Elevator passenger monitoring system based on Internet of Things
Through the IoT elevator passenger monitoring system, combined with the time-frequency domain joint analysis and multi-source fusion algorithm of edge nodes, the problems of poor environmental adaptability and insufficient real-time performance are solved, high-precision elevator passenger counting and behavioral risk identification are achieved, and the efficiency of elevator operation management is improved.
Patent Information
- Application Number
- CN202511111368.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-08
- Publication Date
- 2025-10-10
AI Technical Summary
In the existing technology, the elevator passenger monitoring system has poor environmental adaptability, lacks behavioral characteristics and is not real-time, resulting in a high error rate and an inability to meet real-time intervention needs.
An IoT-based elevator passenger monitoring system is used, combined with time-frequency domain joint analysis of edge nodes. The number of elevator passengers is calculated through a multi-source fusion algorithm, and the risk of personnel behavior is identified through abnormal retention and fall judgments. Infrared verification is dynamically triggered, data transmission delay is optimized, and high-precision real-time monitoring is achieved.
It achieves high-precision elevator passenger counting in complex scenarios, reduces the accident rate, improves elevator operation management efficiency, and meets the needs of real-time monitoring and immediate handling of abnormal situations.
Smart Images
Figure CN120756953A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of Internet of Things sensing technology, and more specifically, to an elevator passenger monitoring system based on the Internet of Things. Background Art
[0002] In the process of urbanization, elevators, as core vertical transportation equipment, require real-time monitoring of passengers, crucial for ensuring operational safety and optimizing dispatch efficiency. Traditional monitoring relies on manual labor or single equipment, making it difficult to address complex scenarios. However, the multi-device collaboration and real-time analysis capabilities of IoT technology offer a new path for precise monitoring.
[0003] In existing technologies, infrared array sensors count through the pyroelectric effect. Although they are sensitive to static human bodies, they are significantly affected by ambient temperature, resulting in a large error rate. LiDAR counts people through point cloud modeling. Its advantage is that it is resistant to light interference, but its disadvantage is that it is sensitive to the reflection of clothing materials, especially metal reflections, which are easy to misjudge and have a high false detection rate. Neither of them is associated with human behavioral characteristics, such as sudden falls and deliberate delays, and the data needs to be processed in the cloud, with high one-way delay, which cannot meet the needs of real-time intervention.
[0004] Therefore, in response to the problems of poor environmental adaptability, lack of behavioral characteristics and insufficient real-time performance in existing technologies, there is an urgent need for an elevator passenger monitoring system based on the Internet of Things, which combines the time-frequency domain joint analysis of edge nodes to achieve high-precision real-time monitoring of the number of people and behavioral status. Summary of the Invention
[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides an elevator passenger monitoring system based on the Internet of Things, which solves the problems raised in the above-mentioned background technology through the following scheme.
[0006] To achieve the above objectives, the present invention provides the following technical solutions: an elevator passenger monitoring system based on the Internet of Things, comprising: Initialization module: Deploys perception terminals, edge nodes, and a management platform based on the classic architecture of IoT perception, transmission, and applications; Number monitoring module: The edge node calculates the number of people in the elevator through a multi-source fusion algorithm; Abnormal judgment module: It can identify the risk of personnel behavior through abnormal judgment of detention and fall; Frequency domain interference suppression module: Filters effective radar signals based on the energy ratio of the main frequency band, dynamically triggers infrared verification, and reduces radar interference misjudgment; Transmission optimization module: Dynamically adjusts transmission delay based on data redundancy and compression rate to ensure real-time data transmission.
[0007] Preferably, the sensing terminal is used to collect multi-dimensional data of the elevator environment, including: collecting the micro-Doppler frequency of the echo signal through the motion sensing terminal The radial distance r(t) from the target, where i=1, 2, 3, corresponds to the movement frequency of limbs, trunk, and head respectively; the temperature gradient of 64 pixel blocks is collected by the thermal distribution sensing terminal and infrared temperature entropy H, where k = 1, 2, ..., 64, , Indicates the pixel ratio of temperature interval i; collects instantaneous air pressure through the spatial pressure sensing terminal ; The edge node is connected to the perception terminal in communication, receives perception data through the ZigBee protocol, and performs time-frequency domain joint analysis and fusion algorithm; the management platform: receives the processing results of the edge node to achieve real-time monitoring and historical data storage.
