A method and system for detecting termites based on microwave radar penetrating walls

Abstract

The present invention discloses a method and system for detecting termites based on microwave radar through walls, and belongs to the technical field of microwave radar through-wall detection. The detection method of the present invention comprises: using a multi-output microwave radar detection device to scan the area to be detected to obtain a radar echo signal, performing signal reception processing on the signal to obtain a discrete time domain intermediate frequency signal; bandpass filtering to remove background clutter; signal reconstruction to achieve DOA estimation of the source; amplitude screening of termite targets and measuring their target distance; and sending termite detection results including target distance and DOA estimation to the target user. The detection system of the present invention comprises a microwave radar detection device and a user terminal, wherein the microwave radar detection device is used to scan the area to be detected and perform signal processing and analysis on the radar echo signal to obtain the termite detection result, and send it to the user terminal. The present invention adopts a non-invasive detection method to realize fast, convenient and low-cost accurate termite self-detection, so as to significantly reduce the harmfulness of termite disasters.

Description

Technical Field

[0001] The present invention relates to the technical field of microwave radar through-wall detection, and in particular to a termite detection method and system based on microwave radar through-wall detection. Background Art

[0002] Termites, as a pest hidden underground and in wood, have become a serious threat worldwide with their amazing reproductive ability and destructiveness. Termites can not only erode the structure of houses and commercial buildings, resulting in high repair costs, and even in severe cases, put the building at risk of collapse, but also seriously affect the integrity of wooden furniture and decorations, causing inestimable financial losses. Traditional prevention and control methods require waiting until obvious mud ant paths appear on the surface of the wall, or have caused significant damage to the wall, and there is obvious population gathering before carrying out pest control work. At this time, the cost of pest control work is high, the scale is large, and the negative impact is more significant. Other termite control technologies that have appeared on the market in recent years include characteristic gas analysis, acoustic wave detection, radar detection, etc. Among them, the detection threshold of characteristic gas analysis is high, and it is difficult to make effective detection when the termite population is not large. Acoustic wave detection uses ultrasound to detect the sound waves generated by the activities of termites. It is also difficult to effectively detect when the termite population is small, and it is greatly disturbed by noise and has poor penetration into cement materials.

[0003] At present, there are devices that use the relatively primitive single-transmitter and single-receiver microwave frequency modulated continuous wave (FMCW) radar technology for termite detection. This type of equipment can complete certain detection tasks, but its wall penetration performance is poor and the detection accuracy is low, resulting in limited detection capabilities for termites inside the wall and low data value. At the same time, due to the backward detection algorithm, the detection effect of this type of equipment is not good, so it has not been commercialized. In summary, traditional termite prevention and control methods are inefficient and have high subsequent disinfection costs, while existing termite detection methods have high environmental requirements and poor detection effects, and have not been widely used. Therefore, a low-cost, easy-to-implement, and high-recognition accuracy termite detection method has broad application prospects. Summary of the invention

[0004] The present invention provides a termite detection method and system based on microwave radar through walls, aiming to provide an accurate termite detection solution that is easy to operate and can be used by users themselves, and is used to accurately locate and remove termite nests and ant trails when they are initially formed, so as to reduce time costs and building structure losses, and significantly reduce the harmfulness of termite disasters.

[0005] In one aspect, the present invention provides a method for detecting termites based on microwave radar through walls, the method comprising the following steps:

[0006] Step 1, using a multi-input multi-output microwave radar detection device to scan the area to be detected, obtain radar echo signals and perform receiving signal processing on them to obtain the original discrete time domain intermediate frequency signal; wherein the antenna array of the microwave radar detection device is a uniform array integrating transmission and reception;

[0007] Step 2, bandpass filtering the original discrete time domain intermediate frequency signal, and then performing discrete Fourier transform to obtain a frequency domain intermediate frequency signal containing target orientation information;

[0008] Step 3, performing signal reconstruction processing on the frequency domain intermediate frequency signal to obtain the estimated results of the arrival directions of K information sources; wherein the number of information sources K is a preset value, and each information source is used to represent a candidate detection target of termites;

[0009] And based on the arrival direction estimation result, the signal amplitude corresponding to each information source is estimated to obtain the frequency domain signal of each information source;

[0010] Step 4, taking the signal source whose signal amplitude is greater than or equal to the preset amplitude threshold as the termite target, obtaining the target distance of each termite target, and obtaining the termite detection result of the to-be-detected area based on the target distance of the termite target and the corresponding arrival direction estimation result; wherein the target distance is used to characterize the distance between the signal source corresponding to the termite target and the reference array element of the antenna array of the microwave radar detection device;

[0011] Step 5: Send the termite detection results of the area to be detected to the target user for visual display.

[0012] The present invention utilizes a multi-input multi-output (MIMO) microwave radar detection device to detect a target object (such as a wall), and then combines a high-resolution radar signal processing algorithm to analyze and present the infestation of termites in the wall, thereby providing a more efficient and intelligent solution for building protection and pest management.

