阅读说明：本技术 一种基于毫米波雷达的智能消防安全监控系统及监控方法 (Intelligent fire safety monitoring system and monitoring method based on millimeter wave radar ) 是由 韩俊峰 于 2020-07-29 设计创作，主要内容包括：一种基于毫米波雷达的智能消防安全监控系统及监控方法,监控系统包括智能火灾预警及救援辅助模块、无线收发设备、云端服务器和控制终端,智能火灾预警及救援辅助模块包括毫米波雷达检测跟踪装置、烟雾检测装置、温度检测装置、洒水控制器、电源装置、报警装置；当火灾发生时,毫米波雷达检测跟踪装置探测用户位置信息和状态信息并通过无线收发设备发送给云端服务器,云端服务器记录信息并判断现场人员状况发送给控制终端,控制终端能够显示室内人数、火灾发生位置、人员运动轨迹及人员健康状态。本发明能够在火灾发生时有效监测室内人员的数量以及分布情况,跟踪室内人员的运动轨迹,检查室内人员的呼吸体征,同时不会涉及到用户的个人隐私。(An intelligent fire safety monitoring system and a monitoring method based on millimeter wave radar are disclosed, wherein the monitoring system comprises an intelligent fire early warning and rescue auxiliary module, wireless transceiving equipment, a cloud server and a control terminal, and the intelligent fire early warning and rescue auxiliary module comprises a millimeter wave radar detection tracking device, a smoke detection device, a temperature detection device, a water spray controller, a power supply device and an alarm device; when a fire disaster occurs, the millimeter wave radar detection tracking device detects user position information and state information and sends the user position information and the state information to the cloud server through the wireless receiving and sending equipment, the cloud server records the information and judges the situation of field personnel and sends the situation to the control terminal, and the control terminal can display the number of indoor people, the position where the fire disaster occurs, the movement track of the personnel and the health state of the personnel. The invention can effectively monitor the number and distribution condition of indoor personnel when a fire disaster happens, track the motion track of the indoor personnel, check the breathing signs of the indoor personnel, and simultaneously, the invention can not relate to the individual privacy of users.)

1. The utility model provides an intelligence fire control safety monitored control system based on millimeter wave radar which characterized in that: the intelligent fire early warning and rescue auxiliary module comprises a millimeter wave radar detection tracking device, a smoke detection device, a temperature detection device, a water sprinkling controller, a power supply device and an alarm device; when a fire disaster occurs, the smoke detection device and the temperature detection device detect abnormal information, the alarm device gives an alarm, the sprinkling controller controls field sprinkling equipment to put out a fire in an emergency, the millimeter wave radar detection tracking device detects user position information and state information and sends the user position information and the state information to the cloud server through the wireless transceiving equipment, the cloud server records the information and judges the situation of field personnel and sends the situation to the control terminal, and the control terminal can display the number of indoor people, the fire disaster occurrence position, the personnel movement track and the personnel health state; and the power supply device is used for supplying power to each device of the intelligent fire early warning and rescue auxiliary module.

2. An intelligent fire safety monitoring system based on millimeter wave radar according to claim 1, characterized in that: the intelligent fire early warning and rescue auxiliary modules are distributed in different areas, the cloud server can realize cross-area remote interconnection through wireless transceiver equipment, and the wireless transceiver equipment adopts NB-IOT-based wireless transceiver equipment.

3. An intelligent fire safety monitoring system based on millimeter wave radar according to claim 1, characterized in that: the radar adopted by the millimeter wave radar detection tracking device is a 76G-81G millimeter wave radar.

4. An intelligent fire safety monitoring system based on millimeter wave radar as claimed in claim 1, wherein the intelligent fire early warning and rescue auxiliary module is started in the following three ways:

the active start is controlled by a control terminal;

the passive start is triggered and started through a smoke detection device and a temperature detection device;

and (4) spontaneous starting, wherein abnormal conditions are found in the self-checking process and the starting is carried out.

5. An intelligent fire safety monitoring system based on millimeter wave radar according to claim 4, characterized in that: and the self-checking time and frequency of the spontaneous starting are set through the control terminal.

6. An intelligent fire safety monitoring system based on millimeter wave radar according to claim 1, characterized in that: the power supply device sets the threshold electric quantity to be 10%, and when the electric quantity is lower than the threshold electric quantity, the charging is automatically carried out.

