

[0092] Due to the limited number of antennas on the radar, the angular resolution of traditional beamforming is only 10°-15°. This is because traditional beamforming maintains the time resolution of the original signal by sacrificing the angular resolution. p C Make an accurate measurement to perfect Figure 3The geometric model shown, so time resolution is not necessary for us.

[0093] Therefore, we use a multi-signal classification algorithm to achieve higher accuracy in angle estimation. The multi-signal classification algorithm exploits the characteristic structure of the covariance matrix of the received signal over a period of time and uses the noise information to provide a more accurate estimate. In other words, the algorithm uses longer signal measurements and prior knowledge of the number of targets to obtain finer angular resolution.

[0094] Specifically, we use one transmitting antenna and four receiving antennas on the radar to calculate the angle. First, we extract the multi-antenna signal from the two range bins detected by the range-Fourier transform. Then, we obtain the multi-antenna signal covariance matrix and its eigenvalues ​​of the line-of-sight signal and the non-line-of-sight signal respectively. The largest eigenvalues ​​correspond to the line-of-sight signal and the non-line-of-sight signal of the target, and the rest correspond to the noise that constitutes the noise matrix U. Next, we scan the angle ψ from -90° to 90° at 0.1° intervals and use Calculate the spectral function, where D represents the steering vector, and then find the angle with the maximum power in P(ψ) as a fine-grained estimate of the target arrival angle. l , n After that, we can get ∠TR p C=|ψ n -ψ l |.

[0095] 2D vibration index estimation:

[0096] After signal preprocessing, we get the phase sequence of line-of-sight and non-line-of-sight signals over time. Then we subtract the phase of the first timestamp from them to get the relative phase change Δφ of the line-of-sight and non-line-of-sight signals. l (t) and Δφ n (t), and then use Calculate the length change Δr of the line-of-sight and non-line-of-sight paths l (t) and Δr n (t), finally, the real-time two-dimensional vibration Δx(t) and Δy(t) of the target along the x-axis and y-axis are calculated using the geometric model of the measured two-dimensional vibration and the estimated fine-grained parameters.

[0097] Next, we calculate the vibration frequency and peak-to-peak displacement of the target on the x-axis and y-axis over a period of time. We perform Fourier transform on the vibration displacement sequence Δx(t) and Δy(t) and obtain the vibration frequency by searching for the maximum value in the spectrum. In order to extract the peak-to-peak displacement, we first perform peak detection on the displacement change sequence to detect the peaks and troughs, and set the peak threshold and minimum peak interval to avoid the influence of signal jitter on peak detection. Finally, the average value of the difference between the peak and trough displacements during the period is calculated as the peak-to-peak displacement during the period.

[0098] In addition to obtaining the vibration frequency and vibration displacement in orthogonal axes, we can also recover the two-dimensional vibration trajectory of the target by combining the real-time measurements of the displacement changes Δx(t) and Δy(t).

[0099] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A two-dimensional vibration measurement method based on a single millimeter wave radar, characterized in that: The steps include: S1. Deploy the designed millimeter wave signal reflector; S2. The millimeter wave radar sends a frequency modulated continuous wave. After the signal propagates to the object, it is reflected back to the radar. The transmitted signal and the received signal are mixed to obtain an intermediate frequency signal. The intermediate frequency signal is Fourier transformed and beamformed to calculate the distance and angle of the target in the scene, and to separate the vibration signal. S3. Extract vibration from line-of-sight and non-line-of-sight signals; S4. Accurately estimate the distance and angle of the target; S5. convert the vibration signals in the line-of-sight and non-line-of-sight directions into a set of orthogonal coordinate axes, construct a geometric model using the geometric relationship between the radar, the object, and the reflecting surface, and use the distance angle parameters calculated in S4 to accurately calculate the two-dimensional vibration; S6. Perform Fourier transform on the vibration signals on the orthogonal coordinate axes x and y to obtain the vibration frequencies of the target in the x-axis and y-axis directions; perform peak detection on the vibration signals on the orthogonal coordinate axes x and y, and calculate the peak-to-peak value as the vibration displacement on the x-axis and y-axis.

2. A two-dimensional vibration measurement method based on a single millimeter wave radar according to claim 1, characterized in that: In step S1: the millimeter wave signal reflector includes two corner reflectors, and the angle between the two corner reflectors is θ; a reflecting surface is set near the target to be measured, and the signal reflector is deployed on the target, so that the center line of one corner reflector faces the radar, and the center line of the other corner reflector faces the environmental reflecting surface, thereby enhancing the strength of the line-of-sight signal and the non-line-of-sight signal at the same time; The radar faces the reflecting surface and uses the high antenna gain in the direction of the radar centerline to further compensate for the severe attenuation of multiple reflections of the non-line-of-sight signal.

