A underground space chamber security monitoring system

Abstract

The present invention discloses a cavern security monitoring system for underground spaces, which includes an intrusion monitoring unit, a fire sensing unit, a water hazard sensing unit, an electrical safety monitoring unit, an abnormality alarm unit, and a user data management and control unit; each monitoring unit and abnormality alarm unit of the system are respectively connected to the user data management and control unit in communication; after the user data management and control unit receives and stores the early warning signals sent by each monitoring unit, it sends an alarm signal to the abnormality alarm unit; after receiving the alarm signal, the abnormality alarm unit activates an audible and visual alarm, and pushes the alarm information to the security monitoring platform of the management party and the mobile terminal device carried by the user through the built-in Internet of Things module. The cavern security monitoring system for underground spaces of the present invention can overcome the problems existing in existing monitoring systems, realize the timely discovery and disposal of safety hazards and sudden safety incidents, and link and control multiple monitoring devices to eliminate potential safety risks.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent monitoring and security surveillance of underground spaces, and in particular to a security surveillance system for underground chambers. Background Art

[0002] Smart underground security uses sensing devices and the internet as a link, integrating artificial intelligence, network communications, security monitoring, and automatic control technologies to interconnect relevant underground facilities and build an efficient underground security management system. Currently, underground space development has brought enormous social and economic benefits, but underground space security management faces many challenges. There is an urgent need to integrate new technologies such as the Internet of Things, artificial intelligence, and security surveillance to improve underground space security, ensure overall safety management, and enhance security monitoring in key areas of underground space.

[0003] Existing underground spaces have few security measures. While some equipment chambers have security monitoring capabilities, their monitoring systems have numerous problems. For example, the entrances to equipment chambers are generally unsupervised, leading to door lock failures over time. Furthermore, equipment chambers lack intrusion detection capabilities. The convenience of security monitoring leads managers to overlook the aging of electrical equipment, resulting in over-operation and unnecessary fires or potential safety hazards. Existing security monitoring systems have simple safety protection measures and lack networking and overall safety risk analysis capabilities, making them unable to predict the risk of safety accidents. Safety hazards and emergency safety incidents within underground spaces cannot be promptly notified to remote managers or owners, preventing them from being eliminated.

[0004] In response to the problems existing in existing underground spaces, the present invention proposes an intelligent security monitoring system for underground chambers. The system uses functions such as intrusion detection, fire monitoring, and water hazard monitoring to achieve timely discovery and disposal of safety hazards and sudden safety incidents, and controls multiple monitoring devices in an interconnected manner to eliminate potential safety risks. At the same time, the system transmits the abnormal alarm signals sent by the monitoring equipment to the terminal device carried by the owner or the monitoring platform of the management party through the Internet of Things module, and by analyzing the historical data sent by multiple monitoring devices, it can predict safety hazards, thereby constructing an intelligent security monitoring system for underground chambers. Summary of the Invention

[0005] In response to the shortcomings and defects in the existing underground space chamber security monitoring technology, the present invention provides an underground space chamber intelligent security monitoring system, which utilizes new technologies such as the Internet of Things and artificial intelligence, combined with multi-source monitoring data fusion to achieve timely discovery and emergency response of safety hazards and sudden safety incidents, and realize intelligent security of underground spaces.

[0006] The present invention provides a cavern security monitoring system for an underground space, which includes: an intrusion monitoring unit, a fire sensing unit, a water hazard sensing unit, an electrical safety monitoring unit, a user data management and control unit, and an abnormal alarm unit; the intrusion monitoring unit, the fire sensing unit, the water hazard sensing unit, and the electrical safety monitoring unit all have built-in data networking units, and are respectively communicatively connected to the user data management and control unit; the fire sensing unit and the water hazard sensing unit are respectively communicatively connected to the electrical safety monitoring unit, and when the electrical safety monitoring unit receives the early warning signal sent by the fire sensing unit and the water hazard sensing unit through the data networking unit, it starts the power-off lockout protection; the abnormal alarm unit is communicatively connected to the user data management and control unit; after the user data management and control unit receives and stores the early warning signal sent by the intrusion monitoring unit, the fire sensing unit, the water hazard sensing unit, and the electrical safety monitoring unit, it sends an alarm signal to the abnormal alarm unit, and the user data management and control unit uses a self-learning strategy to analyze the signal data within a certain period of time to obtain the statistical characteristics of the signal data; after receiving the alarm signal, the abnormal alarm unit starts the sound and light alarm, and sends the alarm information to the security monitoring platform and the mobile terminal device through the built-in Internet of Things module.

[0007] Furthermore, the intrusion monitoring unit includes an infrared human monitoring module, a face recognition module, an abnormal sound monitoring module, an abnormal vibration monitoring module, and a personnel behavior analysis module; the infrared human monitoring module monitors whether there is a person approaching the chamber door through a built-in pyroelectric sensor. When the pyroelectric sensor detects that there is a person approaching the chamber door, the face recognition module, abnormal sound monitoring module, abnormal vibration monitoring module, and personnel posture analysis module in the intrusion monitoring unit start working.

