Abnormal heart rate detection method based on distributed MIMO radar

Abstract

The present application discloses an abnormal heart rate detection method based on distributed MIMO radar, and relates to the field of signal processing technology. The present application includes placing a MIMO radar in an observation area; obtaining the location information of the MIMO radar and the observation area, and collecting the received signal of the MIMO radar; inputting the received signal into a signal model, performing ML estimation on the abnormal heart rate after the abnormal change time of the heart rate parameter based on historical data, and performing abnormal heart rate detection based on the fastest change detection of GLRT using the distribution change before and after the abnormal change of the heart rate parameter; and analyzing the asymptotic performance of the fastest change detection of GLRT, and adjusting the position of the MIMO radar antenna by analyzing the asymptotic performance of the fastest detection of abnormal heart rate GLRT. The present application applies the fastest change detection based on GLRT to the problem of abnormal heart rate detection of human body, and combines it with distributed MIMO radar to greatly improve the detection performance. It also provides a performance evaluation of abnormal heart rate detection and uses it for the design of the position of the MIMO radar antenna.

Description

Technical Field

[0001] The present application relates to the field of signal processing technology, and in particular to an abnormal heart rate detection method based on distributed MIMO radar. Background Art

[0002] Human health is becoming a focus of attention. Among various common chronic diseases, the most aggressive are sudden cardiovascular diseases accompanied by abnormal heart rate. Radar's non-invasive and non-contact nature brings great advantages to human health detection and has attracted much attention. Multiple Input Multiple Output (MIMO) radar has been proven to bring performance gains in target detection compared to single-base radar.

[0003] Therefore, how to apply the MIMO radar system to the detection of abnormal human heart rate has become an urgent problem to be solved. Summary of the invention

[0004] The technical problem to be solved by the present application is: applying the MIMO radar system to the abnormal heart rate detection of the human body. The present application provides an abnormal heart rate detection method based on distributed MIMO radar to solve the above problem.

[0005] This application is implemented through the following technical solutions:

[0006] In a first aspect, the present application provides an abnormal heart rate detection method based on distributed MIMO radar, comprising:

[0007] Place MIMO radars in the observation area;

[0008] Obtain the location information of the MIMO radar and the observation area, and collect the receiving signal of the MIMO radar;

[0009] Input the received signal to the signal model, perform ML estimation on the abnormal heart rate after the abnormal change of heart rate parameters based on historical data, and perform abnormal heart rate detection based on GLRT using the distribution change before and after the abnormal change of heart rate parameters; analyze the asymptotic performance of GLRT's fastest change detection;

[0010] The MIMO radar antenna position is adjusted by analyzing the asymptotic performance of the fastest detection of abnormal heart rate GLRT.

[0011] Optionally, placing a MIMO radar in the observation area includes:

[0012] The observation area is divided into multiple areas, and a coordinate system is established, and multiple receiving antennas of the MIMO radar are placed corresponding to the divided observation areas.

[0013] Optionally, obtaining the location information of the MIMO radar and the observation area includes:

[0014] A coordinate set of a plurality of points in a plurality of regions is obtained, and location information of corresponding regions of a plurality of receiving antennas and transmitting antennas of a MIMO radar is obtained.

[0015] Optionally, collecting a received signal of a MIMO radar includes:

[0016] Obtain the transmitting antenna sampling value corresponding to the sampling time and the carrier frequency of the transmitting signal;

[0017] Obtain the reflection coefficients of the corresponding areas of the transmitting antenna and the receiving antenna;

[0018] Obtaining the displacement data of the observation area at the corresponding moment, including respiratory displacement data and heartbeat displacement data;

[0019] Get the amplitude of the harmonics of breathing and heartbeat in the corresponding area.