[0013] Furthermore, in step 2, a bandpass filter is performed with a passband of 0.1-10 Hz to effectively remove background clutter, increase the signal ratio of the changing termite signal, and thus improve the detection performance.

[0014] Furthermore, step 3 specifically includes the following steps:

[0015] The original signal matrix is ​​constructed and matrix enhancement is performed on the frequency domain intermediate frequency signal to construct an augmented signal matrix ;

[0016] For the augmented signal matrix Perform singular value decomposition and decompose the left singular matrix The first K column vectors of ;

[0017] The matrix triangular decomposition method is used to calculate the augmented signal matrix Perform triangular decomposition and based on the left singular matrix The left matrix (i.e., the lower triangular matrix) obtained by triangular decomposition is split to obtain the Vandermonde matrix containing the termite signal vector ; That is, based on the left singular matrix The lower triangular matrix is ​​extracted by the extraction method, and the Vandermonde matrix containing the termite signal vector is obtained based on the first K columns of the lower triangular matrix. ;

[0018] By cutting and splicing the signal subspace Combined with the Vandermonde matrix Construct a matrix bundle, and obtain two sets of spatial frequency information sequences containing azimuth angles and elevation angles based on the generalized eigenvalues ​​of the matrix bundle, wherein the lengths of the two sets of spatial frequency information sequences are both K;

[0019] Based on orthogonality, two sets of spatial frequency information sequences are paired to obtain K pairs of spatial frequency pairs in two mutually orthogonal directions relative to the antenna array. The arrival azimuth is estimated based on the matched spatial frequency pairs to obtain the arrival direction estimation result of each target.

[0020] Furthermore, the estimation of the arrival azimuth angle according to the matched spatial frequency pair is specifically as follows:

[0021] definition and To match the two spatial frequencies in the spatial frequency pair, , They are the antenna arrays corresponding to , The array element spacing in the direction, the subscript k is the source index;

[0022] Estimate the elevation and azimuth of source k:

[0023] , ;

[0024] in, , are the elevation angle and azimuth angle of the signal source k, respectively, and the auxiliary parameters , The expressions are: , , represents the imaginary unit, is the natural logarithm.

[0025] In the present invention, the time correlation and autocorrelation between radar echo signals are captured by matrix enhancement processing, and the signal feature extraction of termite detection signals (radar echo signals) is realized, which is more conducive to the detection and analysis of termites in the detection area to be detected; the matrix enhancement processing can increase the signal dimension while maintaining low computational complexity while improving the signal detection resolution and anti-noise interference capability (such as background wall clutter interference), further ensuring that the detection method of the present invention can realize the recognition and separation of termite signals; and singular value decomposition is used to distinguish mutually orthogonal signal subspaces and noise subspaces to achieve the purpose of optimizing detection capabilities; the constructed Vandermonde matrix combines the two-dimensional spatial frequency with the augmented signal matrix The left matrix obtained by the triangular decomposition is combined with the constructed matrix bundle to further realize the generalized eigenvalue decomposition, which improves the extraction efficiency of the corresponding frequency of the extracted subspace.

[0026] Furthermore, the original signal matrix is ​​constructed in the following manner: the original signal matrix is ​​constructed according to the relative position relationship between each array element of the antenna array and the reference array element.

[0027] Further, the reference array element of the antenna array is set to the first array element at the upper left corner of the antenna array.

[0028] Furthermore, in step 3, the number of information sources K is determined as follows:

[0029] For the augmented signal matrix The singular values ​​obtained by singular value decomposition are squared and arranged in descending order, and then the cumulative variance explanation rate is calculated. When the cumulative variance explanation rate exceeds 90%, it is truncated, and the value of the number of information sources K is determined based on the number of singular values ​​included at the time of truncation.

[0030] Furthermore, in step 3, the two groups of spatial frequency information sequences are paired based on orthogonality as follows: the two groups of spatial frequency information sequences are allocated into K pairs, the two spatial frequencies of each spatial frequency pair come from different spatial frequency information sequences, the vectors corresponding to each spatial frequency pair are multiplied and accumulated to obtain the return value of the objective function, and the final matching result is obtained based on the matching method corresponding to the minimum objective function.

[0031] Furthermore, in step 4, the target distance of the termite target is calculated as follows:

[0032] According to the formula Calculate the phase difference between the adjacent frequency scales of the signal source corresponding to the termite target p ,in is the frequency step size of Fourier transform, is the propagation delay of the signal source corresponding to the termite target p relative to the transmission path and echo path of the antenna array;

[0033] According to the formula The target distance of the termite target p is calculated, where c is the speed of light constant, is the imaginary part extraction operator.