7. The monitoring method of the millimeter wave radar-based intelligent fire safety monitoring system according to any one of claims 1 to 6, characterized by comprising the following steps:

step 1: environmental parameters are detected through a smoke detection device and a temperature detection device, when a fire disaster happens, an alarm device gives an alarm, and a sprinkling controller controls field sprinkling equipment to emergently extinguish the fire; detecting user position information and state information by a millimeter wave radar detection tracking device;

step 2: when a user is located indoors, the millimeter wave radar detection tracking device sends detected position information and state information of the user to a cloud server through wireless receiving and sending equipment, and the cloud server monitors fire conditions of multiple areas in real time;

and step 3: after receiving the information sent by the intelligent fire early warning and rescue auxiliary module, the cloud server records all information in real time and analyzes the breathing information of the personnel so as to judge the physical condition of the personnel;

and 4, step 4: the cloud server sends the data to the control terminal, the control terminal displays the number of people in the room, the position of the fire, the movement track of people and the health state information of the people, and escape path planning and guiding broadcasting are carried out according to the information;

and 5: the fire fighter obtains real-time fire data according to control terminal information, realizes accurate fire extinguishing and accurate rescue.

8. The monitoring method according to claim 7, characterized in that: and 3, the cloud-end server judges the physical condition of the personnel according to whether the human breath is in the normal frequency range or not, if the breath frequency is in the normal range, the personnel state is judged to be good, and if the breath frequency exceeds the normal range, the personnel state is judged to be dangerous.

9. The monitoring method according to claim 7, characterized in that: the process of detecting the user position information and the state information by the millimeter wave radar detection tracking device comprises target detection and target tracking;

the target detection is based on a capon algorithm, a parameter measurement process from distance measurement to angle measurement and then to speed measurement is adopted, and the number, distance, angle, Doppler information and signal-to-noise ratio of detected targets are obtained through a constant false alarm detection method and are used for target tracking; the target tracking adopts a group target tracking algorithm, and comprises four steps of prediction, joint distribution, track initiation and updating maintenance.

10. The monitoring method according to claim 9, characterized in that:

the target detection steps are as follows:

ranging: the receiving signal and the sending signal are subjected to difference frequency by using a frequency mixing circuit to obtain an intermediate frequency signal, and distance information is obtained by using a distance formula, wherein the specific formula is as follows:

in the formula f_IFIntermediate frequency, C light speed, S frequency modulation slope and R distance; performing AD sampling on the intermediate frequency signal, and performing one-dimensional FFT on a data matrix obtained by sampling in the distance direction to obtain a one-dimensional data matrix 1DFFT containing distance information;

angle measurement: the angle measurement utilizes capon algorithm to receive the power of the signal in each direction, and the expression is as follows:

wherein R is X^H，X＝{x₁,x₂,x₃,x₄…x_n}^TA vector of signal components representing the target at different antennas, α (θ) {1, e }^{i2πdcos(θ)×2},e^{i2πdcos(θ)×3},…,e^{i2πdcos(θ)×(N-1)}}^TIs a guide vector; t is transposition operation;

searching the power of each angle for a peak value to determine the angle;

the operation is carried out on the one-dimensional data matrix 1DFFT to obtain a distance and azimuth thermodynamic diagram rangeAzimuth HeatMap which contains distance and angle information, and the distance and azimuth thermodynamic diagram rangeAzimuth HeatMap is respectively subjected to CFAR detection in the distance direction and the azimuth direction, so that the detection of a plurality of targets and the measurement of the distance and the angle are realized;

speed measurement: the calculation is made according to the following formula:

wherein V is velocity, λ is wavelength,for change of angle, T_cIs a pulse period;

the target tracking steps are as follows:

and (3) prediction: estimating the tracking group centroid at the time n by utilizing an extended Kalman filtering prediction process based on the state at the time n-1 and a process covariance matrix; the state and covariance prediction equations for the kalman filter algorithm are as follows:

s_apr(n)＝Fs(n-1)