[0020] Optionally, the signal model includes:

[0021] The sampling samples, sampling time delay and breathing frequency of the received signal are obtained, and a useful signal including the heart rate frequency before the abnormal change of the heart rate is constructed for the sampling samples. The useful signals of multiple sampling samples are processed to obtain the MIMO radar observation vector for the abnormal change of the heart rate information, and the signal model is obtained by modeling the fastest change detection problem of abnormal heart rate through the observation vector;

[0022] The signal model includes a signal model in which the heart rate does not change and maintains a normal value and a signal model in which the heart rate changes from a normal value to an abnormal value.

[0023] Optionally, the GLRT-based fastest change detection uses distribution changes of heart rate parameters before and after abnormal changes to perform abnormal heart rate detection, including:

[0024] A likelihood function of the observation vector before and after the abnormal change in heart rate is constructed, and the log-likelihood ratio of the likelihood function corresponding to the two hypotheses of abnormal heart rate change and no change is processed. The historical data of heart rate is used to obtain the ML estimation of the abnormal heart rate after the change. The fastest change detection based on GLRT is used to obtain the existence judgment result of the unknown abnormal heart rate, and the existence judgment result is used to characterize whether an abnormal heart rate occurs.

[0025] In a second aspect, the present application provides a MIMO radar for abnormal heart rate detection, comprising a plurality of transmitting antennas and receiving antennas arranged in an observation area, wherein the transmitting antennas and receiving antennas collect sample information, and the MIMO radar is used to implement any of the methods described in the first aspect.

[0026] In a third aspect, an abnormal heart rate detection device based on a distributed MIMO radar comprises:

[0027] MIMO radars installed in the observation area;

[0028] An acquisition module is used to obtain the location information of the MIMO radar and the observation area, and collect the receiving signal of the MIMO radar;

[0029] A processing module, for inputting a received signal into a signal model, performing ML estimation of an abnormal heart rate after an abnormal change in heart rate parameters based on historical data, and performing abnormal heart rate detection based on a GLRT-based fastest change detection using distribution changes before and after an abnormal change in heart rate parameters;

[0030] An evaluation module for analyzing the asymptotic performance of GLRT's fastest change detection;

[0031] The control module is used to adjust the MIMO radar antenna position by analyzing the asymptotic performance of the fastest detection of abnormal heart rate GLRT.

[0032] In a fourth aspect, an abnormal heart rate detection system based on a distributed MIMO radar includes: a processor and a memory;

[0033] The memory stores computer-executable instructions;

[0034] When the processor executes the computer-executable instructions stored in the memory, the method as described in any one of the first aspects is implemented.

[0035] In a fifth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the method as described in any one of the first aspects.

[0036] This application has the following advantages and beneficial effects:

[0037] This application applies the GLRT-based fastest change detection to the problem of abnormal heart rate detection in humans, and combines it with a distributed MIMO radar system to bring performance gains.

[0038] This application applies the MIMO radar system to the detection of abnormal heart rate in the human body, greatly improving the detection performance.

[0039] This application also provides a performance evaluation of the MIMO radar system applied to human heart rate abnormality detection and uses it in the design of the MIMO radar antenna position. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The drawings described herein are used to provide a further understanding of the embodiments of the present application, constitute a part of the present application, and do not constitute a limitation on the embodiments of the present application. In the drawings:

[0041] Figure 1 A flowchart of an abnormal heart rate detection method based on distributed MIMO radar is provided in an embodiment of the present application.

[0042] Figure 2 A schematic diagram of heartbeat, breathing and overall chest displacement curves with abnormal heart rate changes provided in an embodiment of the present application.

[0043] Figure 3 A schematic diagram of simulation results of the MTFA-threshold provided in an embodiment of the present application using a 2×2 MIMO radar.

[0044] Figure 4 A schematic diagram of simulation results of ADD-MTFA using 2×2 MIMO radar provided in an embodiment of the present application.

[0045] Figure 5 A schematic diagram of an asymptotic performance curve of average detection delay at 15 dB for 2×2 and 3×3 MIMO radar antennas located at different positions provided in an embodiment of the present application. DETAILED DESCRIPTION

[0046] Before any embodiment of the present application is described in detail, it should be understood that the application of the present application is not limited to the details of the structures shown in the following description or drawings. The present application may adopt other embodiments and may be implemented or executed in various ways. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative improvements are within the scope of protection of the present application.

