Underground personnel and equipment consistency monitoring method based on millimeter waves and UWB

Abstract

The invention relates to the technical field of underground personnel identification, in particular to an underground personnel equipment consistency monitoring method based on millimeter waves and UWB. The method comprises the following steps: sending a radar signal to underground personnel through a millimeter-wave radar to carry out information acquisition, obtaining an echo signal of the underground personnel, carrying out short-time Fourier transform and fast Fourier transform processing, and respectively constructing a gait data set of the underground personnel containing identity information and distance information; acquiring identity information and position information of underground personnel through a multi-branch network and a target detection algorithm; information reading is carried out based on a preset UWB tag on the underground equipment, and identity information and position information corresponding to the equipment are obtained; and performing consistency judgment according to the identity information and the position information of the underground personnel and the identity information and the position information of the equipment. According to the method and the device, the influence of weak underground illumination intensity on gait recognition can be eliminated, abundant human gait information and position information are acquired, and thus consistent monitoring of underground personnel and equipment is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of underground personnel identification, and in particular to a method for monitoring the consistency of underground personnel equipment based on millimeter waves and UWB. Background Art

[0002] At present, there are some problems such as people going down into coal mines not carrying personnel location identification cards, people and cards not matching, or one person having multiple cards. These problems can easily lead to over-staffing in coal mines, and the inability to accurately grasp the number and location of trapped people underground in a timely manner after an accident occurs, thereby increasing the difficulty of accident rescue. By real-time monitoring and positioning of the miners' position, the underground personnel positioning system can provide key safety guarantees. In an emergency, rescue personnel can quickly locate the miners and launch rescue operations.

[0003] The existing single underground personnel positioning technologies include radio frequency identification, face recognition, iris recognition, and video-based gait recognition. However, the radio frequency identification method is easy to lose, forge, and replace because the system recognizes the identification card, and it is difficult to determine whether the person has no card or multiple cards. Fingerprint recognition has the following disadvantages: it is easy to imitate fingerprint film, easy to obtain, and the imitation technology is mature and low-cost; most of the underground personnel are engaged in physical labor, and their fingerprints are severely worn, which easily leads to recognition failure; the underground environment is poor, and the personnel's fingers are covered with coal dust and dust, resulting in fingerprints that cannot be recognized. Face recognition is limited by the technology itself. Factors such as the angle of the facial image, the covering on the face (hats, masks, glasses, hairstyles, etc.), and the intensity of light on the scene will affect the recognition success rate. Since the faces of coal mine personnel are easily contaminated with coal dust, and because the underground personnel wear dust masks and helmets, it cannot be used underground. Currently, facial recognition equipment is installed at the entrance of coal mines, which cannot achieve full mine coverage; iris recognition is less convenient to use, has a slightly lower recognition accuracy, a slower response speed, and its monitoring distance is within 1 meter, requiring the human eye to be within the recognition frame. Therefore, it is currently only used at the entrance of the mine, and cannot achieve full mine coverage; video-based gait recognition is greatly affected by lighting conditions, and the accuracy of gait recognition in low-light environments underground is low.

[0004] Based on the difficulties existing in the above-mentioned existing technologies in underground identification, it is impossible to achieve mutual coordination and unified management between underground personnel, vehicles, equipment, and materials, so as to solve the problems such as scheduling of underground personnel and equipment by the above-ground dispatcher when the precise location of the underground personnel and equipment is clear. In addition, the current consistency monitoring technology mainly monitors a single target, and most of the technologies cannot be applied to the underground environment. Therefore, there is an urgent need for a method that can realize consistency monitoring of underground personnel and equipment as well as personnel and areas. Summary of the invention

[0005] In order to solve the problems existing in the prior art, the present invention provides a method for monitoring the consistency of underground personnel and equipment based on millimeter waves and UWB. The scheme includes: sending radar signals to underground personnel through millimeter wave radar to collect information, obtaining their echo signals for short-time Fourier transform and fast Fourier transform processing, and constructing gait data sets of underground personnel containing identity information and distance information respectively; obtaining the identity information and location information of underground personnel through multi-branch networks and target detection algorithms; reading information based on the preset UWB tags on the underground equipment, and obtaining the identity information and location information corresponding to the equipment; and making consistency judgments based on the identity information, location information of the underground personnel and the identity information and location information of the equipment. The present application can eliminate the influence of weak illumination intensity underground on gait recognition, obtain rich human gait information and location information, and thus realize consistency monitoring of underground personnel and equipment.