P_apr(n)＝FP(n-1)F^T+Q(n-1)

wherein s is_apr(n) is a prior value of the state, P_apr(n) is the prior value of covariance, F is the state transition matrix; q is the state noise covariance, s is the target state, and P is the covariance;

association and allocation: assuming that there are one or more trajectories and associated predicted state vectors, for each given trajectory, forming a gate with respect to the predicted centroid, building an ellipsoid in a three-dimensional measurement space tracking the cluster centroid using the cluster residual covariance matrix, the ellipsoid representing a gating function to define a single measurement observed at time n, for measurements within the gate, computing a normalized distance function as a cost function associating the measurement with each trajectory;

starting a flight path: for any track independent metrics, a new group tracker will be assigned and initialized;

updating and maintaining: when the measurement at time n is available, the state and error covariance estimates are updated as follows:

a) the measurement residue was calculated as follows:

y(n)＝u(n)-H(s_apr(n))；

in the formula, H is a measurement matrix;

b) the innovation covariance is calculated as follows:

C(n)＝H(s_apr(n))P_apr(n)H^T(s_apr(n))+R(n)

c) the kalman gain is calculated as follows:

wherein inv [ ] is an inversion operation;

d) the posterior state vector is calculated as follows:

s(n)＝s_apr(n)+K(n)y(n)

e) calculating the posterior covariance:

P(n)＝P_apr(n)-K(n)s_apr(n)inv(K(n))

each trace goes through the life cycle of an event, changing state or deleting traces that are no longer in use during the maintenance phase.

Technical Field

The invention belongs to the field of indoor positioning and fire fighting, and particularly relates to an intelligent fire fighting safety monitoring system and method based on a millimeter wave radar.

Background

The frequency range of the millimeter wave is between microwave and infrared, and the millimeter wave radar has the advantages of both the microwave radar and the photoelectric radar. Meanwhile, the millimeter wave radar has the characteristics of small volume, light weight and strong anti-interference capability. Under the condition that the conflagration took place, the working ability of millimeter wave radar in high temperature and smog environment is extremely strong, consequently, the millimeter wave radar can make statistics of the number information in the room and room personnel's breathing sign well, is favorable to helping the fire fighter to carry out rescue work.

Up to now, the technology of indoor intelligent monitoring is still not mature enough. Most people do not install a surveillance camera indoors because of the personal privacy of the indoor space. And the millimeter wave radar can well protect personal privacy and accurately and conveniently count various information of indoor personnel. In most cases, when a fire disaster occurs, people mostly find the fire and give an alarm at the first time, and then the fire alarm center gives an alarm according to information provided by an alarm person. However, people often cannot accurately grasp the specific occurrence of fire in the alarming process, so that fire-fighting units cannot accurately extinguish fire according to the fire. In addition, if the fire disaster happens, the fire disaster often spreads and brings serious loss if the fire disaster happens without the on-site unattended operation.

The safe city is a large-scale and highly comprehensive management system which is developed at present, and relates to various aspects in city management, in particular to the field of emergency treatment of disaster accidents. Under the background of the times, it is very important to establish an intelligent fire safety monitoring system, and the system needs to be capable of carrying out fire monitoring and early warning in a certain area and can be connected through a cloud platform, so that the fire occurrence conditions in the whole country can be monitored in real time.

Disclosure of Invention

The invention aims to provide an intelligent fire safety monitoring system and a monitoring method based on a millimeter wave radar aiming at the problem that the number and the distribution condition of indoor personnel cannot be accurately monitored in the prior art, so that fire alarm and field environment monitoring can be carried out, and fire fighters can remotely acquire real-time data through terminals to help realize accurate fire extinguishing and personnel rescue.

In order to achieve the purpose, the invention has the following technical scheme:

an intelligent fire safety monitoring system based on a millimeter wave radar comprises an intelligent fire early warning and rescue auxiliary module, a wireless transceiver, a cloud server and a control terminal, wherein the intelligent fire early warning and rescue auxiliary module comprises a millimeter wave radar detection tracking device, a smoke detection device, a temperature detection device, a water spray controller, a power supply device and an alarm device; when a fire disaster occurs, the smoke detection device and the temperature detection device detect abnormal information, the alarm device gives an alarm, the sprinkling controller controls field sprinkling equipment to put out a fire in an emergency, the millimeter wave radar detection tracking device detects user position information and state information and sends the user position information and the state information to the cloud server through the wireless transceiving equipment, the cloud server records the information and judges the situation of field personnel and sends the situation to the control terminal, and the control terminal can display the number of indoor people, the fire disaster occurrence position, the personnel movement track and the personnel health state; and the power supply device is used for supplying power to each device of the intelligent fire early warning and rescue auxiliary module.