[0008] Furthermore, the abnormal sound monitoring module collects sound signals near the chamber door through a built-in microphone, and the abnormal vibration monitoring module collects vibration signals of the chamber door through a built-in vibration sensor; when the abnormal sound monitoring module determines that there is abnormal sound in the sound signal, and the abnormal vibration monitoring module determines that there is abnormal vibration in the vibration signal, the intrusion monitoring unit sends an abnormal intrusion warning signal to the user data management and control unit through the built-in data networking unit.

[0009] Furthermore, the face recognition module and the personnel behavior analysis module are interconnected. The face recognition module collects video images of the outside of the chamber door through a built-in camera. When a person is detected in the video image, the face recognition module starts face detection and comparison, and sends the video image to the personnel behavior analysis module; after receiving the video image, the personnel behavior analysis module adopts a human skeleton key point detection algorithm to obtain the key points of the person in the image, and combines the built-in ranging device to obtain the relative position of the key points of the human body in the three-dimensional space, and calculates the change pattern of the relative position of the same key point of the human body in the three-dimensional space in adjacent video frames; when the face detection and comparison result of the face recognition module is less than the set confidence level, and the personnel behavior analysis module determines that the change pattern of the human body is abnormal behavior, the intrusion monitoring unit sends a personnel intrusion warning signal to the user data management and control unit through the built-in data networking unit.

[0010] Furthermore, the fire sensing unit includes a temperature monitoring module, a particulate matter monitoring module, a smoke monitoring module, a flame monitoring module, and a harmful gas monitoring module, and each monitoring module contains a battery management unit and a data networking unit; the temperature monitoring module, the particulate matter monitoring module, the smoke monitoring module, the flame monitoring module, and the harmful gas monitoring module send the collected monitoring data to the fire sensing unit through the data networking unit, and the fire sensing unit normalizes the data collected by each monitoring module at the same time after receiving it, and inputs the normalized data into the pre-trained fire classifier, and outputs the fire hazard coefficient through the fire classifier; the fire sensing unit compares the fire hazard coefficient with the fire alarm threshold. If the fire alarm threshold is exceeded, the fire warning signal is sent to the electrical safety monitoring unit and the user data management and control unit through the built-in data networking unit.

[0011] Furthermore, the water hazard sensing unit includes a flow rate monitoring device, a water level monitoring device, and a humidity monitoring device; the flow rate monitoring device, the water level monitoring device, and the humidity monitoring device are installed on the top plate of the chamber or the side wall close to the top plate, and all have a built-in battery management unit and a data networking unit; the water level monitoring device collects water level data H once every interval t, then the water inflow V = L(H1-H0), the water inflow speed S = (H1-H0) / t, L is the area of ​​the chamber bottom plate, H1 and H0 are the water level at the end time and the water level at the start time of the interval t respectively;

[0012] The flow rate monitoring device, water level monitoring device and humidity monitoring device collect monitoring data in real time and send it to the water hazard sensing unit through the built-in data networking unit. The water hazard sensing unit calculates the average value of the flow rate and humidity monitoring data collected at N intervals of time t, obtains the average flow rate and average humidity, and calculates the average water inflow. and average water inflow velocity The water hazard sensing unit normalizes the average flow rate, average humidity, average water inflow and average water inflow speed, and inputs the normalized data into the pre-trained water hazard classifier, and outputs the water hazard risk coefficient through the water hazard classifier; the water hazard sensing unit compares the water hazard risk coefficient with the water hazard alarm threshold. If the water hazard alarm threshold is exceeded, the water hazard warning signal is sent to the electrical safety monitoring unit and the user data management unit through the built-in data networking unit.

[0013] Furthermore, the electrical safety monitoring unit includes a current transformer, a current meter, and a power-off lockout protection device. The current transformer is directly connected to the power supply circuit of the electrical equipment to be tested, and the current meter is connected to the secondary side of the current transformer. The electrical safety monitoring unit determines the switch state and working state of the electrical equipment based on the current data measured by the current meter. After the electrical safety monitoring unit receives the early warning signal sent by the fire sensing unit or the water hazard sensing unit, the power-off lockout protection device activates the power-off lockout protection.

[0014] When there is current in the power supply circuit and the instantaneous current value is greater than the set current safety threshold, the electrical safety monitoring unit sends the electrical safety warning signal to the user data management and control unit through the built-in data networking unit; when there is current in the power supply circuit and the instantaneous current value is less than the normal operating current threshold, the average current value per unit time is calculated. If the average current value is still less than the normal operating current threshold, it is determined that the electrical equipment to be tested is in an abnormal state, the power-off lockout protection device starts the power-off lockout protection, and the electrical safety monitoring unit sends the electrical safety warning signal to the user data management and control unit through the built-in data networking unit.