[0008] Preferably, the multi-source fusion algorithm includes a millimeter wave radar effective number calculation unit, an infrared thermal imaging effective number calculation unit, an air pressure dynamic correction unit and a multi-source fusion unit; the millimeter wave radar effective number calculation unit screens effective targets and performs weighted correction based on the human motion characteristics of the micro-Doppler effect to obtain the millimeter wave radar effective number. ,in Indicates the total number of targets detected by the radar, Indicates the proportion of effective human targets; the infrared thermal imaging effective number calculation unit calculates the thermal feature matching degree by temperature gradient spatial distribution and entropy value to obtain the effective number of infrared thermal imaging ,in Indicates the total number of heat sources detected by infrared; the air pressure dynamic correction unit is used to eliminate the air pressure interference caused by the operation of the elevator space, and derive the number of people with dynamic air pressure correction based on Boyle's law ,in represents standard atmospheric pressure, represents the air density, represents the acceleration due to gravity, Indicates the running speed of the elevator space over time changes, Indicates the volume of the elevator space, Represents the average volume of the human body.
[0009] Preferably, the multi-source fusion unit calculates the dynamic weight based on the variance of the first three sampling periods, fuses the three types of data and corrects the periodic fluctuations to obtain the multi-source fusion number ,in 、 as well as represents the dynamic weight, , and Similarly, express variance, express mean, Indicates the number of radar users in the first three sampling periods, represents the fluctuation variance of infrared population, represents the variance of air pressure fluctuation, It represents the operation cycle of the elevator space, and t represents the current sampling time.
[0010] Preferably, the abnormal retention determination is: the retention degree is calculated by integrating the radial velocity of the human body to obtain the abnormal retention index , where t represents the duration, represents the radial velocity of the human body relative to the radar, Represents the target radial distance r(t) over time The dynamic changes of Indicates the normal movement threshold; the management platform monitors in real time. And infrared temperature entropy When Indicates the preset value of abnormal retention index, Indicates the infrared temperature entropy threshold, which determines abnormal retention and issues an abnormal retention warning signal.
[0011] Preferably, the abnormal fall judgment is to obtain the sudden fall index by combining the micro-Doppler frequency mutation and the temperature gradient sudden increase characteristics. ,in represents the fall feature time window, represents the average temperature gradient within the time window, Indicates different moments The instantaneous temperature gradient, Represents the micro-Doppler frequency over time The management platform monitors the changes in real time. When Indicates the preset value of the sudden fall index, which is determined as a sudden fall and issues a sudden fall warning signal.
[0012] Preferably, the specific method of dynamic trigger infrared verification is: extracting the energy proportion of the main frequency band of human motion through FFT ,in represents the FFT result of the millimeter wave echo signal, Indicates the number of FFT points; the management platform monitors in real time. < When Indicates the energy percentage threshold, starts the infrared thermal imaging secondary verification and increases the sampling frame. If the secondary verification still meets < , issuing a radar signal abnormality warning signal.
[0013] Preferably, the compression ratio , where C represents data redundancy; the dynamic adjustment of transmission delay ,in Indicates the original data volume, B indicates the transmission bandwidth, Indicates edge node processing delay.
[0014] Technical effects and advantages of the present invention: 1. This invention uses a multi-source fusion algorithm that combines millimeter-wave radar micro-Doppler multi-component analysis, infrared thermal gradient spatial modeling, and air pressure dynamic correction to obtain high-precision monitoring results for the number of elevator passengers. This solves the problem of large counting errors caused by environmental interference on a single sensor, achieving the beneficial effect of consistently reliable passenger counts in complex scenarios, and providing accurate data support for overload warnings and transport capacity scheduling. 2. The present invention uses a dual-feature determination algorithm based on the abnormal retention index and the sudden fall index to obtain real-time identification results of the behavioral risks of elevator passengers. This solves the problem of existing technologies that can only count the number of people but cannot associate them with their safe behaviors, thereby reducing the incidence of elevator accidents. 3. The present invention obtains low-latency monitoring data transmission and analysis results through joint time-frequency domain processing of edge nodes and adaptive data compression algorithm, which solves the problem of high latency and inability to intervene in real time due to data reliance on cloud processing. It achieves the beneficial effect of rapid feedback of monitoring results to the elevator management platform, realizing remote real-time monitoring and immediate handling of abnormal situations, reducing manual inspection costs and improving elevator operation management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] Figure 1 It is a schematic diagram of the overall structure of the present invention. DETAILED DESCRIPTION
[0016] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.