Preferably, the intelligent fire early warning and rescue auxiliary modules are distributed in different areas, the cloud server can realize cross-area remote interconnection through wireless transceiver equipment, and the wireless transceiver equipment adopts NB-IOT-based wireless transceiver equipment.

Preferably, the radar used by the millimeter wave radar detection and tracking device is a 76G to 81G millimeter wave radar.

Preferably, the intelligent fire early warning and rescue auxiliary module has the following three starting modes:

the active start is controlled by a control terminal;

the passive start is triggered and started through a smoke detection device and a temperature detection device;

and (4) spontaneous starting, wherein abnormal conditions are found in the self-checking process and the starting is carried out.

Preferably, the self-checking time and frequency of the spontaneous start are set by a control terminal.

Preferably, the power supply device sets the threshold electric quantity to be 10%, and automatically charges when the electric quantity is lower than the threshold electric quantity.

The invention relates to a monitoring method of an intelligent fire safety monitoring system based on a millimeter wave radar, which comprises the following steps:

step 1: environmental parameters are detected through a smoke detection device and a temperature detection device, when a fire disaster happens, an alarm device gives an alarm, and a sprinkling controller controls field sprinkling equipment to emergently extinguish the fire; detecting user position information and state information by a millimeter wave radar detection tracking device;

step 2: when a user is located indoors, the millimeter wave radar detection tracking device sends detected position information and state information of the user to a cloud server through wireless receiving and sending equipment, and the cloud server monitors fire conditions of multiple areas in real time;

and step 3: after receiving the information sent by the intelligent fire early warning and rescue auxiliary module, the cloud server records all information in real time and analyzes the breathing information of the personnel so as to judge the physical condition of the personnel;

and 4, step 4: the cloud server sends the data to the control terminal, the control terminal displays the number of people in the room, the position of the fire, the movement track of people and the health state information of the people, and escape path planning and guiding broadcasting are carried out according to the information;

and 5: the fire fighter obtains real-time fire data according to control terminal information, realizes accurate fire extinguishing and accurate rescue.

Preferably, in the step 3, the cloud-end server judges the physical condition of the person according to whether the human breath is in the normal frequency range, if the breath frequency is in the normal range, the state of the person is judged to be good, and if the breath frequency exceeds the normal range, the state of the person is judged to be dangerous.

Preferably, the process of detecting the user position information and the state information by the millimeter wave radar detection tracking device comprises target detection and target tracking;

the target detection is based on a capon algorithm, a parameter measurement process from distance measurement to angle measurement and then to speed measurement is adopted, and the number, distance, angle, Doppler information and signal-to-noise ratio of detected targets are obtained through a constant false alarm detection method and are used for target tracking; the target tracking adopts a group target tracking algorithm, and comprises four steps of prediction, joint distribution, track initiation and updating maintenance.

Preferably, the target detection step is as follows:

ranging: the receiving signal and the sending signal are subjected to difference frequency by using a frequency mixing circuit to obtain an intermediate frequency signal, and distance information is obtained by using a distance formula, wherein the specific formula is as follows:

in the formula f_IFIntermediate frequency, C light speed, S frequency modulation slope and R distance; performing AD sampling on the intermediate frequency signal, and performing one-dimensional FFT on a data matrix obtained by sampling in the distance direction to obtain a one-dimensional data matrix 1DFFT containing distance information;

angle measurement: the angle measurement utilizes capon algorithm to receive the power of the signal in each direction, and the expression is as follows:

wherein R is X^H，X＝{x₁,x₂,x₃,x₄…x_n}^TA vector of signal components representing the target at different antennas, α (θ) {1, e }^{i2πdcos(θ)×2},e^{i2πdcos(θ)×3},…,e^{i2πdcos(θ)×(N-1)}}^TIs a guide vector; t is transposition operation;

searching the power of each angle for a peak value to determine the angle;

the operation is carried out on the one-dimensional data matrix 1DFFT to obtain a distance and azimuth thermodynamic diagram rangeAzimuth HeatMap which contains distance and angle information, and the distance and azimuth thermodynamic diagram rangeAzimuth HeatMap is respectively subjected to CFAR detection in the distance direction and the azimuth direction, so that the detection of a plurality of targets and the measurement of the distance and the angle are realized;

speed measurement: the calculation is made according to the following formula:

wherein V is velocity, λ is wavelength,

for change of angle, T_cIs a pulse period;

the target tracking steps are as follows:

and (3) prediction: estimating the tracking group centroid at the time n by utilizing an extended Kalman filtering prediction process based on the state at the time n-1 and a process covariance matrix; the state and covariance prediction equations for the kalman filter algorithm are as follows:

s_apr(n)＝Fs(n-1)