[0015] Furthermore, the warning signals include abnormal intrusion warning signals, personnel intrusion warning signals, fire warning signals, water hazard warning signals, and electrical safety warning signals; the statistical characteristics include the frequency, proportion, and interval of the sending time of the warning signals sent by the intrusion monitoring unit, fire perception unit, water hazard perception unit, and electrical safety monitoring unit within time T, as well as the safety risk coefficient of the cavern; the intervals are 0:00-8:00, 8:00-16:00, and 16:00-24:00, and the user data management and control unit adopts a weighted statistical model to obtain the safety risk coefficient of the cavern; the self-learning strategy is that the user data management and control unit dynamically calculates the frequency of the warning signal based on historical warning signal data.

[0016] Furthermore, the modeling process of the weighted statistical model includes:

[0017] S1: Construct the weight vector W = [α1, α2, α3, α4, α5] of abnormal intrusion warning signal, personnel intrusion warning signal, fire warning signal, water hazard warning signal, and electrical safety warning signal;

[0018] S2: Calculate the frequency of abnormal intrusion warning signals, personnel intrusion warning signals, fire warning signals, water damage warning signals, and electrical safety warning signals within time T, and construct the frequency vector P = [p1, p2, p3, p4, p5] through self-learning strategy;

[0019] S3: Calculate the safety risk coefficient vector Q of the chamber, Where * is the vector product, . is the scalar product; calculate the product of all elements in vector Q to obtain the safety risk factor of the chamber.

[0020] Furthermore, the data networking unit includes a wired communication module and a wireless communication module. When the wired communication module is abnormally interrupted, the wireless communication module starts working.

[0021] The present invention provides a cavern security monitoring system for an underground space that can control a variety of monitoring equipment in an interconnected manner, thereby eliminating potential safety risks; at the same time, the system transmits abnormal alarm signals sent by the monitoring equipment to the terminal device carried by the owner or the monitoring platform of the management party through the Internet of Things module, and by analyzing the historical data sent by a variety of monitoring equipment, it can predict safety hazards, thereby constructing a smart security monitoring system for caverns in underground spaces. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A schematic diagram of a security monitoring system for an underground chamber in the present invention.

[0023] Figure 2 The present invention provides a working flow diagram of a security monitoring system for an underground space chamber. DETAILED DESCRIPTION

[0024] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the following will be a clear and complete description of the technical solutions in the embodiments of the present invention. Obviously, the embodiments described are some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work shall fall within the scope of protection of the present invention.

[0025] The schematic diagram of the underground space chamber security monitoring system (10) of the present invention is as follows Figure 1As shown, the system includes: an intrusion monitoring unit (11), a fire sensing unit (12), a water hazard sensing unit (13), an electrical safety monitoring unit (14), a data networking unit (15), a user data management unit (16), an abnormality alarm unit (17), a security monitoring platform (18) and a mobile terminal device (19); the intrusion monitoring unit (11), the fire sensing unit (12), the water hazard sensing unit (13), the electrical safety monitoring unit (14) and the abnormality alarm unit (17) are respectively connected to the user data management unit (16); the fire sensing unit (12) and the water hazard sensing unit (13) are respectively connected to the electrical safety monitoring unit (14);

[0026] The intrusion monitoring unit (11) comprises an infrared human body monitoring module (110), a face recognition module (111), an abnormal vibration monitoring module (112), an abnormal sound monitoring module (113), and a personnel behavior analysis module (114); the intrusion monitoring unit (11) transmits a personnel intrusion warning signal and an abnormal intrusion warning signal to a user data control unit (16) via a built-in data networking unit (15).

[0027] The infrared human body monitoring module (110) is installed on the outer wall of the chamber door and monitors whether a person approaches the chamber door through a built-in pyroelectric sensor. When the pyroelectric sensor detects that a person approaches the chamber door, the face recognition module (111), abnormal sound monitoring module (112), abnormal vibration monitoring module (113), and personnel posture analysis module (114) in the intrusion monitoring unit start working; the pyroelectric sensor adopts a PIR pyroelectric infrared sensor with a high sensitivity and dual-element compensation structure that is resistant to radio frequency interference, model XYC-PIR203B-S0, and a receiving wavelength of 5.5-14μm.

[0028] The face recognition module (111) is connected to the personnel behavior analysis module (115). When the face recognition module (111) starts working, the face recognition module (111) collects the video image of the outside of the chamber door through the built-in wide-angle camera. When the personnel detection model is used to detect the presence of a person in the video image, the face recognition module (111) sends the personnel video image to the personnel behavior analysis module (115). At the same time, the face recognition module (111) starts face detection and comparison, and determines whether the face detection and comparison results are less than a set confidence level. When the face comparison result is less than the set confidence level, the face recognition module (111) feeds back abnormal information to the intrusion monitoring unit (11).

[0029] The abnormal vibration monitoring module (112) is installed on the inner wall of the chamber door and collects the vibration signal of the chamber door through a built-in vibration sensor. When the abnormal vibration monitoring module (112) determines that abnormal vibration exists in the vibration signal, the abnormal vibration monitoring module (112) feeds back the abnormal information to the intrusion monitoring unit (11).

[0030] The abnormal sound monitoring module (113) is installed on the inner wall of the chamber door and collects sound signals near the chamber door through a built-in microphone. When the abnormal sound monitoring module (113) determines that there is an abnormal sound in the sound signal, the abnormal sound monitoring module (113) feeds back the abnormal information to the intrusion monitoring unit (11).