[0017] As attached Figure 1 The elevator passenger monitoring system shown is based on the Internet of Things and includes an initialization module, a number monitoring module, an abnormality judgment module, a frequency domain interference suppression module, and a transmission optimization module.
[0018] The initialization module deploys perception terminals, edge nodes, and a management platform based on the classic architecture of perception, transmission, and application of the Internet of Things; Specifically, it should be noted that the sensing terminal is used to collect multi-dimensional data of the elevator environment, including: collecting the micro-Doppler frequency of the echo signal through the motion sensing terminal; The radial distance r(t) from the target, where i=1, 2, 3, corresponds to the movement frequency of limbs, trunk, and head respectively; the temperature gradient of 64 pixel blocks is collected by the thermal distribution sensing terminal and infrared temperature entropy H, where k = 1, 2, ..., 64, , Indicates the pixel ratio of temperature interval i; collects instantaneous air pressure through the spatial pressure sensing terminal ; The edge node is connected to the perception terminal in communication, receives perception data through the ZigBee protocol, and performs time-frequency domain joint analysis and fusion algorithm; the management platform: receives the processing results of the edge node to achieve real-time monitoring and historical data storage.
[0019] The number monitoring module: the edge node calculates the number of people in the elevator through a multi-source fusion algorithm; Specifically, it should be noted that the multi-source fusion algorithm includes a millimeter wave radar effective number calculation unit, an infrared thermal imaging effective number calculation unit, an air pressure dynamic correction unit, and a multi-source fusion unit; the multi-source fusion algorithm includes a millimeter wave radar effective number calculation unit, an infrared thermal imaging effective number calculation unit, an air pressure dynamic correction unit, and a multi-source fusion unit; the millimeter wave radar effective number calculation unit is based on the human body motion characteristics of the micro-Doppler effect, screens effective targets and weighted correction to obtain the millimeter wave radar effective number ,in Indicates the total number of targets detected by the radar, Indicates the proportion of effective human targets; it is understandable that different parts of the human body have different movement frequencies (limbs>head>torso). Differentiate and weight ( High-frequency motion is given a higher weight), which is consistent with the physical characteristics of human motion; the exponential function converts the frequency deviation into a correction coefficient of 0-1, mathematically ensuring that the further the deviation from the typical frequency, the stronger the correction; non-human targets (such as shaking luggage, etc.) are eliminated from radar detection. <0.5Hz or >5Hz), correct individual differences in human movement frequency (such as different movement frequencies between children and adults); the infrared thermal imaging effective number calculation unit calculates the thermal feature matching degree by temperature gradient spatial distribution and entropy value to obtain the infrared thermal imaging effective number of people ,in Indicates the total number of heat sources detected by infrared; it can be understood that the temperature gradient of human skin is concentrated in (typical value 2.15℃ / cm), the Gaussian function suppresses gradients that deviate from this range (such as wall heat sources); the larger the edge pixel block k, the lower the weight, reducing the thermal interference reflected by the car wall; when the temperature entropy H is less than 1.2, the heat distribution is orderly (in line with the human body shape), and 2-H strengthens this feature; thereby excluding non-human heat sources (such as direct sunlight and heating) in infrared detection and solving the problem of ambient temperature interference; the air pressure dynamic correction unit is used to eliminate the air pressure interference caused by the operation of the elevator space, and the air pressure dynamic correction number is derived based on Boyle's law. ,in represents standard atmospheric pressure, represents the air density, represents the acceleration due to gravity, Indicates the running speed of the elevator space over time changes, Indicates the volume of the elevator space, Indicates the average volume of the human body; it is understandable that the elevator will cause the air pressure to change ( The interference is the pressure difference caused by altitude change, which needs to be eliminated first. Based on Boyle's law (in a closed space, air pressure is inversely proportional to volume), the physical characteristics of air pressure change are used to supplement the counting deviation of radar / infrared in severe obstruction (such as dense crowds). The multi-source fusion unit calculates the dynamic weight based