[0047] Human health is increasingly becoming a focus of attention. Among various common chronic diseases, the most aggressive are sudden cardiovascular diseases accompanied by abnormal heart rate.

[0048] The present application considers applying the MIMO radar system to abnormal heart rate detection of human body. Due to the non-invasive and non-contact characteristics of radar, the technical solution provided in the embodiment of the present application can be used to solve the problem of abnormal heart rate detection.

[0049] GLRT is a traditional statistical method for solving the fastest change detection problem with unknown parameters. This method requires estimating the unknown parameters first to achieve detection. The fastest change detection method based on GLRT provides an asymptotically optimal performance indicator for related problems. The fastest change detection method based on statistics can use the current data to obtain prior knowledge about the changes before and after, and realize the detection of data changes.

[0050] The present application provides an abnormal heart rate detection method based on distributed MIMO radar. Figure 1 The present invention provides a flowchart of an abnormal heart rate detection method based on distributed MIMO radar. Figure 1 As shown, including:

[0051] S101, placing a MIMO radar in the observation area;

[0052] S102, obtaining location information of the MIMO radar and the observation area, and collecting a received signal of the MIMO radar;

[0053] S103, inputting the received signal into the signal model, performing ML estimation on the abnormal heart rate after the abnormal change of the heart rate parameter based on historical data, and performing abnormal heart rate detection based on the fastest change detection of GLRT using the distribution change before and after the abnormal change of the heart rate parameter;

[0054] S104, analyze the asymptotic performance of the fastest change detection of GLRT;

[0055] S105. Adjust the MIMO radar antenna position by analyzing the asymptotic performance of the abnormal heart rate GLRT fastest detection.

[0056] In a possible implementation, S101 includes dividing the observation area into multiple areas, establishing a coordinate system, and placing multiple receiving antennas of the MIMO radar corresponding to the divided observation areas.

[0057] In a possible implementation, S102 includes obtaining a coordinate set of a plurality of midpoints in a plurality of regions, and obtaining location information of corresponding regions of a plurality of receiving antennas and transmitting antennas of the MIMO radar.

[0058] In a possible implementation, S103 includes obtaining a transmitting antenna sampling value corresponding to a sampling moment and a carrier frequency of a transmitting signal;

[0059] Obtain the reflection coefficients of the corresponding areas of the transmitting antenna and the receiving antenna;

[0060] Obtaining the displacement data of the observation area at the corresponding moment, including respiratory displacement data and heartbeat displacement data;

[0061] Get the amplitude of the harmonics of breathing and heartbeat in the corresponding area.

[0062] In a possible implementation, the signal model in S103 includes the following detailed explanation:

[0063] The signal model obtains the sampling samples, sampling time delay and breathing frequency of the received signal, constructs a useful signal including the heart rate frequency before the abnormal change of the heart rate for the sampling samples, obtains the MIMO radar observation vector for the abnormal change information of the heart rate by processing the useful signals of multiple sampling samples, and obtains the signal model by modeling the fastest change detection problem of abnormal heart rate through the observation vector;

[0064] The signal model includes a signal model in which the heart rate does not change and maintains a normal value and a signal model in which the heart rate changes from a normal value to an abnormal value.

[0065] In a possible implementation, S103 also includes constructing a likelihood function of the observation vector before and after the abnormal change in heart rate, processing the log-likelihood ratio of the likelihood function corresponding to the two hypotheses that the abnormal change in heart rate occurs and that the change does not occur, using the historical data of heart rate to obtain the ML estimate of the abnormal heart rate after the change, and using the GLRT-based fastest change detection to obtain a judgment result on the existence of the unknown abnormal heart rate, wherein the judgment result is used to characterize whether an abnormal heart rate occurs.