[0031] After receiving a video image of a person, the person behavior analysis module (114) uses a human skeleton key point detection algorithm to obtain key points of the person in the image, combines a built-in distance measuring device to obtain the relative position of the key points of the human body in three-dimensional space, and calculates the change pattern of the relative position of the same key point of the human body in three-dimensional space in adjacent video frames. If the person behavior analysis module (114) determines that there is abnormal behavior in the change pattern of the human body, the abnormal behavior warning signal is fed back to the intrusion monitoring unit (11).

[0032] The fire sensing unit (12) comprises a temperature monitoring module (120), a particle monitoring module (121), a smoke monitoring module (122), a flame monitoring module (123), and a harmful gas monitoring module (124); each of the monitoring modules internally comprises a battery management unit and a data networking unit (15); when a fire occurs, the power supply of each monitoring module is cut off and the internal battery management unit directly provides working power.

[0033] The temperature monitoring module (120) is used to collect temperature data in the chamber, and can use any one of distributed optical fiber sensors, PT100 thermocouple temperature sensors, infrared thermal imagers, infrared thermoelectric sensors, or infrared thermometers. The temperature sensors are distributed and installed on the top plate of the chamber. Fire monitoring is achieved by utilizing the fact that the indoor temperature rises significantly when a fire occurs.

[0034] The particle monitoring module (121) is used to collect particle size and concentration data of the particles in the chamber. It can use particle detection sensors such as light scattering detection, piezoelectric crystal oscillator detection, and beta ray detection. The particle detection sensors are installed on the ceiling of each chamber in the underground space. Fire monitoring is achieved by utilizing the characteristic that indoor particles increase significantly when a fire occurs.

[0035] The smoke sensing module (122) is used to collect smoke emitted during an early fire in a chamber, and an MQ-2 smoke sensor can be used; the MQ-2 smoke sensor is a high-performance sensor with high sensitivity and low cost. The principle of the MQ-2 smoke sensor is to output according to the change of the conductivity of the gas-sensitive material in the sensor as the detected gas concentration changes.

[0036] The flame monitoring module (123) is used to detect flame signals in real time and feed the detected signals back to the fire sensing unit in real time; the flame monitoring module uses a camera to collect video images, and a video image recognition device to recognize flames, and an intelligent image fire detector of Chengdu Century Chaoyang Technology Co., Ltd. can be used.

[0037] The harmful gas monitoring module (124) includes a CO sensor, a CO2 sensor, and a SO2 sensor. The harmful gas monitoring sensors are installed on the ceiling of each chamber in the underground space. Fire monitoring is achieved by utilizing the fact that the concentrations of CO, CO2, and SO2 gases in the chamber increase significantly when a fire occurs.

[0038] The water hazard sensing unit (13) comprises a flow rate monitoring device (130), a water level monitoring device (131), and a humidity monitoring device (132); each of the monitoring modules internally comprises a battery management unit and a data networking unit (15); when a water hazard occurs, the power supply of each monitoring module is cut off and the internal battery management unit directly provides working power.

[0039] The flow rate monitoring device (130) is installed on the top plate of the chamber or the side wall close to the top plate, and adopts IP67 or IP68 waterproof design; Doppler velocity radar or ultrasonic velocity measurement equipment can be selected, and the angle between the Doppler velocity radar antenna and the tunnel floor is not less than 45 degrees and not more than 60 degrees. When a flood occurs, the external water source invades the chamber and the water flow rate is relatively high, so as to realize the flood monitoring.

[0040] The water level monitoring device (131) is installed on the top plate of the chamber or the side wall close to the top plate, and adopts IP67 or IP68 waterproof design; ranging radar or ultrasonic ranging equipment can be selected, and the ultrasonic transmitting and receiving probe is installed perpendicular to the tunnel floor. When a flood occurs, the water source invades the chamber and the water body gradually accumulates, so as to realize the flood monitoring.

[0041] The humidity monitoring module (132) is installed on the top plate of the cavern or the side wall close to the top plate, and is used to collect real-time humidity data in the cavern. A resistive or capacitive humidity sensor can be used. The humidity sensor is distributed and installed in each cavern in the underground space. The characteristic that the humidity inside the cavern increases before a flood occurs is utilized to realize flood monitoring.

[0042] The electrical safety monitoring unit (14) includes a current transformer (140), a current meter (141) and a power-off locking protection device (142);

[0043] The current transformer (140) consists of a closed iron core and a winding. The primary winding has a smaller number of turns and is directly connected to the power supply circuit of the electrical equipment to be tested. The secondary winding has a larger number of turns and is connected in series to the circuit of the current meter (141).

[0044] The current meter (141) is connected to the secondary side of the current transformer and is used to measure the current of the power supply circuit. The electrical safety monitoring unit (14) determines the switch state and working state of the electrical equipment based on the current data measured by the current meter (141).