3. The two-dimensional vibration measurement method based on a single millimeter wave radar according to claim 1, characterized in that: In step S2, two objects can be parsed from the radar signal, one of which corresponds to the line-of-sight direction, which is the path for the signal to reach the object directly; the other corresponds to the non-line-of-sight direction, which is the path for the signal to reach the object after being reflected by the reflecting surface.

4. The two-dimensional vibration measurement method based on a single millimeter wave radar according to claim 1, characterized in that: In step S3, due to vibration, the complex signal forms an arc in the in-phase orthogonal domain, and a circle is fitted using the Taubin method and the Levenberg-Marquardt method. Then, the center of the circle is subtracted from the complex signal to remove the static signal component, thereby obtaining a purified vibration signal.

5. The two-dimensional vibration measurement method based on a single millimeter wave radar according to claim 1, characterized in that: In step S4, Chirp-Z transform and multi-signal classification algorithm are performed on the intermediate frequency signal to accurately estimate the distance and arrival angle of the line-of-sight signal and the non-line-of-sight signal to the radar, and the distance angle parameters are sent to the geometric model to calculate the two-dimensional vibration on the orthogonal coordinate axes, and the two-dimensional vibration index is obtained.

6. The two-dimensional vibration measurement method based on a single millimeter wave radar according to claim 1, characterized in that: In step S5: in addition to the line-of-sight path between the radar and the vibration target, another non-line-of-sight path is constructed using the reflective surface in the environment. This non-line-of-sight path can be regarded as coming from a virtual radar and is used to provide vibration observation in a second dimension. A geometric model is established using the geometric relationship between the radar, the object, and the reflective surface, that is, T is the vibration target, R p is the location of the radar, W is the reflection surface, C is the reflection point, R v It is a virtual radar constructed by the reflecting surface. Due to the reflection principle, the triangle R is constructed. p TR v , and |CR p |=|CR v |; The coordinates of the vibrating target are (x(t), y(t)), which change with time due to vibration; the signal from R p Go directly to T and then return to R p is the line-of-sight signal, half of its length is recorded as r l (t); signal from R p First go to C, then reflect to T and return to R along the same route p is a non-line-of-sight signal, half of its length is recorded as r n (t), by r n1 (t) and r n2 (t) consists of two parts; Real Radar R p and Virtual Radar R v The distance between them is denoted as r d , in triangle R p TR v In the above equation, y(t) can be expressed as: in Represents triangle R p TR v The area of ​​triangle R p TR v The half perimeter of According to the Pythagorean theorem, x(t) can be expressed as:

7. A two-dimensional vibration measurement method based on a single millimeter wave radar according to claim 6, characterized in that: Taylor expansion is performed on x(t) and y(t) at (x0, y0), and the two-dimensional vibration displacements Δx(t) and Δy(t) are obtained as follows: where r l0 and r n0 r l (t) and r n The initial value of (t) is obtained by Fourier transforming the intermediate frequency signal; Δr l (t) and Δr n (t) respectively represent the r caused by target vibration l (t) and r n The change of (t) can be determined by the phase change Δφ between the line-of-sight signal and the non-line-of-sight signal. l (t) and Δφ n (t) is calculated as: The partial derivatives used to calculate Δx(t) and Δy(t) in the formula are expressed as follows: Among them, y0 and p0 correspond to the initial values ​​of y(t) and p, respectively, that is, The above formula shows that the four partial derivatives are given by and r d calculated; and Through radar ranging, in triangle R p TR v Applying the cosine theorem, we get r d It is expressed as follows: Where ∠R p TR v Unknown, therefore, in triangle ∠R p Using the cosine theorem to express ∠R in TC p TR v as follows: in and is still unknown, but together they constitute Therefore, we can solve the following equations to calculate and In the above equations, ∠TR p C can be calculated by radar angle measurement.

8. The two-dimensional vibration measurement method based on a single millimeter wave radar according to claim 6, characterized in that: The millimeter wave signal reflector includes two corner reflectors, and the combined angle θ of the two corner reflectors is 45°; each corner reflector is composed of three mutually perpendicular intersecting surfaces, which can make the incident signal reflect back to its source along its reverse direction after multiple reflections, and the corner reflector achieves high signal enhancement within the range of ±30° near its center line.

Citation Information

Cited By

  • Micro-vibration real-time measurement method based on mobile phone sound signals

    CN119689488A

  • Anti-vibration interference displacement measurement method and device based on millimeter wave radar structure

    CN121541160A