P_apr(n)＝FP(n-1)F^T+Q(n-1)

wherein s is_apr(n) is a prior value of the state, P_apr(n) is the prior value of covariance, F is the state transition matrix; q is the state noise covariance, s is the target state, and P is the covariance;

association and allocation: assuming that there are one or more trajectories and associated predicted state vectors, for each given trajectory, forming a gate with respect to the predicted centroid, building an ellipsoid in a three-dimensional measurement space tracking the cluster centroid using the cluster residual covariance matrix, the ellipsoid representing a gating function to define a single measurement observed at time n, for measurements within the gate, computing a normalized distance function as a cost function associating the measurement with each trajectory;

starting a flight path: for any track independent metrics, a new group tracker will be assigned and initialized;

updating and maintaining: when the measurement at time n is available, the state and error covariance estimates are updated as follows:

a) the measurement residue was calculated as follows:

y(n)＝u(n)-H(s_apr(n))；

in the formula, H is a measurement matrix;

b) the innovation covariance is calculated as follows:

C(n)＝H(s_apr(n))P_apr(n)H^T(s_apr(n))+R(n)

c) the kalman gain is calculated as follows:

wherein inv [ ] is an inversion operation;

d) the posterior state vector is calculated as follows:

s(n)＝s_apr(n)+K(n)y(n)

e) calculating the posterior covariance:

P(n)＝P_apr(n)-K(n)s_apr(n)inv(K(n))

each trace goes through the life cycle of an event, changing state or deleting traces that are no longer in use during the maintenance phase.

Compared with the prior art, the invention has the following beneficial effects: when the conflagration takes place, detect indoor user's positional information and state information through millimeter wave radar detection tracking means, the millimeter wave radar interference killing feature who uses is strong, and the electromagnetic wave of transmission can pass smog well, can be under the condition that the conflagration takes place, effectively monitor indoor personnel's quantity and distribution condition, trail indoor personnel's movement track, and the breathing sign of inspection indoor personnel can not involve user's individual privacy simultaneously. The watering controller that sets up among intelligence fire early warning and the rescue auxiliary module can control on-the-spot watering equipment emergency fire extinguishing, slows down the condition of a fire and enlarges, and alarm device can remind the user that unknown condition of a fire takes place well, and help people flee from the fire area very first time. The millimeter wave radar detection tracking device is connected with the cloud server through the wireless transceiving equipment, the cloud server records information and judges the status of field personnel and sends the status to the control terminal, and firefighters can observe the room location where a fire disaster occurs, the number of the personnel in the room and vital signs in the room in real time through the control terminal. The wireless transceiving equipment and the cloud server are interconnected, so that the fire occurrence condition in the whole country can be monitored in real time, and the construction of a safe city is promoted. The fire fighter can obtain these real-time data from control terminal, and then help the fire fighter carry out accurate fire extinguishing and personnel rescue, has fine application prospect.

Drawings

FIG. 1 is a schematic diagram of a system architecture for building applications of the present invention;

FIG. 2 is a block diagram of the intelligent fire warning and rescue aid module of the present invention;

FIG. 3 is a block diagram of the millimeter wave radar detection tracking device of the present invention;

FIG. 4 is a flow chart of a monitoring method according to an embodiment of the present invention;

FIG. 5 is a flow chart of a method for detecting an object according to the present invention;

FIG. 6 is a flow chart of a target tracking method of the present invention;

FIG. 7 is a flow chart of a method for monitoring respiratory information according to the present invention.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and examples.

Referring to fig. 1, the intelligent fire safety monitoring system based on the millimeter wave radar of the invention comprises an intelligent fire early warning and rescue auxiliary module, a wireless transceiver, a cloud server and a control terminal. The real-time positioning monitoring of the indoor environment is achieved under the condition that the privacy of the user is protected, the alarm can be timely given to assist in rescue when a fire occurs, and loss is reduced.