on the variance of the first three sampling periods, fuses the three types of data and corrects the periodic fluctuations to obtain the multi-source fusion number of people. ,in 、 as well as represents the dynamic weight, , and Similarly, express variance, express mean, Indicates the number of radar users in the first three sampling periods, represents the fluctuation variance of infrared population, represents the variance of air pressure fluctuation, represents the operation cycle of the elevator space, and t represents the current sampling time. Among them, the variance reflects the stability of the data (the smaller the variance, the more reliable it is), and the weight is Proportional to achieve more reliable data contribution; the air pressure / radar data in the elevator operation cycle has periodic fluctuations (such as when starting and stopping), the sine term Compensation is mathematically consistent with periodic signal characteristics; therefore, the advantages of the three sensor types are integrated to avoid counting errors caused by the failure of a single data source. Understandably, a single sensor is susceptible to interference from the scene (such as radar misidentifying luggage and infrared misdetecting high-temperature objects). Multi-source fusion improves accuracy through complementary physical principles—radar based on motion characteristics, infrared based on thermal characteristics, and air pressure based on volume characteristics—covering counting requirements in different scenarios. This solves the large counting errors of existing single sensors and enables accurate people counting in complex scenarios (crowded, occluded, and mixed static and dynamic scenes).
[0020] The abnormality judgment module fills the gap in behavior monitoring by judging the abnormality of staying and falling; Specifically, it should be noted that the abnormal retention determination is to calculate the retention degree by integrating the radial velocity of the human body to obtain the abnormal retention index. , where t represents the duration, represents the radial velocity of the human body relative to the radar, Represents the target radial distance r(t) over time The dynamic changes of Indicates the normal movement threshold; the management platform monitors in real time. And infrared temperature entropy When Indicates the preset value of abnormal retention index, Indicates the infrared temperature entropy threshold, which determines abnormal retention and issues an abnormal retention warning signal. It can be understood that: normal walking speed ≥ 0.5m / s ( =0.5), when stationary (| |=0), integral term The influence of the cumulative static time (the longer the time, the larger S), and the square term strengthen the weight of the static state, thereby distinguishing temporary stays (such as pressing the floor button, short t, small S) from abnormal stays (such as fainting, deliberate stay, long t, large S). The abnormal fall judgment: Combine the micro-Doppler frequency mutation and temperature gradient sudden increase characteristics to obtain the sudden fall index ,in represents the fall feature time window, represents the average temperature gradient within the time window, Indicates different moments The instantaneous temperature gradient, Represents the micro-Doppler frequency over time The management platform monitors the changes in real time. When Indicates the preset value of the sudden fall index, which is judged as a sudden fall and issues a sudden fall warning signal. It can be understood that: at the moment of falling, the human body's movement frequency drops sharply. (is the frequency mutation value), and the posture change causes the temperature gradient to increase sharply ( is the average gradient), the product of the two amplifies the fall feature; time window Match the typical duration of a human fall. Therefore, the dual features of motion and shape are used to determine falls, avoiding misjudgment based on a single feature (such as bending over to pick up an object).
[0021] The frequency domain interference suppression module: screens effective radar signals based on the main frequency band energy ratio, dynamically triggers infrared verification, and reduces radar interference misjudgment; Specifically, it should be noted that the specific method of dynamic trigger infrared verification is: extract the energy proportion of the main frequency band of human motion through FFT ,in represents the FFT result of the millimeter wave echo signal, Indicates the number of FFT points; the management platform monitors in real time. < When Indicates the energy percentage threshold, starts the infrared thermal imaging secondary verification and increases the sampling frame. If the secondary verification still meets < , issuing a warning signal of abnormal radar signal. It is understandable that the main frequency band of human motion is 1-3Hz ( =1, =3), FFT energy ratio Reflects the proportion of human features in the radar signal; therefore, the sensor weight is dynamically switched to reduce the interference of non-human motion (such as elevator vibration) on the radar.