[0045] The power-off lockout protection device (142) is used to promptly cut off the relevant power supply when an electrical device fails and causes abnormal current fluctuations in the circuit, thereby preventing harm to human health or fire accidents. The device adopts a self-resetting over-voltage or under-voltage delay protector. At the same time, after the electrical safety monitoring unit (14) receives the early warning signal sent by the fire sensing unit (12) or the water damage sensing unit (13), the power-off lockout protection device (142) also starts the power-off lockout protection.

[0046] The data networking unit (15) includes a wired communication module and a wireless communication module. Under normal working conditions, the wired communication mode is adopted. When the wired communication module is abnormally interrupted, the wireless communication module starts working. At the same time, when the power supply circuit is cut off, the battery power supply can be automatically enabled.

[0047] After receiving and storing the warning signals sent by the intrusion monitoring unit (11), the fire sensing unit (12), the water hazard sensing unit (13), and the electrical safety monitoring unit (14), the user data management and control unit (16) sends an alarm signal to the abnormal alarm unit (17); the warning signals include abnormal intrusion warning signals, personnel intrusion warning signals, fire warning signals, water hazard warning signals, and electrical safety warning signals; the user data management and control unit (16) analyzes the signal data within a certain period of time to obtain statistical characteristics of the signal data; the statistical characteristics include the frequency, proportion, and interval of the sending time of the warning signals sent by the intrusion monitoring unit, the fire sensing unit, the water hazard sensing unit, and the electrical safety monitoring unit within a time T, as well as the safety risk coefficient of the chamber obtained by using a weighted statistical model; the intervals are 0:00-8:00, 8:00-16:00, and 16:00-24:00, and T is greater than 24 hours.

[0048] The abnormal alarm unit (17) is set in the underground monitoring room or the ground monitoring room, and is used to receive the alarm signal sent by the user data control unit (16). When the alarm signal is received, the sound and light alarm is activated, and the alarm information is pushed to the security monitoring platform (18) and the mobile terminal device (19) carried by the owner through the built-in Internet of Things module (171). When the management party or the owner eliminates the abnormal event corresponding to the alarm information, the security monitoring platform (18) or the mobile terminal device (19) sends a cancellation alarm signal to the abnormal alarm unit (17). In addition, the abnormal alarm unit (17) can also realize local alarm and remote alarm at the same time.

[0049] The workflow of the underground space chamber security monitoring system of the present invention is as follows: Figure 2 shown.

[0050] Collecting monitoring data (20): After the system is powered on, the security monitoring data of the underground chambers are collected in real time through each monitoring unit;

[0051] Intrusion monitoring (21): The intrusion monitoring unit (11) obtains the sound signal near the chamber door, the vibration signal of the chamber door, the pyroelectric sensor monitoring signal, and the video image data outside the chamber door;

[0052] Infrared human monitoring (210): using a pyroelectric sensor to monitor whether a person is approaching the chamber door. When the pyroelectric sensor detects that a person is approaching the chamber door, facial recognition (211), abnormal sound monitoring (215), and abnormal vibration monitoring (216) are sequentially executed;

[0053] Face recognition (211): When face recognition is started, the face recognition module (111) collects the video image of the outside of the chamber door. When the personnel detection model detects the presence of a person in the video image, the face recognition module starts face detection and comparison, and sequentially performs whether the face comparison result is less than the set confidence level (213), and sequentially performs personnel behavior analysis (212).

[0054] Personnel behavior analysis (212): After receiving a video image of a person, the key points of the person in the image are obtained using a human skeleton key point detection algorithm. The relative positions of the key points in the three-dimensional space are obtained by combining with the built-in distance measurement device. The relative positions of the same key points in the three-dimensional space of the human body in adjacent video frames are calculated. The human skeleton key point detection algorithm uses the AlphaPose or OpenPose detection model to extract the key points of the human body from the video frame, obtain the key point coordinates of each target image frame, and judge the position of the person in the 2D environment to obtain the depth information of the person. By fusing the head joint points and simplifying the limb joint points, the movement behavior of the same skeleton key point is analyzed, and the judgment of whether there is abnormal behavior is performed in sequence (213).

[0055] Less than the confidence level and abnormal behavior exists (213): When the face comparison result is less than the set confidence level and the human body has abnormal behavior, the human intrusion warning signal is sent (214), otherwise it returns to the infrared human body monitoring (210).

[0056] Sending a personnel intrusion warning signal (214): The intrusion monitoring unit (11) sends the personnel intrusion warning signal to the user data control unit (16) via the built-in data networking unit (15).

[0057] Abnormal sound monitoring (215): The built-in microphone of the abnormal sound monitoring module (113) collects the sound signal near the chamber door and makes an abnormal sound determination (the sound not generated by unlocking or opening the door).

[0058] Abnormal vibration monitoring (216): The vibration sensor built into the abnormal vibration monitoring module (112) collects the vibration signal on the chamber door and makes a judgment on abnormal vibration (vibration not caused by unlocking or opening the door).