Taking the system applied to the building as an example, the intelligent fire early warning and rescue auxiliary module is installed in the center of the ceiling of a room, and comprises a millimeter wave radar detection tracking device for detecting the number of people and the movement track of indoor personnel as shown in fig. 2: the smoke detection device and the temperature detection device are used for monitoring the occurrence of fire; the power supply device is used for supplying power to the equipment; and the alarm device is used for giving an alarm when a fire occurs. The intelligent fire alarm system has the functions of fire alarm, personnel monitoring, data transmission, sprinkling control and the like.

The control terminal comprises a computer terminal and a mobile phone terminal.

As shown in fig. 3, the millimeter wave radar detection tracking device of the present invention includes two transmitting and three receiving antennas, and a 76-81G millimeter wave radar chip: RC7701N32, single chip: STM32H743 × IH6U, SDRAM: IS42S 32800G-6BLI, high speed USB transceiver: USB3320C-EZK, UART to USB: CP2102-GM, USB connector: 47590-0001.

The wireless transceiver device adopts NB-IOT technology for connecting the intelligent fire early warning and rescue auxiliary device and the cloud platform, and the NB-IOT technology supports the connection of low-power consumption devices in a wide area network, so that very comprehensive data coverage can be provided.

After the cloud server receives the signals sent by the devices, a database is established to match the ID of each device. The background control terminal can acquire the position information of the fire room and the related information of the personnel by downloading the cloud data.

The invention can provide the following support:

1. indoor personnel positioning and trajectory tracking;

as shown in fig. 5, in the target detection process of the present invention, a parameter measurement procedure of distance measurement, angle measurement, and speed measurement is adopted, a capon algorithm is adopted to improve the accuracy of angle measurement, and finally, information such as the number of targets to be detected, distance, angle, doppler information, signal-to-noise ratio, and the like, obtained through CFAR detection is transmitted to a target tracking module for the next tracking process. The specific implementation process is as follows:

regarding ranging: the receiving signal and the sending signal are subjected to difference frequency by using a frequency mixing circuit to obtain an intermediate frequency signal, and distance information can be obtained by using a distance formula, wherein the specific formula is as follows:

wherein f is obtained in this way_IFThen, R is obtained. And then, AD sampling is carried out on the intermediate frequency signal, one-dimensional FFT is carried out on a data matrix obtained by sampling in the distance direction to obtain 1DFFT, and a final distance formula is obtained, wherein the specific formula is as follows:

wherein k is_rFor the value of K, F corresponding to the spectral peak in the FFT spectrogram_SAMPIs the sampling frequency. The distance resolution formula is:

regarding angle measurement: the angle measurement mainly utilizes a capon algorithm, a capon beam former (namely a minimum variance distortionless response beam former) forms a beam in a desired observation direction by using part of the degrees of freedom, and forms a null in an interference direction by using the rest of the degrees of freedom. When there are multiple signals incident on the sensor array, the array output power will include the desired signal power and the power of the interfering signal. capon's least variance approach minimizes the output power of the interfering signal to suppress the interfering signal while keeping the gain constant in the observation direction (usually assuming this constant is 1).

The power expression of the received signal in each direction can be expressed as:

wherein R is X^T,X＝{x₁,x₂,x₃,x₄…x_n}^TA vector of signal components representing the target at different antennas, α (θ) {1, e }^{i2πdcos(θ)×2},e^{i2πdcos(θ)×3},…,e^{i2πdcos(θ)×(N-1)}}^TIs a steering vector.

And searching the power of each angle for a peak value, so as to determine the angle.

The 1DFFT carries out the operation to obtain a matrix rangeAzimuth HeatMap, and CFAR detection is carried out on the rangeAzimuth HeatMap in the direction and the azimuth direction respectively, so that detection of a plurality of targets and measurement of parameters such as distance and angle are realized.