[0034] On the other hand, the present invention also provides a termite detection system based on microwave radar through walls, which includes a user terminal and a multi-input multi-output microwave radar detection device;

[0035] The antenna array of the microwave radar detection device is a uniform array integrating transmission and reception; data is transmitted between the user terminal and the microwave radar detection device by wireless or wired means; the user terminal configures the working parameters of the microwave radar detection device;

[0036] The microwave radar detection device is used to scan the area to be detected and receive the radar echo signal, and perform received signal processing on the radar echo signal to obtain the original discrete time domain intermediate frequency signal; and is used to perform steps 2 to 4 of the detection method of the present invention to realize signal extraction processing on the original discrete time domain intermediate frequency signal to obtain the termite detection result of the area to be detected;

[0037] The microwave radar detection device transmits the termite detection results to the user terminal in real time, and the termite detection results are visualized on the user terminal.

[0038] In another aspect, the present invention provides a microwave radar wall-penetrating termite detection system comprising: a multi-input multi-output microwave radar detection device, an edge computing device for performing signal extraction processing, and a user terminal;

[0039] The microwave radar detection device is connected to the user terminal through the edge computing device, and the connection mode between the edge computing device and the microwave radar detection device is a wired mode or a wireless mode; and the connection mode between the edge computing device and the user terminal is a wired mode or a wireless mode;

[0040] Among them, the antenna array of the microwave radar detection device is a uniform array of transmission and reception, and the user terminal sends configuration instructions of working parameters to the microwave radar detection device through the edge computing device;

[0041] The microwave radar detection device is used to scan the area to be detected and receive the radar echo signal, and perform received signal processing on the radar echo signal to obtain the original discrete time domain intermediate frequency signal and send it to the edge computing device;

[0042] The edge computing device is used to execute steps 2 to 4 of the detection method of the present invention to realize signal extraction and processing of the original discrete time domain intermediate frequency signal to obtain the termite detection result of the area to be detected;

[0043] The edge computing device transmits the termite detection results of the area to be detected to the user terminal in real time, or transmits the termite detection results of the area to be detected specified by the request message initiated by the user terminal to the user terminal, so as to realize the visual display of the transmitted termite detection results on the user terminal.

[0044] In summary, the technical solution provided by the present invention brings at least the following beneficial effects:

[0045] The present invention realizes a complete set of radar detection and signal extraction processing schemes for termite detection based on the data characteristics of radar echo signals. The present invention can accurately extract weak termite signals from background noise, effectively facilitating the development of non-invasive termite detection systems; the present invention can help fields that require termite detection, such as construction and agriculture, to conduct fast, convenient and low-cost termite self-tests, saving the expensive costs of professional termite detection agencies. At the same time, it also effectively avoids the irreversible damage to the building itself caused by large-scale invasive detection and treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0047] Figure 1 A schematic flow chart of a method for detecting termites based on microwave radar penetrating walls provided in an embodiment of the present invention.

[0048] Figure 2 A schematic structural diagram of a microwave radar-through-wall termite detection system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] To make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be described in detail and completely in conjunction with the drawings in the implementation of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings can be arranged and designed using different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the present application for which protection is claimed, but merely represents selected embodiments of the present invention.

[0050] The embodiment of the present invention provides a method for detecting termites based on microwave radar through walls. Figure 1 As shown, the method comprises the following processing steps:

[0051] Step S1, scanning the area to be detected to obtain radar echo signals, and receiving and processing the signals to obtain discrete time domain intermediate frequency signals.

[0052] A multi-input multi-output microwave radar detection device is used to scan the area to be detected, obtain radar echo signals and perform received signal processing on them to obtain the original discrete time domain intermediate frequency signal; wherein the antenna array (transmitter and receiver in one) of the microwave radar detection device is a uniform array;

[0053] Step S2: bandpass filtering to remove background clutter.

[0054] The original discrete time-domain intermediate frequency signal is bandpass filtered to filter out background clutter, thereby reducing interference with termite signals and improving the performance of later signal processing. At the same time, the signal obtained after bandpass filtering is discrete Fourier transformed to obtain a frequency domain signal in the form of a complex exponential, that is, a frequency domain intermediate frequency signal containing target azimuth information.

[0055] Step S3: Signal reconstruction to achieve DOA estimation of the information source.

[0056] The frequency domain intermediate frequency signal is reconstructed to obtain the estimated results of the direction of arrival (DOA) of K information sources; the number of information sources K is a preset value, and each information source is used to represent a candidate detection target of termites; and the signal amplitude corresponding to each information source is estimated based on the arrival direction estimation result to obtain the frequency domain signal of each information source;

[0057] Step S4, the termite target is selected by amplitude and its target distance is calculated.

[0058] For information sources whose signal amplitude is greater than or equal to the amplitude threshold, distance measurement is performed based on the frequency domain signal of the information source to obtain the distance estimation value between each information source and the reference array element of the antenna array of the microwave radar detection device; the termite detection result of the area to be detected is obtained based on the distance estimation value of the information source and the corresponding arrival direction estimation result;

[0059] Step S5: Send the termite detection result to the target user.