[0066] The embodiment of the present application provides a method for detecting abnormal heart rate of a human body based on a distributed MIMO radar and using GLRT fastest change detection, including:

[0067] For the convenience of description, the following definitions are first made:

[0068] is the transpose, () H is the conjugate transpose, Re{·} is the real part, Im{·} is the imaginary part, is the gradient operation, It means to find the mathematical expectation, and |·| means to take the absolute value.

[0069] Methods include:

[0070] Step 1:

[0071] Taking the heart of a person at X0 = (x0, y0) as the center, the chest observation area of ​​the person is divided into I areas for placing MIMO radars, where the coordinate set of the midpoint in the i-th (i = 1, ..., I) area is

[0072]

[0073] Among them, the coordinates represents a point in the i-th region.

[0074] Step 2:

[0075] For the distributed MIMO radar system, the signal sampling values ​​received by N receivers placed in I area are arranged in sequence into a column to form a received signal

[0076]

[0077] in,

[0078]

[0079]

[0080]

[0081]

[0082]

[0083]

[0084]

[0085] M is the number of transmitting antennas of the MIMO radar system, and N is the number of receiving antennas of the MIMO radar system; if the mth transmitting antenna is located in the i region, it is recorded as The nth transmitting antenna is located in region i and is denoted as mth i Transmitting antennas in kT s The sampling value at time is in is the transmitted signal power, T s is the sampling time interval, k is the sampling number, Indicates the transmission signal. Indicates m i The carrier frequency of the signal transmitted by the antenna. i Represents the reflection coefficient of region i, represented by the real part α i,Re and the imaginary part α i,Im Composition, the delay at time k By m i ,n i Nominal distance of the path and the chest displacement μ at time k i [k] composition, which includes respiratory displacement and heartbeat displacement where f r Respiratory frequency, f h Indicates heart rate (HR), Indicates the initial phase of breathing. Indicates the initial phase of heartbeat. and represents the amplitude of the pth harmonic of breathing and the qth harmonic of heartbeat in the ith region respectively. Usually, the amplitude of breathing is much larger than the amplitude of heartbeat, and the respiratory frequency f r can be obtained relatively easily. Therefore, it is assumed here that f is known r , and HR is the variable parameter of interest. Assuming f h The value before the change can be obtained based on historical data and is known. Indicates the mth i n i The noise on the path is assumed to be a white complex Gaussian distribution in space and time, with a mean of 0 and a variance of

[0086] Step 3: Construct the following detection problem

[0087]

[0088]

[0089]

[0090] Among them, H0 represents the signal model in which the heart rate does not change and remains at a normal value, and H1 represents the signal model in which the heart rate changes from a normal value to an abnormal value. ah Indicates the abnormal heart rate value after the change. Due to the lack of predictability, in practical applications, f ah Usually unknown.

[0091] Step 4: As the unknown parameter k * and f ah The log-likelihood ratio can be written as

[0092]

[0093] Step 5: Applying an idea similar to GLRT, all historical data is used to obtain the unknown parameters f ah ML estimate of * ,Right now

[0094]

[0095] Among them, "sup" means taking the supremum, and esssupA means taking the essential bound of A, that is, finding the minimum B that satisfies P(A≤B)=1.

[0096] Step 6: The test statistic for the fastest detection based on GLRT can be written as

[0097]

[0098] Step 7: The stopping time, i.e. the estimated change time, is given by

[0099]

[0100] Step 8: Use Representing the asymptotic version of the ML estimate of abnormal HR

[0101]

[0102] Among them, J -1 (f ah ) is f ah Estimated CRB,

[0103]

[0104] Step 9: You can calculate the asymptotic Statistical mean log-likelihood ratio

[0105]

[0106] Step 10: Replace it with its asymptotic version The test statistic becomes

[0107]

[0108] Step 11: The asymptotic stopping time can be given by

[0109]

[0110] Step 12: To evaluate the detection performance, construct a performance metric, Mean Time to False Alarm (MTFA)

[0111]

[0112] and Worst Case Detection Delay (WADD)

[0113]

[0114] Step 13: Stop Time and Substituting into the performance indicators, the performance of the GLRT-based fastest change detection and asymptotic analysis is evaluated, and the upper and lower bounds of the performance indicators are given.