[0059] Abnormality (217): When the abnormal sound monitoring module (113) determines that there is abnormal sound (sound not generated by unlocking or opening the door) in the sound signal, and the abnormal vibration monitoring module (112) determines that there is abnormal vibration (vibration not generated by unlocking or opening the door) in the vibration signal, the abnormal intrusion warning signal (215) is sent, otherwise the infrared human body monitoring (210) is returned to be executed.

[0060] Sending abnormal intrusion warning signal (218): The intrusion monitoring unit (11) sends the abnormal intrusion warning signal to the user data control unit (16) through the built-in data networking unit (15).

[0061] Fire sensing (22): The fire sensing unit (12) obtains the real-time fire data collected by each monitoring module;

[0062] Data normalization (221): The fire sensing unit (12) normalizes the fire data (temperature, particle concentration, smoke concentration, flame value, and harmful gas concentration) at the same moment;

[0063] Input classifier (222): input the normalized fire data at the same moment into the pre-trained classifier, and obtain the fire hazard coefficient at a certain moment through the prediction of the classifier;

[0064] Calculate the fire hazard coefficient (223): After prediction by the classifier, output the fire hazard coefficient corresponding to the fire data at a certain moment;

[0065] Greater than a threshold value (224): the fire sensing unit (12) compares the fire risk factor with the fire alarm threshold value. If the fire risk factor exceeds the fire alarm threshold value, the fire sensing unit (12) executes sending a fire warning signal (225), otherwise returns to executing fire sensing (22).

[0066] Sending a fire warning signal (225): The fire sensing unit (12) sends the fire warning signal to the user data control unit (16) via the built-in data networking unit (15).

[0067] Water hazard sensing (23): The water hazard sensing unit (13) obtains the real-time data collected by each monitoring device;

[0068] Collecting monitoring data (231): The flow rate monitoring device (130), the water level monitoring device (131) and the humidity monitoring device (132) collect the flow rate, water level and humidity data in the chamber in real time, wherein the water level monitoring device collects water level data H once every interval t, and the water inflow volume V = L(H1-H0), the water inflow speed S = (H1-H0) / t, L is the area of ​​the chamber bottom plate, H1 and H0 are the water level at the end time and the water level at the start time of the interval t respectively. At the same time, the flow rate monitoring device (130), the water level monitoring device (131) and the humidity monitoring device (132) send the collected data to the water hazard sensing unit (13) through the built-in data networking unit.

[0069] Calculating the average water inflow volume and average water inflow speed (232): After receiving the data sent by the flow rate monitoring device (130), the water level monitoring device (131) and the humidity monitoring device (132), the water hazard sensing unit (13) calculates the average value of the flow rate and humidity monitoring data collected at N intervals of time t, obtains the average flow rate and average humidity, and calculates the average water inflow volume based on the water inflow volume and water inflow speed. and average water inflow velocity

[0070] Data normalization (233): The water hazard sensing unit (13) normalizes the water hazard data (average flow rate, average humidity, average water inflow volume, and average water inflow speed) at the same moment;

[0071] Calculate the flood hazard coefficient (234): Input the normalized flood hazard data at the same moment into the pre-trained classifier, and obtain the flood hazard coefficient at a certain moment through the prediction of the classifier;

[0072] Exceeding the threshold value (235): When the flood hazard sensing unit (13) compares the flood hazard coefficient with the flood hazard alarm threshold value, if the flood hazard coefficient exceeds the flood hazard alarm threshold value, it executes sending a flood hazard warning signal (236), otherwise it returns to execute collecting monitoring data (231).

[0073] Sending a flood warning signal (236): The flood sensing unit (13) sends the flood warning signal to the user data control unit (16) via the built-in data networking unit (15).

[0074] Electrical appliance safety monitoring (24): The current transformer (140) in the electrical appliance safety monitoring unit (14) is directly connected to the power supply circuit of the electrical appliance to be tested, and the current meter (141) is connected to the secondary side of the current transformer (140) and measures the current data of the power supply circuit in real time.

[0075] Measuring current data (241): The current meter (141) monitors the current data in the circuit to be measured in real time, and uses the current data to determine the switch state of the electrical device (242).

[0076] Electrical device switch state (242): When current flows through the power supply circuit, the electrical device is in the working state; when no current flows through the power supply circuit, the electrical device is in the off state.

[0077] The instantaneous current value is greater than the safety threshold (243): when the electrical appliance switch state (242) is in the working state, it is determined whether the instantaneous current value in the measured current data (241) is greater than the set current safety threshold. When the instantaneous current value is greater than the set current safety threshold, the electrical appliance safety monitoring unit executes sending an electrical appliance abnormality warning signal (246); otherwise, it executes calculating the average current value (244).

[0078] Calculating an average current value (244): When current flows through the power supply circuit and the instantaneous current value is less than the normal operating current threshold, the average current value per unit time is calculated based on the current data collected over a period of time.

[0079] Average current value < working threshold value (245): The electrical safety monitoring unit (14) compares the average current value with the working current threshold value based on the calculated average current value. If the average current value is less than the normal working current threshold value, it is determined that the electrical device to be tested is in an abnormal state and executes the sending of an electrical abnormality warning signal (246). Otherwise, it returns to execute the electrical safety monitoring (24).