[0006] The present invention adopts the following technical solution, a method for monitoring the consistency of underground personnel equipment based on millimeter wave and UWB, including: The millimeter-wave radar sends radar signals to underground personnel for information collection and obtains their echo signals; Performing short-time Fourier transform processing on the echo signal to construct a gait data set containing identity information of underground personnel; A gait recognition model is constructed using a pre-trained multi-branch network, and a gait data set containing identity information of underground personnel is input into the gait recognition model to obtain the identity information of the underground personnel; Performing fast Fourier transform processing on the echo signal to construct a gait data set containing underground personnel distance information; Using CA-CFAR target detection to process the gait data set containing the distance information of the underground personnel, and obtain the position information of the underground personnel; Read information based on the preset UWB tags on the underground equipment to obtain the identity information and location information corresponding to the equipment; The consistency is judged based on the identity information and location information of the underground personnel and the identity information and location information of the equipment.

[0007] Furthermore, the echo signal includes gait information and distance information of the underground personnel.

[0008] Furthermore, the gait information includes gait information of underground personnel at multiple walking speeds; the multiple walking speeds include slow speed, normal speed and fast speed.

[0009] Furthermore, constructing a multi-branch network for gait recognition based on the gait data set to obtain the identity information of the underground personnel includes: Inputting the gait data set into a multi-branch network to perform preliminary feature extraction to obtain shallow gait features; The preliminary features are subjected to deep feature extraction by multiple branches in a multi-branch network to obtain deep gait features; The deep gait feature is compared with a preset personnel gait data set to determine the identity information of the underground personnel.

[0010] Furthermore, the multi-branch network includes: a residual convolution layer, a first branch, a second branch and a third branch; The residual convolution layer is used to perform preliminary extraction on the gait dataset to obtain shallow gait features; The first branch is used to extract global features of shallow gait features; The second branch divides the shallow gait feature level into two parts for local feature extraction; The third branch divides the shallow gait feature level into four parts for local feature extraction.

[0011] Furthermore, the location information of underground personnel is obtained, including: The CA-CFAR target detection algorithm is used to detect underground personnel in a gait data set containing the position information of underground personnel, and an angle measurement algorithm is used to measure the angle between the corresponding echo signal in the gait data set containing the position information of underground personnel and the position of the radar, and the position of the underground personnel corresponding to the echo signal is determined based on the angle.

[0012] Furthermore, information is read based on the preset UWB tags on the downhole equipment, specifically: Use TOA ranging method to perform UWB positioning on underground equipment and obtain the location information of underground equipment; The identity information of the downhole equipment is determined according to the UWB tag corresponding to the downhole equipment.

[0013] The beneficial effects of the present invention are as follows: the present invention collects personnel gait information by using millimeter-wave radar, and processes the gait information into a micro-Doppler spectrum to realize personnel identification, thereby eliminating the influence of weak light intensity underground on gait identification; millimeter-wave radar has a high speed resolution, is easy to obtain rich human gait information and position information, and has the characteristics of contactless identification. Combined with UWB technology to obtain equipment ID and position information, it can monitor whether personnel are carrying their own unique equipment, and determine whether personnel can enter restricted areas, thereby realizing consistency monitoring of personnel and equipment as well as personnel and areas underground. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art 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 paying creative labor.