Regarding speed measurement: the phase difference of the same target intermediate frequency signal between two adjacent chirp signals is as follows:

it can be seen from the above equation that the phase is very sensitive to small changes in distance. And the target speed can be calculated therefrom:

therefore, the targets at the same distance and different speeds can be distinguished only by performing FFT on the targets at the same distance in the chirp direction. The specific operation is as follows: and performing beam synthesis on the 1DFFT in the antenna direction, wherein the weight value is calculated to P (theta), and performing FFT in the chirp direction to obtain a 2DFFT matrix. The velocity resolution needs to be noted as:

the target detection part can obtain information such as the number, the distance, the angle, the Doppler and the like of the detected target, the information data is transmitted to the target tracking module, and the target tracking module is used for completing the tracking of the target.

As shown in fig. 6, when the target tracking is performed, the tracking algorithm mainly uses a group target tracking algorithm, and the specific implementation steps mainly include four parts, namely prediction, association and allocation, track initiation, and update and maintenance:

with respect to the prediction: and estimating the tracking group centroid at the time n by utilizing a Kalman filtering prediction process based on the state at the time n-1 and the process covariance matrix. The a priori state and covariance estimates for each trackable object are calculated, while the metrology vector estimates are also calculated at this step. The state and covariance prediction equations for the kalman filter algorithm are as follows:

s_apr(n)＝Fs(n-1)

P_apr(n)＝FP(n-1)F^T+Q(n-1)

wherein s is_apr(n) is a prior value of the state, P_apr(n) is the prior value of the covariance and F is the state transition matrix.

Regarding association and allocation: the combined part mainly comprises a door mechanism and a scoring mechanism. It is assumed that one or more trajectories and associated prediction state vectors exist. For each given track, a gate is formed with respect to the predicted centroid. The gate should consider target motion, population dispersion and measurement noise, utilize the group residual covariance matrix to establish an ellipsoid in the three-dimensional measurement space tracking the group centroid, the ellipsoid represents a gating function to define the single measurement value observed at n moments, for the measurement values in the gate, calculate a normalized distance function as a cost function, and associate the measurement value with each orbit.

With respect to track initiation: for metrics unrelated to any tracks (located outside of any existing gates), a new group tracker will be assigned and initialized. First a dominant measurement point is selected and a centroid is set. The radial velocities of the other candidate points are spread out using the radial velocities of the leading points. Each time it is first checked whether one candidate point is within the speed range (speed check, followed by distance check). If so, the centroid is recalculated and the point is added to the cluster. Some pass tests are then performed for the cluster. A minimum number, a strong enough combined signal-to-noise ratio and a minimum amount of centroid dynamics need to be measured. If by creating (assigning) a new tracking object and initializing the scatter matrix with the relevant points, clusters with fewer points will be ignored.

Regarding update maintenance: when the measured value at time n is available, the state and error covariance estimates will be updated in the following measurement update process:

calculating the measurement residue:

y(n)＝u(n)-H(s_apr(n))

calculating innovation covariance:

C(n)＝H(s_apr(n))P_apr(n)H^T(s_apr(n))+R(n)

calculating a Kalman gain:

calculating the posterior state vector:

s(n)＝s_apr(n)+K(n)y(n)

calculating the posterior covariance:

P(n)＝P_apr(n)-K(n)s_apr(n)inv(K(n))

each trajectory goes through the life cycle of an event.

During the maintenance phase, the state may be changed or the traces that are no longer in use may be deleted.

And tracking the target by the four parts.

In the project implementation, a constant acceleration model is adopted, and a state vector under a cartesian coordinate system at a time n is as follows:

the transition matrix is:

the measurement vector is:

the relationship between the state vector and the metrology vector can be expressed as:

u(n)＝H(s(n))+v(n)

wherein:

to linearize the relationship between the state vector and the measurement vector, the bias matrix for H (s (n)) is calculated as follows:

the relationship between the measurement vector and the state vector can be expressed as:

u(n)＝J_H(s)s(n)+v(n)

at this time, a linear relationship is obtained, and iteration can be performed by using the above four parts, so as to finally realize the tracking of the target. The data are transmitted back to the control terminal through the cloud server, so that the motion trail of the user can be clearly and visually obtained.

2. Monitoring human body breathing information;

the millimeter wave radar detection tracking device has two modes, wherein the first mode is the target detection and tracking mode, when the target is detected to be stationary for a long time, a target user is suspected to be coma possibly caused by factors such as smoke, temperature and the like, the millimeter wave radar starts a second working mode, namely vital sign monitoring, so that the respiratory characteristics of a human body are mainly tested, respiratory information is transmitted back to an upper computer through a cloud end in real time, and the background can acquire the vital signs of the user in real time.