[0060] In this step, the termite detection results of the area to be detected can be directly transmitted to the target user in real time for visual display; or based on the detection result request message initiated by the user, the termite detection results of the area to be detected specified by the user can be sent to the target user for visual display.

[0061] In one embodiment, step S3 specifically includes the following steps:

[0062] Step S301, for the frequency domain signal obtained after filtering, the original signal matrix is ​​constructed and matrix enhancement processing (ME, Matrix Enhancement) is performed to construct an augmented matrix. The purpose of matrix enhancement is to expand the rank of the matrix and increase the dimension of the signal to enhance the time correlation of the radar data in the matrix and improve the processing efficiency. In this embodiment, the Hankle matrix and the sub-matrix are concatenated to construct an augmented matrix for subsequent processing.

[0063] Step S302, performing singular value decomposition (SVD) on the augmented matrix, extracting singular values ​​of the augmented matrix, performing dimension reduction and decomposition processing on the augmented matrix, and distinguishing the signal subspace and the noise subspace through the singular values.

[0064] Step S303, using a matrix triangular decomposition method to process the augmented matrix to construct a Vandermonde matrix, which greatly reduces the difficulty of subsequent processing and the complexity of calculation.

[0065] Step S304, constructing a matrix pencil by cutting and splicing the signal subspace and combining the obtained Vandermonde matrix and solving the generalized eigenvalues, thereby obtaining two sets of spatial frequency information sequences containing azimuth angles and elevation angles, that is, obtaining a series of generalized eigenvalues ​​containing two-dimensional spatial frequency information through matrix deformation operation and derivation.

[0066] Step S305, the frequency information obtained by solving the generalized eigenvalue is paired through a matching algorithm, and the arrival azimuth is estimated based on the matched frequency pairs, and then the arrival direction (DOA) is estimated based on the relationship between the spatial frequency and the arrival direction, and the arrival direction estimation result of each target is obtained. When matching the frequency pairs, for the same source, if the two spatial frequencies match, the vectors corresponding to the spatial frequencies should be orthogonal (two different directions), so the multiplication result should be 0 in theory. In this embodiment, for the two sets of frequency information sequences obtained, the two sets of spatial frequency pairs corresponding to the vector multiplication are traversed based on the set objective function, that is, the two sets of spatial frequency information sequences are allocated into K pairs, and the vectors corresponding to each spatial frequency pair are multiplied and accumulated to obtain the return value of the objective function, and the final matching result is obtained based on the matching method corresponding to the minimum objective function.

[0067] In the embodiment of the present invention, the mathematical model of the frequency domain intermediate frequency signal containing the target position information can be expressed as:

[0068] (1)

[0069] in, The antenna array Line The frequency domain intermediate frequency signal corresponding to the antenna in the column, the row index of the antenna array , column index , M and N are the number of row and column elements of the antenna array respectively. represents the source index, i.e. , represents the signal amplitude, represents the imaginary unit, Indicates the source The azimuth of and Source Relative to the spatial frequency in two mutually orthogonal directions of the antenna array, it characterizes the phase change characteristics and also contains the directional arrival angle information (elevation angle and azimuth angle), which is the parameter to be estimated.

[0070] Signal reconstruction is based on the sampling characteristics of FMCW radar. The intermediate frequency signal sampled in a snapshot is spliced ​​into a The first element is a reference element. Of course, other elements in the array may also be used as reference elements, which is not limited in the embodiment of the present invention. The initial signal matrix corresponding to the antenna array obtained by signal reconstruction is in the form of:

[0071] (2)

[0072] Matrix enhancement is to process the initial signal matrix through mathematical methods, expand the signal dimension, and increase time redundancy. This embodiment uses the construction of Hankel matrix combined with matrix splicing for enhancement. The property that the diagonal elements of Hankel matrix are equal makes the subsequent singular value decomposition easier. First, extract the dimension of the initial matrix X The submatrix of is used as the expansion matrix of each element of the initial matrix, recorded as: (3)

[0073] Where i and j are the initial values ​​of the row and column indices of the matrix elements, respectively.

[0074] The matrix Further, as the elements of the newly added matrix, the row and column indexes of the matrix elements are The new wide matrix is ​​in the following form:

[0075] (4)

[0076] in, To augment the signal matrix, the sum of K and L should satisfy the number of sampling points (i.e., the number of frequency domain intermediate frequency signals containing target position information) minus one. In this embodiment, K and L are both taken as half of N based on experience. K and L should be the same signal dimension to ensure that the augmented matrix has the same rank as the initial signal matrix.