[0015] Figure 1 A schematic flow chart of a method for monitoring the consistency of underground personnel equipment based on millimeter waves and UWB according to an embodiment of the present invention; Figure 2 A schematic diagram of a multi-branch network structure according to an embodiment of the present invention; Figure 3 Schematic diagram of the detection principle of a one-dimensional CA-CFAR according to an embodiment of the present invention; Figure 4 A schematic diagram of an angle measurement principle of an embodiment of the present invention; Figure 5 The figure is a schematic diagram of the principle of a TOA ranging method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] The flowchart of the method for monitoring the consistency of underground personnel equipment based on millimeter wave and UWB in the embodiment of the present invention is as follows: Figure 1 As shown in the figure, the framework of the underground personnel and equipment identification, positioning and consistency monitoring method based on millimeter wave and UWB proposed in the present invention is mainly divided into two parts: gait recognition based on micro-Doppler and consistency monitoring of underground personnel, equipment and personnel areas. In the gait recognition stage, data preprocessing is performed to complete the construction of the micro-Doppler spectrum gait data set, and the neural network is used to extract features and classify to obtain personnel identity information (ID); after preprocessing the information collected by the millimeter wave radar, the position coordinates of the personnel are obtained through the CFAR target detection algorithm and horizontal angle estimation; UWB technology is used to obtain the ID and position coordinates of the equipment, so as to finally make a consistency judgment on the personnel and equipment information, which specifically includes: The millimeter-wave radar sends radar signals to underground personnel for information collection and obtains their echo signals; Performing short-time Fourier transform processing on the echo signal to construct a gait data set containing identity information of underground personnel; A gait recognition model is constructed using a pre-trained multi-branch network, and a gait data set containing identity information of underground personnel is input into the gait recognition model to obtain the identity information of the underground personnel; Performing fast Fourier transform processing on the echo signal to construct a gait data set containing underground personnel distance information; Using CA-CFAR target detection to process the gait data set containing the distance information of the underground personnel, and obtain the position information of the underground personnel; Read information based on the preset UWB tags on the underground equipment to obtain the identity information and location information corresponding to the equipment; The consistency is judged based on the identity information and location information of the underground personnel and the identity information and location information of the equipment.

[0018] In the embodiment of the present invention, a radar signal is sent to the underground personnel by a millimeter wave radar to collect information, and an echo signal thereof is obtained. The echo signal is subjected to short-time Fourier transform processing to construct a gait data set containing the identity information of the underground personnel, which specifically includes: The millimeter wave radar will send a signal to the underground personnel and receive a target echo signal, wherein the echo signal contains the gait information of the underground personnel and the distance information between different underground personnel; since the echo signal will be affected by various factors during the propagation process, the amplitude and phase of the echo signal will change to form a complex signal, so the embodiment of the present invention first samples the echo signal and performs Fourier transform on it, and obtains the distance information between different underground personnel therefrom, thereby obtaining a time-distance diagram of the underground personnel, and then uses a moving target indicator (MTI) method to suppress clutter using a clutter suppression filter to improve the signal-to-clutter ratio of the radar signal, and finally selects the distance unit where the human body is walking in the time-distance diagram, performs short-time Fourier transform on the signal of each distance unit and sums them to obtain the final time-frequency spectrum diagram, which can reflect the process of energy intensity and frequency of each scattering point of the underground personnel changing with time.

[0019] Short-time Fourier transform is a commonly used time-frequency analysis method, which uses a sliding window to perform Fourier transform (FFT) on the signal in the time window to obtain a time-frequency spectrum.

[0020] The acquired underground personnel atlas is constructed to obtain a gait data set containing the identity information of the underground personnel. Among them, since the walking speed of the personnel cannot be kept consistent, there are differences in gait. In the embodiment of the present invention, the walking speed is uniformly set, so that gaits at three walking speeds can be obtained, including slow, normal and fast.