[0115] In the process of providing the above method, in a specific example, the embodiment of the present application provides another abnormal heart rate detection method based on distributed MIMO radar, including:

[0116] Assuming that the object is stationary, take the same height H as the human chest, establish a two-dimensional Cartesian coordinate system, with the human heart located at X0 = (x0, y0) as the center, and the human chest observation area is divided into I areas, where the coordinate set of the midpoint in the i-th area is

[0117]

[0118] Among them, the coordinates Represents a point in the i-th region. Consider a distributed MIMO radar system for contactless heart rate estimation, M is the number of transmitting antennas of the MIMO radar system, and N is the number of receiving antennas of the MIMO radar system; if the m-th transmitting antenna is located in the i-th region Record it as The nth transmitting antenna is located in region i Record it as mth i Transmitting antennas in kT s The sampling value at time is in is the transmitted signal power, T s is the sampling time interval, and k is the sampling number.

[0119] Therefore, due to the transmission from the transmitting antenna m i Transmitting and receiving antenna n i The low-pass equivalent of the observed signal can be written as

[0120]

[0121] Among them, α i =α i,Re +jα i,Im =Re(α i )+jIm(α i ) is the unknown target complex reflection coefficient determined in region i, Indicates m i The carrier frequency of the signal transmitted by the antenna. Indicates the mth i n i The noise on the path is assumed to be a white complex Gaussian distribution in space and time, with a mean of 0 and a variance of The delay at time k By m i ,n i Nominal distance of the path and the chest displacement μ at time k i [k] composition,

[0122]

[0123] in, represents the respiratory displacement and represents the heartbeat displacement of region i. and is assumed to have frequency f r and f h Initial Phase and A zero-mean periodic function, which can be represented by a Fourier sine series,

[0124]

[0125]

[0126] in, and represents the amplitude of the pth harmonic of breathing and the qth harmonic of heartbeat in the ith region respectively. Usually, the amplitude of breathing is much larger than the amplitude of heartbeat, and the respiratory frequency f r can be obtained relatively easily. Therefore, it is assumed here that f is known r And HR is the variable parameter of interest. Assume f h The value before the change can be obtained from historical data and is known. For simplicity and without loss of generality, let p=1,q=1, so we have

[0127] The useful signal in formula (2) is defined as

[0128]

[0129] Taking samples from k = 1 to K, the observation vector of the distributed MIMO radar system can be written as

[0130]

[0131] in,

[0132] Assume that if HR changes from a normal value to an abnormal value, the change occurs at time 1≤k * ≤ K. Therefore, the fastest detection problem of abnormal HR alarms can be modeled as follows

[0133]

[0134]

[0135]

[0136] Among them, H0 represents the signal model in which the heart rate does not change and remains at a normal value, and H1 represents the signal model in which the heart rate changes from a normal value to an abnormal value.ah ∈[abnormal heart rate interval] represents the value of abnormal heart rate after the change. Due to the lack of predictability, in practical applications, f ah Usually unknown.

[0137] As described above, in a possible implementation provided by the present application, the fastest change detection detects the change as quickly as possible by distinguishing the distribution changes caused by the heart rate parameters before and after the change. The present application takes into account that the abnormal HR after the change time is unknown, and adopts the fastest change detection method based on GLRT to replace the unknown abnormal HR with its ML estimate.

[0138] In a possible implementation provided by the present application, in order to facilitate performance analysis, the approximate distribution of the ML estimate of HR is used to derive the asymptotic performance of the fastest change detection based on GLRT. The bounds of the mean time to false alarm (MTFA) and the worst-case average detection delay (WADD) are derived.