[0077] Singular value decomposition is a common method for extracting the principal components of a signal, and is usually used for non-square matrices. For square matrices, an alternative method is eigenvalue decomposition, which is used to extract the significant components in the signal, usually corresponding to the signal source. The significant eigenvalues ​​correspond to the corresponding number of target signal sources, and the remaining eigenvalues ​​represent noise. For non-square matrices, since there is no concept of eigenvalues, singular value decomposition is used, and the decomposition process can be expressed as:

[0078] (5)

[0079] Among them, A is the matrix to be decomposed, assuming its dimension is , orthogonal matrix Each column is The singular vectors of are , orthogonal matrix Each column is The singular vectors of are , is a diagonal matrix containing singular values, including or The eigenvalue square root of is . After obtaining the singular value matrix, in order to distinguish the noise and signal subspaces, the singular values ​​are squared and arranged in descending order, and the cumulative variance explanation rate is calculated. The threshold is set to 90%. When the cumulative variance explanation rate exceeds 90%, it is truncated, that is, the signal subspace and the noise subspace are distinguished. Specifically, if there are r singular values ​​corresponding to the signal subspace, and a total of N singular values, then the signal subspace consists of the first r columns of the left singular matrix, and the noise subspace consists of the first (Nr) columns of the left singular matrix. The right singular matrix is ​​the inversion of the left singular matrix. The variance explanation rate comes from the concept of principal component analysis (PCA). The covariance matrix is ​​constructed through the original data matrix, and then the eigenvalue decomposition is performed. The eigenvalue obtained represents the variance explanation rate of each eigenvector corresponding to the direction. Since the singular values ​​obtained by SVD and the eigenvalues ​​obtained by PCA have a relationship that the latter is equal to the square of the former, SVD can also use this method to divide the signal subspace and the noise subspace. This embodiment obtains a left singular matrix, a singular value matrix and a right singular matrix by singular value decomposition. Assuming that the singular value matrix can obtain the first K columns under a preset threshold, then the first K column vectors of the left singular matrix are the signal subspace .

[0080] Next, the parameters to be estimated are separated by matrix pencil (MP) and generalized eigenvalue decomposition (GED) and . The matrix bundle can be defined as follows:

[0081] (6)

[0082] Where A and B are square matrices of order N, and the generalized eigenvalues ​​of the matrix bundle are belongs to the set of complex numbers. Then, the set of generalized eigenvalues is defined as:

[0083] (7)

[0084] in, To solve a generalized eigenvalue, The complex domain.

[0085] Before constructing the specific matrix bundle, the triangle matrix method is first used to augment the signal matrix Decompose the left matrix to obtain the estimated parameters and The combination method is the same as the left singular matrix obtained by singular value decomposition. Therefore, the left matrix obtained by decomposition can be split according to the singular value matrix result to obtain the Vandermonde matrix containing the termite signal vector .

[0086] To separate For example, the matrix bundle can be constructed as follows:

[0087] (8)

[0088] (9)

[0089] (10)

[0090] (11)

[0091] Among them, the matrix For the matrix , Unknown matrices of the same rank (whose values ​​can be obtained by formulas (12) and (13)), matrices , The signal subspace matrix is The first L rows and the last L rows of the matrix, For the matrix The first L rows of the matrix, where L has the same value as L during matrix enhancement.

[0092] because

[0093] (12)

[0094] (13)

[0095] Therefore, the constructed matrix bundle is:

[0096] (14)

[0097] Among them, the auxiliary matrix To contain only The diagonal matrix of series terms, is the unit matrix. Based on formula (13), the auxiliary matrix The solutions of all generalized eigenvalues ​​of are A similar construction method can be used to separate all Value, that is, replace the rows in formulas (9) to (11) with columns to solve , let the matrix , The signal subspace matrix is The first L columns and the last L columns of the matrix, For the matrix The first L columns of the matrix, and then based on the matrix bundle The auxiliary matrix in All generalized feature values ​​are obtained A collection of values.

[0098] Based on the generalized eigenvalues ​​of the constructed matrix bundle, two sets of spatial frequency information sequences with a length of K can be obtained. Then, the matching between the two sets of eigenvalues ​​is optimized according to the objective function to obtain the two most orthogonal subspaces, thereby obtaining K pairs of matched spatial frequency pairs.

[0099] Define two auxiliary parameters for direction of arrival estimation:

[0100] (15)

[0101] (16)

[0102] in, and is the auxiliary parameter corresponding to the extracted generalized eigenvalue, which is used to characterize the phase change characteristics of the kth source in space. and are the estimated parameters obtained from the above matrix bundle separation.

[0103] Based on the auxiliary parameters and Get a direction of arrival estimate:

[0104] (17)

[0105] (18)

[0106] in, , are the elevation angle and azimuth angle of the signal source k, respectively. and are the geometric spacing of the array elements in two mutually orthogonal directions in the antenna array, that is, , Corresponding to , The array element spacing in the direction, Is an imaginary unit.