[0021] In the embodiment of the present invention, the gait data set containing the identity information of the underground personnel is input into a pre-trained gait recognition model to obtain the identity information of the underground personnel, specifically including: A multi-branch network is used to construct a gait recognition model for gait recognition, thereby obtaining the identity information of underground personnel. The structure of the multi-branch network is as follows: Figure 2 As shown, it includes a branch for global feature representation and two branches for local feature representation; a multi-granularity local feature representation is obtained by dividing the image into multiple strips and changing the number of strips of different local branches; first, the backbone network of the multi-branch network uses the ResNet-50 convolution structure, and takes the first four groups of convolution layers to perform preliminary feature extraction on the gait data set. After the preliminary extraction, it is divided into three branches to extract features respectively. The first branch extracts global features, and the second and third branches extract local features. Among them, the second branch and the third branch divide the feature level into two and four parts respectively for feature learning. The present invention further uses a normalized loss function (softmax loss) and a triplet loss function (triplet loss) in the design of the loss function. loss), and the loss functions corresponding to each branch are different. The feature maps belonging to the triplet loss function include the global features in the first, second and third branches, and the feature maps belonging to the normalized loss function include the global features in the first, second and third branches, as well as the final blocks of the horizontal segmentation feature maps of the second and third branches; finally, the deep gait features output by the gait recognition model are compared with the preset personnel gait dataset to determine the identity information of the underground personnel.

[0022] In an embodiment of the present invention, the echo signal is subjected to fast Fourier transform processing to construct a gait data set containing the position information of the underground personnel; the gait data set containing the position information of the underground personnel is processed by CA-CFAR target detection to obtain the position information of the underground personnel, specifically including: The echo signal sent and received by the millimeter-wave radar also contains the location information of the personnel. Therefore, in the embodiment of the present invention, the echo signal in the echo signal is further subjected to two fast Fourier transforms in the range and Doppler directions to obtain the speed information of the underground personnel, and generate a range-speed spectrum diagram of the underground personnel, thereby constructing a gait data set containing the location information of the underground personnel. The clutter is suppressed by using a clutter suppression filter using a moving target display method, and then the background noise interference is further filtered out and the CA-CFAR target detection algorithm is used to obtain the location information of the personnel.

[0023] like Figure 3The figure shows a schematic diagram of the detection principle of one-dimensional CA-CFAR, which obtains a background noise level by calculating the signal statistical characteristics (such as the mean or median) within a reference window, and sets a fixed false alarm probability based on this level. Then, after the target signal passes a certain distance, a window is used to detect the existence of the target signal, and the signal in the window is compared with the background noise level. If the strength of the signal in the window exceeds the threshold value corresponding to the set false alarm probability, it is determined that there is a target signal in the window, otherwise it is considered that there is only background noise in the window. The judgment criteria of this detection principle are: ; When the detection value Greater than threshold , then it is judged as the target, Less than value When it is judged as noise, the threshold can be calculated by the following formula: ; Where N represents the number of adjacent cells, represents the false alarm probability.

[0024] Furthermore, the azimuth angle refers to the angle between the direction in which the radar antenna points to the target and the direction in which the radar is located. The millimeter-wave radar system usually contains two or more receiving antennas to receive the signals emitted by the target and calculate the azimuth angle of the target by measuring the phase difference between these signals. Figure 4 In the two receiving antenna configurations shown, when the distance between the target and the radar is far enough and the angle is small, the echo signals of the two solid lines can be approximately considered to be parallel. is the receiving antenna separation distance, is the target arrival angle, then the phase difference for: ; in, The wavelength is the target's location, and the target's position can be determined quickly and accurately by calculating the target's angle.

[0025] In the embodiment of the present invention, information is read based on the preset UWB tag on the downhole equipment to obtain the identity information and location information corresponding to the equipment, specifically: In the embodiment of the present invention, UWB tags are pre-fixed on the downhole equipment. Each tag has its own unique ID, so the downhole equipment to be located can be linked to the tag through the ID. UWB positioning uses TOA ranging technology. TOA (Time of Arrival) ranging technology is a time-based distance measurement method used to determine the propagation time of the signal from the transmitter to the receiver, and calculate the distance of the signal through the propagation time. Its principle is as follows: Figure 5 As shown, it can be applied to different communication systems and positioning systems, such as wireless communication, radar and positioning systems, Figure 5 The distances between the three base stations BS and the MS are r1, r2, and r3 respectively. Three circles are drawn with the distances measured by each base station as the center and the radius as the radius. The intersection of the circles is the location of the MS. When the three base stations are LOS base stations, the estimated location of the MS can generally be calculated based on the least squares (LS) algorithm. Assume that the location coordinates of the MS are (x, y) and the location coordinates of the N BSs are (x i ,y i ), according to their geometric meaning, the relationship between them is: ; In this embodiment, consistency judgment is performed based on the identity information, location information of the underground personnel and the identity information and location information of the equipment. Specifically, after using millimeter wave radar and UWB technology to obtain the information of personnel and equipment, the identity information is compared first. If they are consistent, then calculate whether the position difference of the personnel and equipment is less than a preset threshold. If it is less than the threshold, the personnel and equipment are consistent, and consistency monitoring of personnel equipment and personnel areas can be achieved.