[0080] Sending an electrical appliance abnormality warning signal (246): The power-off lockout protection device (142) in the electrical appliance safety monitoring unit (14) starts the power-off lockout protection, and at the same time, the electrical appliance safety monitoring unit (14) sends the electrical appliance abnormality status signal to the user data control unit (16) through the built-in data networking unit (15).

[0081] Analyzing signal data (25): After receiving the abnormal intrusion warning signal, the personnel intrusion warning signal, the fire warning signal, the water hazard warning signal, and the electrical safety warning signal, the user data control unit (16) sends a corresponding alarm signal to the abnormal alarm unit. At the same time, the user data control unit (16) uses a self-learning strategy to analyze the signal data within a certain period of time to obtain the statistical characteristics of the signal data; the modeling process of the weighted statistical model includes:

[0082] S1: Construct the weight vector W = [α1, α2, α3, α4, α5] of abnormal intrusion warning signal, personnel intrusion warning signal, fire warning signal, water hazard warning signal, and electrical safety warning signal;

[0083] S2: Calculate the frequency of abnormal intrusion warning signals, personnel intrusion warning signals, fire warning signals, water damage warning signals, and electrical safety warning signals within time T, and construct a frequency vector P = [p1, p2, p3, p4, p5] through a self-learning strategy; the self-learning strategy is that the user data management unit dynamically calculates the frequency of warning signals based on historical warning signal data;

[0084] S3: Calculate the safety risk coefficient vector Q of the chamber, Where * is the vector product, . is the scalar product; calculate the product of all elements in vector Q to obtain the safety risk factor of the chamber.

[0085] Abnormal alarm (26): After receiving the alarm signal, the abnormal alarm unit (17) activates the sound and light alarm and sends the alarm information to the security monitoring platform (18) and the mobile terminal device (19) through the built-in Internet of Things module.

Claims

1. A security monitoring system for an underground space, comprising: Intrusion monitoring unit, fire sensing unit, water hazard sensing unit, electrical safety monitoring unit, user data management and control unit and abnormal alarm unit; the intrusion monitoring unit, fire sensing unit, water hazard sensing unit and electrical safety monitoring unit all have built-in data networking units, and are respectively communicated with the user data management and control unit; the fire sensing unit and the water hazard sensing unit are respectively communicated with the electrical safety monitoring unit, and when the electrical safety monitoring unit receives the early warning signal sent by the fire sensing unit and the water hazard sensing unit through the data networking unit, it starts the power-off lockout protection; the abnormal alarm unit is communicated with the user data management and control unit; after the user data management and control unit receives and stores the early warning signal sent by the intrusion monitoring unit, fire sensing unit, water hazard sensing unit and electrical safety monitoring unit, it sends an alarm signal to the abnormal alarm unit, and the user data management and control unit adopts a self-learning strategy to analyze the signal data within a certain period of time and obtain the statistical characteristics of the signal data; after receiving the alarm signal, the abnormal alarm unit starts the sound and light alarm, and sends the alarm information to the security monitoring platform and mobile terminal device through the built-in Internet of Things module; The statistical features include the frequency, proportion, and time interval of warning signals sent by the intrusion monitoring unit, fire sensing unit, water hazard sensing unit, and electrical safety monitoring unit within a time period T, as well as the safety risk coefficient of the chamber; the self-learning strategy dynamically calculates the frequency of warning signals based on historical warning signal data; The safety risk coefficient is obtained using a weighted statistical model. The modeling process of the weighted statistical model includes: S1: Construct the weight vector W = [α1, α2, α3, α4, α5] of abnormal intrusion warning signal, personnel intrusion warning signal, fire warning signal, water hazard warning signal, and electrical safety warning signal; S2: Calculate the frequency of abnormal intrusion warning signals, personnel intrusion warning signals, fire warning signals, water damage warning signals, and electrical safety warning signals within time T, and construct the frequency vector P = [p1, p2, p3, p4, p5] through self-learning strategy; S3: Calculate the safety risk coefficient vector Q of the chamber, Where * is the vector product, . is the scalar product; Calculate the product of all elements in vector Q to obtain the safety risk factor of the chamber.

2. An underground space chamber security monitoring system as described in claim 1, wherein the intrusion monitoring unit includes an infrared human body monitoring module, a face recognition module, an abnormal sound monitoring module, an abnormal vibration monitoring module, and a personnel behavior analysis module; the infrared human body monitoring module monitors whether there is a person approaching the chamber door through a built-in pyroelectric sensor. When the pyroelectric sensor detects that there is a person approaching the chamber door, the face recognition module, abnormal sound monitoring module, abnormal vibration monitoring module, and personnel posture analysis module in the intrusion monitoring unit start working.

3. In an underground space cavern security monitoring system as described in claim 2, the abnormal sound monitoring module collects sound signals near the cavern door through a built-in microphone, and the abnormal vibration monitoring module collects vibration signals of the cavern door through a built-in vibration sensor; when the abnormal sound monitoring module determines that there is abnormal sound in the sound signal, and the abnormal vibration monitoring module determines that there is abnormal vibration in the vibration signal, the intrusion monitoring unit sends an abnormal intrusion warning signal to the user data management and control unit through the built-in data networking unit.