Firstly, the phase difference between two chirp at two adjacent same positions is as follows:

it can be seen that the moving distance of the object is linear with the phase change. Therefore, the idea of breathing measurement for the millimeter wave radar is as follows: the fluctuation and vibration of the chest are detected by the radar, the phase change of the reflected signal at the position with the same distance is actually detected, and the distance change is deduced in a reverse way, and the specific flow of the algorithm is shown in fig. 7.

The frequency of breathing is then derived by the following processing algorithm. The specific implementation is as follows:

one second will transmit 20 detection frames, each frame period is 50 ms, each ms has only two chirp, that is to say, the first and the second chirp, during detection, only the first chirp is used first, then in each frame, after the first chirp data is obtained, an FFT is made first to obtain 1DFFT, a unit of the same distance is analyzed in the frame gap time, on the unit of the same distance, the corresponding phase change is detected, the phase change is detected, thereby the distance change is obtained, and the fluctuation change of the thoracic cavity is obtained, because the heartbeat and the respiration frequency are different, two corresponding band pass filters are used to filter out the contained information and then the corresponding detection algorithm (such as FFT) is used to analyze the heartbeat frequency respiration frequency, and then the heartbeat and the respiration of the person can be obtained through the following calculation, because the detection accuracy of the heartbeat is not very high in a fire scene, the invention only needs to collect the respiratory information.

The following is illustrated by way of example:

in the millimeter wave radar-based intelligent fire safety monitoring system, the fire control monitoring of the whole building room is integrally monitored through a background control terminal, the intelligent fire early warning and rescue auxiliary module in each room is installed in the center of the ceiling of the room, has an independent ID number matched with the room, and is connected with a cloud server through an NB-IOT wireless transceiver.

As shown in fig. 4, when a fire in a room occurs, the smoke sensor detects smoke dust in the air, the temperature sensor detects that the temperature is too high, at the moment, the system sends out an alarm to start emergency measures for sprinkling to extinguish the fire, and sends a starting command to the millimeter wave radar detection tracking device, the device starts to detect the number of people in the room and the movement track of the fire after being started, meanwhile, the breathing vital signs of the people in the room are detected, the obtained data are sent to the cloud through the NB-IOT, and then the position of the fire room, the number of trapped people and the state information of the people can be observed on the control terminal.

The background control terminal can obtain real-time data from the cloud server, and simultaneously controls devices in adjacent rooms to start to remind people in the rooms near a fire to evacuate; the terminal can plan an escape path for the trapped people through the broadcasting system, and an escape strategy is provided. During fire rescue, the firefighters download cloud data through the background control terminal to obtain the distribution, vital signs and other conditions of trapped people, and take corresponding rescue measures. When a serious fire disaster occurs, the rescue time may be prolonged, which requires a longer working time of the millimeter wave radar detection tracking device, so that the threshold value for setting the automatic charging of the device cannot be too low, the working time of the threshold value electric quantity is ensured to be more than 1 hour, and therefore, the threshold value electric quantity should be set to be more than 10%.

The millimeter wave radar adopted by the invention has strong anti-interference capability, the emitted electromagnetic waves can well penetrate smoke, the quantity and distribution condition of indoor personnel can be effectively monitored under the condition of fire, the motion trail of the indoor personnel is tracked, the breathing signs of the indoor personnel are checked, and meanwhile, the individual privacy of a user cannot be involved. The alarm device and the water sprinkling device can well remind users of unknown fire, and help people escape from a fire area at the first time. The control terminal can observe the room location where the fire breaks out, the number of people in the room and vital signs in real time. The invention can not only monitor and early warn the fire, but also monitor the fire occurrence condition in the whole country in real time by the connection of the cloud platform, thereby being beneficial to the construction of safe cities. The fire fighter can obtain these real-time data from control terminal, carries out accurate fire extinguishing and personnel's rescue.

The embodiments described above are only a part of the embodiments of the present invention, and not all of them. The scope of the present invention includes, but is not limited to, the foregoing detailed description. Other embodiments based on the embodiments of the present invention, which can be made by those skilled in the art without inventive efforts, also belong to the scope of the present invention.