[0026] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for monitoring the consistency of underground personnel equipment based on millimeter wave and UWB, characterized in that: include: The millimeter-wave radar sends radar signals to underground personnel for information collection and obtains their echo signals; Performing short-time Fourier transform processing on the echo signal to construct a gait data set containing identity information of underground personnel; A gait recognition model is constructed using a pre-trained multi-branch network, and a gait data set containing identity information of underground personnel is input into the gait recognition model to obtain the identity information of the underground personnel; Performing fast Fourier transform processing on the echo signal to construct a gait data set containing underground personnel distance information; Using CA-CFAR target detection to process the gait data set containing the distance information of the underground personnel, and obtain the position information of the underground personnel; Read information based on the preset UWB tags on the underground equipment to obtain the identity information and location information corresponding to the equipment; The consistency is judged based on the identity information and location information of the underground personnel and the identity information and location information of the equipment.

2. The method for monitoring the consistency of underground personnel equipment based on millimeter wave and UWB according to claim 1 is characterized in that: The echo signal includes gait information and distance information of the underground personnel.

3. The method for monitoring the consistency of underground personnel equipment based on millimeter wave and UWB according to claim 2 is characterized in that: The gait information includes gait information of underground personnel at multiple walking speeds; the multiple walking speeds include slow speed, normal speed and fast speed.

4. The method for monitoring the consistency of underground personnel equipment based on millimeter wave and UWB according to claim 1 is characterized in that: A gait recognition model is constructed using a pre-trained multi-branch network, and a gait data set containing the identity information of underground personnel is input into the gait recognition model to obtain the identity information of underground personnel, including: Inputting the gait data set into a multi-branch network to perform preliminary feature extraction to obtain shallow gait features; The preliminary features are subjected to deep feature extraction by multiple branches in a multi-branch network to obtain deep gait features; The deep gait feature is compared with a preset personnel gait data set to determine the identity information of the underground personnel.

5. The method for monitoring the consistency of underground personnel equipment based on millimeter wave and UWB according to claim 1, characterized in that: The multi-branch network includes: a residual convolution layer, a first branch, a second branch and a third branch; The residual convolution layer is used to preliminarily extract the gait data set containing the identity information of underground personnel to obtain shallow gait features; The first branch is used to extract global features of shallow gait features; The second branch divides the shallow gait feature level into two parts for local feature extraction; The third branch divides the shallow gait feature level into four parts for local feature extraction.

6. The method for monitoring the consistency of underground personnel equipment based on millimeter wave and UWB according to claim 2 is characterized in that: Obtain the location information of underground personnel, including: The CA-CFAR target detection algorithm is used to detect underground personnel in a gait data set containing the position information of underground personnel, and an angle measurement algorithm is used to measure the angle between the corresponding echo signal in the gait data set containing the position information of underground personnel and the position of the radar, and the position of the underground personnel corresponding to the echo signal is determined based on the angle.

7. The method for monitoring the consistency of underground personnel equipment based on millimeter wave and UWB according to claim 2, characterized in that: Information is read based on the preset UWB tags on the downhole equipment, specifically: Use TOA ranging method to perform UWB positioning on underground equipment and obtain the location information of underground equipment; The identity information of the downhole equipment is determined according to the UWB tag corresponding to the downhole equipment.

Images