[0107] In one embodiment, in step S3, when DOA estimation is performed based on the phase difference of the obtained time domain data, the frequency containing angle information corresponding to each target is also solved, and the separation of multiple targets has actually been completed. Therefore, the complex amplitude corresponding to each signal source can be estimated by the least squares method, and the frequency domain signal sequence of each target source (signal source) can be separated accordingly, so as to obtain the frequency domain signal corresponding to each signal source.

[0108] The distance calculation of each termite target (signal source with signal amplitude not less than the threshold) to the origin of the spherical coordinate system (i.e., the reference element of the antenna array) is obtained. Since the influence of distance is mainly reflected in the delay and the phase difference caused by the delay, this embodiment considers solving the phase difference in the frequency domain signal.

[0109] First, the signal component corresponding to a certain frequency scale of the total signal in the frequency domain is modeled:

[0110] (19)

[0111] in, Indicates the frequency domain signal on the frequency scale The complex exponential signal under is the echo signal of source k The amplitude of the corresponding component under this frequency scale, the subscript i is the frequency index, Indicates the starting frequency of the bandwidth, represents the propagation delay of the source k relative to the radar’s transmission path and echo path, is the delay of reflection, which is a constant, j is an imaginary unit, is the frequency step size of the Fourier transform.

[0112] As for the representation of independent information sources in the frequency domain, each summed term in formula (19) can be taken out separately, and it can be expressed as:

[0113] (20)

[0114] in, Indicates that the frequency domain signal corresponding to the source k is on the frequency scale The complex exponential signal below.

[0115] When solving DOA, the time domain data received by each receiving array element at a certain moment has been processed with a frequency step size of The corresponding frequency domain signal is obtained by performing discrete Fourier transform. In addition, the MP (Matching Pursuit) algorithm is also used to decompose the auxiliary parameter pair for each source ( , ). The complex amplitude (i.e., signal amplitude) of the source at each frequency scale can be estimated based on this. In this embodiment, the signal amplitude is estimated by the least squares method, which can be expressed as:

[0116] (twenty one)

[0117] in, represents the information source k ( ) of the signal amplitude, It is used to characterize the contribution of the echo signals of K sources, that is, the amount of DOA information contained in each source solved by the generalized eigenvalue method. It can be further expressed as: In the left matrix of the above formula, from top to bottom, first follow The index is divided into M groups, from 0 to (M-1); within each group, The index traverses from 0 to (N-1). It can be seen that each element of the column vector output by the matrix multiplication on the left side of the formula corresponds to the element at the corresponding position of the matrix on the right side, that is, x(m;n), which corresponds to the echo signal received at the array element position (m,n).

[0118] After obtaining the estimated signal amplitude of each termite target, in fact, the time domain sub-signals corresponding to each signal source have been obtained, that is, based on The frequency domain signal sequence composed of has been obtained and is expressed as . Then we can base on Construct the corresponding Hankle matrix, denoted as matrix , and starting from this matrix, construct a matrix bundle and solve the generalized eigenvalue. Specifically, the corresponding matrix bundle is as follows:

[0119] (twenty two)

[0120] in, is a matrix Delete the first line, is a matrix Delete the last line, is the generalized eigenvalue to be solved. The obtained generalized eigenvalue is the phase difference caused by adjacent frequency scales required for distance measurement. , where the subscript p is the index of the termite target, that is, the index of the source whose signal amplitude is greater than or equal to the amplitude threshold. In this embodiment, it is considered that the phase difference mainly comes from the propagation delay. Therefore, for the same source, the phase difference caused by each adjacent two frequency scales ( ) is a constant and can be expressed as:

[0121] (twenty three)

[0122] in, is the frequency step size of Fourier transform, is the propagation delay of the signal source corresponding to the termite target p relative to the radar's transmitting path and echo path, that is, the total distance delay from the transmitting antenna to the termite target and then to the receiving antenna.

[0123] And then you can Inverse the distance:

[0124] (twenty four)

[0125] in, is the distance from the transmitting antenna to the termite target, is the distance from the termite target to the receiving antenna, c is the speed of light constant, is the imaginary part extraction operator, is the natural logarithm.

[0126] Ignoring the error caused by the geometric distance between the transmitting and receiving antennas, the target distance of the termite target can be obtained as:

[0127] (25)