[0139] This application uses the following steps to detect abnormal changes in heart rate using MIMO radar combined with the fastest change detection based on GLRT and gives its asymptotic analysis performance:

[0140] Step 1: At the changing time k * The likelihood function before and after is

[0141]

[0142] Step 2: As the unknown parameter k * and f ah The log-likelihood ratio can be written as

[0143]

[0144] Step 3: Apply the GLRT fastest detection method, and all historical data are used to obtain the unknown parameters f ah ML estimate of * ,Right now

[0145]

[0146] Among them, "sup" means taking the supremum, and esssupA means taking the essential bound of A, that is, finding the minimum B that satisfies P(A≤B)=1.

[0147] Step 4: Replace the unknown abnormal HR with its ML estimate, and the test statistic for the fastest detection based on GLRT can be written as

[0148]

[0149] Step 5: The stopping time, i.e. the estimated change time, is given by

[0150]

[0151] Step 6: Use Representing the asymptotic version of the ML estimate of abnormal HR

[0152]

[0153] Among them, J -1 (f ah ) is f ah Estimated CRB,

[0154]

[0155] Step 7: You can calculate the asymptotic Statistical mean log-likelihood ratio

[0156]

[0157] Step 8: Replace it with its asymptotic version The test statistic becomes

[0158]

[0159] Step 9: The asymptotic stopping time can be given by

[0160]

[0161] Step 10: To evaluate the detection performance, construct the performance metric Mean Time to False Alarm (MTFA)

[0162]

[0163] and Worst Case Detection Delay (WADD)

[0164]

[0165] Among them, the average detection delay (ADD)

[0166]

[0167] Step 11: Stop Time and Substitute them into the performance indicators to evaluate the performance of the fastest change detection and asymptotic analysis based on GLRT.

[0168] Step 12: The fastest change detection based on GLRT has the following asymptotic performance bounds

[0169]

[0170]

[0171] In a specific possible implementation provided in the embodiment of the present application, Figure 3 A schematic diagram of simulation results of an MTFA-threshold using a 2×2 MIMO radar provided in an embodiment of the present application. Figure 4 A schematic diagram of simulation results of an ADD-MTFA using a 2×2 MIMO radar provided in an embodiment of the present application. Figure 3 , Figure 4 As shown, the simulation parameters are set as follows:

[0172] Assume that the position of the human heart is X0 = (x0, y0) = (1 / 2, 1 / 2) m in a two-dimensional coordinate system. The observation area of ​​the target heart is divided into I = 12 equally spaced areas around X0 for the radar to transmit and receive signals in the corresponding areas.

[0173] Consider two radar configurations. One is a single-input single-output (SISO) radar system with M=1 transmit antenna located at (0,0)m in region 1 and N=1 receive antenna located at (0.125,0.125)m in region 1. The other system is a MIMO radar with M=2 transmit antennas located at (0,0)m in region 1 and (0.35,0)m in region 2, and N=2 receive antennas located at (0.125,0.125)m in region 1 and (0.475,0.125)m in region 2.

[0174] A set of sinusoidal pulse signals are used for transmission

[0175]

[0176] The pulse width is T′=0.01s, and the frequency offset between adjacent transmitted signals is f Δ =200GHz, T s =10 - 4 s. The signal-to-noise ratio is defined as For all m i make

[0177] Assume that the target maintains a normal heart rate f at the beginning h =1.22Hz and respiratory rate f r =0.35Hz, then the heart rate changes from normal to abnormal f ah =2.5Hz, and other parameters remain unchanged. The heartbeat signal (solid line), respiratory signal (dotted line), and overall chest displacement (solid line) are plotted on Figure 2 In. You can Figure 2 It can be seen that the respiratory signal frequency is dominant, so it is difficult to observe the changes in heart rate from the overall chest displacement graph.

[0178] 100 uninterrupted received signal samples with variation time are collected for abnormal HR alarm. The first 50 samples are from the normal heart rate stage, and the last 50 samples are from the abnormal heart rate stage, which means that the abnormal HR occurs at k * =51.