4. An underground space chamber security monitoring system as described in claim 2, wherein the face recognition module and the personnel behavior analysis module are interconnected, and the face recognition module collects video images of the outside of the chamber door through a built-in camera. When a person is detected in the video image, the face recognition module starts face detection and comparison, and sends the video image to the personnel behavior analysis module; after receiving the video image, the personnel behavior analysis module adopts a human skeleton key point detection algorithm to obtain the key points of the person in the image, and combines the built-in ranging device to obtain the relative position of the key points of the human body in three-dimensional space, and calculates the change pattern of the relative position of the same key point of the human body in adjacent video frames in three-dimensional space; when the face detection and comparison result of the face recognition module is less than the set confidence level, and the personnel behavior analysis module determines that the change pattern of the human body is abnormal behavior, the intrusion monitoring unit sends a personnel intrusion warning signal to the user data management and control unit through the built-in data networking unit.

5. An underground space chamber security monitoring system as described in claim 1, wherein the fire sensing unit includes a temperature monitoring module, a particulate matter monitoring module, a smoke monitoring module, a flame monitoring module, and a harmful gas monitoring module, and each monitoring module contains a battery management unit and a data networking unit; the temperature monitoring module, the particulate matter monitoring module, the smoke monitoring module, the flame monitoring module, and the harmful gas monitoring module send the collected monitoring data to the fire sensing unit through the data networking unit, and the fire sensing unit normalizes the data collected by each monitoring module at the same time after receiving it, and inputs the normalized data into the pre-trained fire classifier, and outputs the fire hazard coefficient through the fire classifier; the fire sensing unit compares the fire hazard coefficient with the fire alarm threshold. If the fire alarm threshold is exceeded, the fire warning signal is sent to the electrical safety monitoring unit and the user data management and control unit through the built-in data networking unit.

6. An underground space cavern security monitoring system according to claim 1, wherein the water hazard sensing unit includes a flow rate monitoring device, a water level monitoring device, and a humidity monitoring device; the flow rate monitoring device, the water level monitoring device, and the humidity monitoring device are installed on the cavern roof or the side wall near the roof, and each has a built-in battery management unit and a data networking unit; the water level monitoring device collects water level data H once every interval t, where the water inflow volume V = L(H1-H0), the water inflow speed S = (H1-H0) / t, L is the area of ​​the cavern floor, H1 and H0 are the water level at the end and starting times of the interval t, respectively; The flow rate monitoring device, water level monitoring device and humidity monitoring device collect monitoring data in real time and send it to the water hazard sensing unit through the built-in data networking unit. The water hazard sensing unit calculates the average value of the flow rate and humidity monitoring data collected at N intervals of time t, obtains the average flow rate and average humidity, and calculates the average water inflow. and average water inflow velocity The water hazard sensing unit normalizes the average flow rate, average humidity, average water inflow and average water inflow speed, and inputs the normalized data into the pre-trained water hazard classifier, and outputs the water hazard risk coefficient through the water hazard classifier; the water hazard sensing unit compares the water hazard risk coefficient with the water hazard alarm threshold. If the water hazard alarm threshold is exceeded, the water hazard warning signal is sent to the electrical safety monitoring unit and the user data management unit through the built-in data networking unit.

7. An underground space chamber security monitoring system according to claim 1, wherein the electrical safety monitoring unit comprises a current transformer, a current meter, and a power-off lockout protection device, the current transformer being directly connected to the power supply circuit of the electrical equipment to be tested, and the current meter being connected to the secondary side of the current transformer, the electrical safety monitoring unit determining the switch state and operating state of the electrical equipment based on current data measured by the current meter; and the power-off lockout protection device activating power-off lockout protection upon receiving a warning signal from the fire sensing unit or the water hazard sensing unit by the electrical safety monitoring unit; When there is current in the power supply circuit and the instantaneous current value is greater than the set current safety threshold, the electrical safety monitoring unit sends the electrical safety warning signal to the user data management and control unit through the built-in data networking unit; when there is current in the power supply circuit and the instantaneous current value is less than the normal operating current threshold, the average current value per unit time is calculated. If the average current value is still less than the normal operating current threshold, it is determined that the electrical equipment to be tested is in an abnormal state, the power-off lockout protection device starts the power-off lockout protection, and the electrical safety monitoring unit sends the electrical safety warning signal to the user data management and control unit through the built-in data networking unit.

8. An underground space chamber security monitoring system as described in claim 1, wherein the early warning signal includes an abnormal intrusion early warning signal, a personnel intrusion early warning signal, a fire early warning signal, a water hazard early warning signal, and an electrical safety early warning signal; the intervals are 0:00-8:00, 8:00-16:00, and 16:00-24:

00.

9. An underground space chamber security monitoring system as described in claim 1, wherein the data networking unit includes a wired communication module and a wireless communication module, and when the wired communication module is abnormally interrupted, the wireless communication module starts working.

Images