[0128] On the other hand, the embodiment of the present invention also provides a termite detection system based on microwave radar through walls, such as Figure 2As shown, it includes a user terminal and a multi-input multi-output microwave radar detection device; wherein the antenna array of the microwave radar detection device is a uniform array of transceivers; data is transmitted between the user terminal and the microwave radar detection device by wireless or wired means; the user terminal configures the working parameters of the microwave radar detection device, such as working frequency, pulse repetition frequency, pulse width, transmission power, antenna beam width, etc.; then the microwave radar detection device scans the area to be detected (such as a wall) and receives the radar echo signal, and the microwave radar detection device performs received signal processing on the radar echo signal to obtain the original discrete time domain intermediate frequency signal; then the original discrete time domain intermediate frequency signal is bandpass filtered and discrete Fourier transformed to obtain The microwave radar detection device reconstructs the intermediate frequency signal in the frequency domain to obtain the estimated results of the arrival directions of the specified number of information sources; then, the signal amplitude corresponding to each information source is estimated based on the arrival direction estimation results to obtain the frequency domain signal of each information source; finally, for the information sources (termite targets) whose signal amplitude is greater than or equal to the amplitude threshold, the distance is measured based on the frequency domain signal of the information source to obtain the target distance of each information source; the termite detection result of the area to be detected is obtained based on the target distance of the information source and the corresponding arrival direction estimation result. At the same time, the microwave radar detection device transmits the termite detection result to the user terminal in real time, and visualizes the termite detection result on the user terminal.

[0129] In an embodiment of the present invention, a microwave radar detection device includes an antenna array, a data transmission interface, a wireless module, a radar data processing chip (radar board), a signal extraction processing unit and a power supply module; wherein the power supply module is used to supply power to the microwave radar detection device, the data transmission interface realizes wired data transmission between the microwave radar detection device and a user terminal based on a USB transmission protocol, the wireless module realizes wireless data transmission between the microwave radar detection device and the user terminal based on a WIFI module or a Bluetooth module, and the radar data processing chip is used to perform analog-to-digital conversion on the radar echo signal received by the antenna array to obtain a time domain signal, and then obtain the original discrete time signal through down-conversion processing (matched filtering and mixing processing). The signal extraction processing unit is used to perform bandpass filtering and discrete Fourier transform on the discrete time domain intermediate frequency signal, and then obtain the estimated results of the arrival direction of each source and the signal amplitude; then the termite target is screened out based on the signal amplitude and the target distance of the termite target is calculated, and the termite detection result of the area to be detected is obtained based on the target distance and the corresponding arrival direction estimation result, and the termite detection result is transmitted to the user terminal through a wireless module or a wired module, or after receiving the request message of the termite detection result initiated by the user terminal, the termite detection result of the area to be detected specified in the request message is returned to the user terminal. Preferably, the bandpass filter is specifically set to a bandpass filter of 0.1Hz-10Hz, so that the termite signal and the static background clutter can be better distinguished.

[0130] It should be noted that, in this embodiment, the signal extraction processing unit can also be deployed on an external device, such as an edge master device. The edge master device is connected to the microwave radar detection device and the user terminal respectively by wired or wireless means, obtains the original discrete time domain intermediate frequency signal from the radar data processing chip of the microwave radar detection device, and transmits the obtained termite detection result to the user terminal in real time. On the user terminal, the termite detection result can be visualized in two dimensions or three dimensions through a polar coordinate diagram and presented on the user terminal.

[0131] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0132] The above are only some embodiments of the present invention. For those skilled in the art, several modifications and improvements can be made without departing from the creative concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A method for detecting termites based on microwave radar through walls, characterized in that: The following steps are involved: Step 1, using a multi-input multi-output microwave radar detection device to scan the area to be detected, obtain radar echo signals and perform receiving signal processing on them to obtain the original discrete time domain intermediate frequency signal; wherein the antenna array of the microwave radar detection device is a uniform array integrating transmission and reception; Step 2, bandpass filtering the original discrete time domain intermediate frequency signal, and then performing discrete Fourier transform to obtain a frequency domain intermediate frequency signal containing target orientation information; Step 3, performing signal reconstruction processing on the frequency domain intermediate frequency signal to obtain the estimated results of the arrival directions of K information sources; wherein the number of information sources K is a preset value, and each information source is used to represent a candidate detection target of termites; And based on the arrival direction estimation result, the signal amplitude corresponding to each information source is estimated to obtain the frequency domain signal of each information source; Step 4, taking the signal source whose signal amplitude is greater than or equal to the preset amplitude threshold as the termite target, obtaining the target distance of each termite target, and obtaining the termite detection result of the to-be-detected area based on the target distance of the termite target and the corresponding arrival direction estimation result; wherein the target distance is used to characterize the distance between the signal source corresponding to the termite target and the reference array element of the antenna array of the microwave radar detection device; Step 5, sending the termite detection results of the area to be detected to the target user for visual display; Wherein, step 3 specifically includes the following steps: The original signal matrix is ​​constructed and matrix enhancement is performed on the frequency domain intermediate frequency signal to construct an augmented signal matrix ; For the augmented signal matrix Perform singular value decomposition and decompose the left singular matrix The first K column vectors of ; The matrix triangular decomposition method is used to calculate the augmented signal matrix Perform triangular decomposition, and obtain the Vandermonde matrix containing the termite signal vector based on the first K columns of the lower triangular matrix obtained by triangular decomposition ; By cutting and splicing the signal subspace Combined with the Vandermonde matrix Construct a matrix bundle, and obtain two sets of spatial frequency information sequences containing azimuth angles and elevation angles based on the generalized eigenvalues ​​of the matrix bundle, wherein the lengths of the two sets of spatial frequency information sequences are both K; Based on orthogonality, two sets of spatial frequency information sequences are paired to obtain K pairs of spatial frequency pairs in two mutually orthogonal directions relative to the antenna array. The arrival azimuth is estimated based on the matched spatial frequency pairs to obtain the arrival direction estimation result of each target. The pairing process of two sets of spatial frequency information sequences based on orthogonality is specifically as follows: the two sets of spatial frequency information sequences are allocated into K pairs, the two spatial frequencies of each spatial frequency pair come from different spatial frequency information sequences, the vectors corresponding to each spatial frequency pair are multiplied and then accumulated to obtain the return value of the objective function, and the final matching result is obtained based on the matching method corresponding to the minimum objective function; The estimation of the arrival azimuth according to the matched spatial frequency pair is as follows: definition and To match the two spatial frequencies in the spatial frequency pair, , They are the antenna arrays corresponding to , The array element spacing in the direction, the subscript k is the source index; Estimate the elevation and azimuth of source k: ， ； in, , are the elevation angle and azimuth angle of the signal source k, respectively, and the auxiliary parameters , The expressions are: , , represents the imaginary unit, is the natural logarithm; The number of information sources K is determined as follows: For the augmented signal matrix The singular values ​​obtained by singular value decomposition are squared and arranged in descending order, and then the cumulative variance explanation rate is calculated. When the cumulative variance explanation rate exceeds 90%, it is truncated, and the value of the number of information sources K is determined based on the number of singular values ​​included at the time of truncation.