[0179] Figure 3 The GLRT-based approach and the asymptotic analysis of MTFA versus threshold using a 2×2 MIMO radar are shown, considering the first 50 samples under the assumption H0.

[0180] The solid and dashed curves represent the GLRT-based approach and asymptotic analysis. The black dashed line represents the theoretical lower bound obtained in equation (21). Figure 3 It can be seen that MTFA increases with the increase of threshold and takes a maximum value of 50, while for a given threshold, MTFA increases with the increase of SNR, which means that the false alarm rate decreases. In addition, the fastest detection based on GLRT and asymptotic analysis show that the MTFA at different SNRs is higher than the theoretical limit e h , which satisfies the bounds obtained in formula (21). Numerical results show that the fastest MTFA based on GLRT and when h = ln51, the asymptotic analysis can reach e h It is noteworthy that the GLRT-based method gives false alarms earlier than the asymptotic analysis, which means that the asymptotic analysis can provide the achievable asymptotically optimal performance for the GLRT-based method when the signal-to-noise ratio is large enough.

[0181] exist Figure 3 , showing that MTFA can reach the lower limit of formula (21). Therefore, the lower limit of formula (21) is used to achieve a sufficiently large MTFA time. Figure 4 In the paper, the ADD and Figure 4 It shows that for a fixed MTFA constraint, ADD decreases as SNR increases. This result means that as the SNR increases, the system can alert abnormal heart rate more quickly. For a 2×2 MIMO radar system, the ADD of the GLRT-based method is The curves at 0dB, 15dB and 25dB are close to asymptotic analysis. Figure 4 The GLRT-based method and asymptotic analysis are demonstrated for radars using 2×2 MIMO. The curves can be bounded by the theoretical limit formula (22) at different signal-to-noise ratios. In addition, when the signal-to-noise ratio increases, the GLRT method and the asymptotic analysis The curve is closer to the theoretical limit in formula (22). These results verify the correctness of formula (22).

[0182] The correctness of the asymptotic performance of the fastest detection based on GLRT has been verified. Next, the influence of MIMO radar antenna position on the abnormal HR alarm performance is discussed. Since the 2×2 MIMO radar can achieve asymptotic performance below 15dB, such as Figure 4 As shown, for simplicity, asymptotic performance is used to analyze the impact of MIMO radar antenna position on abnormal HR alarm performance at 15dB. Two MIMO radar setups, namely 2×2 and 3×3 radar systems, are considered for abnormal HR alarm. For 2×2 MIMO radar, two scenarios are considered. For scenario 1, a pair of transmit and receive antennas are placed in area 1, and another pair of transmit and receive antennas are placed in area 2. For scenario 2, a pair of transmit and receive antennas are placed in area 1, and another pair of transmit and receive antennas are placed in area 3.

[0183] For 3×3 MIMO radar, three scenarios are considered, namely, two pairs of antennas are placed the same as in scenario 1 (areas 1 and 2), and the third pair of transmit and receive antennas are located in areas 3 or 4, or two pairs of antennas are placed the same as in scenario 2 (areas 1 and 3), and the third pair of transmit and receive antennas are located in area 4. 2×2 and 3×3 MIMO radar antennas are placed in different areas The asymptotic performance curve is as follows Figure 5 shown. Figure 5 Not only did it confirm that the 3×3 radar with more antennas has a smaller detection delay, but it also verified the impact of different antenna positions on the abnormal HR alarm performance. These results are due to the different heartbeat amplitudes observed in different areas. Therefore, if the number of antennas is limited, you can consider placing them in areas where the heartbeat amplitude can be clearly observed, which helps to quickly issue abnormal alarms.