2. A method for detecting termites based on microwave radar through walls as claimed in claim 1, characterized in that: In step 2, bandpass filtering is performed using a passband of 0.1-10 Hz.

3. A method for detecting termites based on microwave radar through walls as claimed in claim 1, characterized in that: The original signal matrix is ​​constructed in the following manner: the original signal matrix is ​​constructed according to the relative position relationship between each array element of the antenna array and the reference array element.

4. A method for detecting termites based on microwave radar through walls as claimed in claim 1, characterized in that: In step 4, the target distance of the termite target is calculated as: According to the formula Calculate the phase difference between the adjacent frequency scales of the signal source corresponding to the termite target p ,in is the frequency step size of Fourier transform, is the propagation delay of the signal source corresponding to the termite target p relative to the transmission path and echo path of the antenna array; According to the formula The target distance of the termite target p is calculated, where c is the speed of light constant, is the imaginary part extraction operator.

5. A termite detection system based on microwave radar through walls, characterized in that: It includes a user terminal and a multi-input multi-output microwave radar detection device; The antenna array of the microwave radar detection device is a uniform array integrating transmission and reception; data is transmitted between the user terminal and the microwave radar detection device by wireless or wired means; the user terminal configures the working parameters of the microwave radar detection device; The microwave radar detection device is used to scan the area to be detected and receive the radar echo signal, and perform received signal processing on the radar echo signal to obtain the original discrete time domain intermediate frequency signal; and is used to perform steps 2 to 4 in the microwave radar through-the-wall termite detection method according to any one of claims 1 to 4, so as to realize signal extraction processing of the original discrete time domain intermediate frequency signal and obtain the termite detection result of the area to be detected; The microwave radar detection device transmits the termite detection results to the user terminal in real time, and the termite detection results are visualized on the user terminal.

6. A termite detection system based on microwave radar through walls, characterized in that: It includes a multi-input and multi-output microwave radar detection device, an edge computing device for performing signal extraction processing, and a user terminal; in, The microwave radar detection device is connected to the user terminal through the edge computing device, and the connection mode between the edge computing device and the microwave radar detection device is a wired mode or a wireless mode; and the connection mode between the edge computing device and the user terminal is a wired mode or a wireless mode; The antenna array of the microwave radar detection device is a uniform array that integrates transmission and reception. The user terminal sends configuration instructions of working parameters to the microwave radar detection device through the edge computing device; The microwave radar detection device is used to scan the area to be detected and receive the radar echo signal, and perform received signal processing on the radar echo signal to obtain the original discrete time domain intermediate frequency signal and send it to the edge computing device; The edge computing device is used to execute steps 2 to 4 in the method for detecting termites based on microwave radar through walls as described in any one of claims 1 to 4, so as to realize signal extraction and processing of the original discrete time domain intermediate frequency signal, and obtain the termite detection result of the area to be detected; The edge computing device transmits the termite detection results of the area to be detected to the user terminal in real time, or transmits the termite detection results of the area to be detected specified by the request message initiated by the user terminal to the user terminal, so as to realize the visual display of the transmitted termite detection results on the user terminal.

Images