[0184] The specific implementation methods described above further illustrate the purpose, technical solutions and beneficial effects of the present application in detail. It should be understood that the above description is only the specific implementation method of the present application and is not intended to limit the scope of protection of the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A method for detecting abnormal heart rate based on distributed MIMO radar, characterized in that: include: Place MIMO radars in the observation area; Obtain the location information of the MIMO radar and the observation area, and collect the received signal of the MIMO radar, including: Obtain the transmitting antenna sampling value corresponding to the sampling time and the carrier frequency of the transmitting signal; Obtain the reflection coefficients of the corresponding areas of the transmitting antenna and the receiving antenna; Obtaining the displacement data of the observation area at the corresponding moment, including respiratory displacement data and heartbeat displacement data; Obtain the amplitude of the harmonics of breathing and heartbeat in the corresponding area; Input the received signal to the signal model, perform ML estimation of the abnormal heart rate after the abnormal change of heart rate parameters based on historical data, and perform abnormal heart rate detection based on the fastest change detection of GLRT using the distribution changes before and after the abnormal change of heart rate parameters, including: Constructing the likelihood function of the observation vector before and after the abnormal change of heart rate, processing the log-likelihood ratio of the likelihood function under the two assumptions that the abnormal change of heart rate occurs and does not occur, using the historical data of heart rate to obtain the ML estimation of the abnormal heart rate after the change, using the GLRT fastest change detection to obtain the existence judgment result of the unknown abnormal heart rate, and the existence judgment result is used to characterize whether the abnormal heart rate occurs; Analyze the asymptotic performance of GLRT for fastest change detection; Adjust the MIMO radar antenna position by detecting the asymptotic performance of the fastest change of abnormal heart rate GLRT; The signal model includes: The sampling samples, sampling time delay and breathing frequency of the received signal are obtained, and a useful signal including the heart rate frequency before the abnormal change of the heart rate is constructed for the sampling samples. The useful signals of multiple sampling samples are processed to obtain the MIMO radar observation vector for the abnormal change of the heart rate information, and the signal model is obtained by modeling the fastest change detection problem of abnormal heart rate through the observation vector; The signal model includes a signal model in which the heart rate does not change and maintains a normal value and a signal model in which the heart rate changes from a normal value to an abnormal value.

2. The method according to claim 1, characterized in that Placing a MIMO radar in the observation area includes: The observation area is divided into multiple areas, and a coordinate system is established, and multiple receiving antennas of the MIMO radar are placed corresponding to the divided observation areas.

3. The method according to claim 2, characterized in that The obtaining of the location information of the MIMO radar and the observation area includes: The coordinate sets of the midpoints of the multiple areas are obtained, and the position information of the corresponding areas of the multiple receiving antennas and transmitting antennas of the MIMO radar is obtained.

4. A MIMO radar for abnormal heart rate detection, characterized in that: It comprises a plurality of transmitting antennas and receiving antennas arranged in an observation area, wherein the transmitting antennas and receiving antennas collect sample information, and the MIMO radar is used to implement the method according to any one of claims 1 to 3.

5. An abnormal heart rate detection device based on distributed MIMO radar, characterized in that: The device is used to implement the method according to any one of claims 1 to 3, including: MIMO radar installed in the observation area; An acquisition module is used to obtain the location information of the MIMO radar and the observation area, and collect the receiving signal of the MIMO radar; A processing module, used for inputting a received signal into a signal model, performing ML estimation of an abnormal heart rate after an abnormal change of a heart rate parameter based on historical data, and performing abnormal heart rate detection based on the fastest change detection of GLRT using a distribution change before and after an abnormal change of a heart rate parameter; An evaluation module for analyzing the asymptotic performance of GLRT’s fastest change detection; The control module is used for adjusting the MIMO radar antenna position by detecting the asymptotic performance of the fastest change of the abnormal heart rate GLRT.

6. An abnormal heart rate detection system based on distributed MIMO radar, characterized in that: include: Processor and memory; The memory stores computer-executable instructions; When the processor executes the computer-executable instructions stored in the memory, the method according to any one of claims 1 to 3 is implemented.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 3 when executed by a processor.

Images