[0022] The transmission optimization module dynamically adjusts the transmission delay based on data redundancy and compression rate to ensure real-time data transmission.
[0023] Specifically, it should be noted that the compression ratio , where C represents data redundancy; the dynamic adjustment of transmission delay ,in Indicates the original data volume, B indicates the transmission bandwidth, It indicates the processing delay of the edge node. It can be understood that: the higher the data redundancy C (the more repeated frames), the greater the compression rate r (the combination of tanh and exponential functions achieves smooth adjustment); the transmission delay By data volume , bandwidth B and processing delay The decision can reduce the transmission volume while ensuring the validity of data and meet the time requirements of real-time warning.
[0024] Secondly: The drawings of the embodiments disclosed in the present invention only involve structures related to the embodiments disclosed in the present invention. Other structures may refer to conventional designs. The same embodiment and different embodiments of the present invention may be combined with each other without conflict. Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. An elevator passenger monitoring system based on the Internet of Things, characterized in that: include: Initialization module: Deploys perception terminals, edge nodes, and a management platform based on the classic architecture of IoT perception, transmission, and applications; Number monitoring module: The edge node calculates the number of people in the elevator through a multi-source fusion algorithm; Abnormal judgment module: It can identify the risk of personnel behavior through abnormal judgment of detention and fall; Frequency domain interference suppression module: Filters effective radar signals based on the energy ratio of the main frequency band, dynamically triggers infrared verification, and reduces radar interference misjudgment; Transmission optimization module: Dynamically adjusts transmission delay based on data redundancy and compression rate to ensure real-time data transmission.
2. The elevator passenger monitoring system based on the Internet of Things according to claim 1 is characterized by: The sensing terminal is used to collect multi-dimensional data of the elevator environment, including: collecting the micro-Doppler frequency of the echo signal through the motion sensing terminal The radial distance r(t) from the target, where i=1, 2, 3, corresponds to the movement frequency of limbs, trunk, and head respectively; the temperature gradient of 64 pixel blocks is collected by the thermal distribution sensing terminal and infrared temperature entropy H, where k = 1, 2, ..., 64, , Indicates the pixel ratio of temperature interval i; collects instantaneous air pressure through the spatial pressure sensing terminal ; The edge node is connected to the perception terminal in communication, receives perception data through the ZigBee protocol, and performs time-frequency domain joint analysis and fusion algorithm; the management platform: receives the processing results of the edge node to achieve real-time monitoring and historical data storage.
3. The elevator passenger monitoring system based on the Internet of Things according to claim 2 is characterized by: The multi-source fusion algorithm includes a millimeter wave radar effective number calculation unit, an infrared thermal imaging effective number calculation unit, an air pressure dynamic correction unit, and a multi-source fusion unit; the millimeter wave radar effective number calculation unit screens effective targets and performs weighted correction based on the human motion characteristics of the micro-Doppler effect to obtain the millimeter wave radar effective number. ,in Indicates the total number of targets detected by the radar, Indicates the proportion of effective human targets; the infrared thermal imaging effective number calculation unit calculates the thermal feature matching degree by temperature gradient spatial distribution and entropy value to obtain the effective number of infrared thermal imaging ,in Indicates the total number of heat sources detected by infrared; The air pressure dynamic correction unit is used to eliminate the air pressure interference caused by the operation of the elevator space, and derive the air pressure dynamic correction number based on Boyle's law. ,in Indicates standard atmospheric pressure, represents the air density, represents the acceleration due to gravity, Indicates the running speed of the elevator space over time changes, Indicates the volume of the elevator space, Represents the average volume of the human body.
4. The elevator passenger monitoring system based on the Internet of Things according to claim 3 is characterized by: The multi-source fusion unit calculates the dynamic weight based on the variance of the first three sampling periods, fuses the three types of data and corrects the periodic fluctuations to obtain the multi-source fusion number ,in 、 as well as represents the dynamic weight, , and Similarly, express variance, express mean, Indicates the number of radar users in the first three sampling periods, represents the fluctuation variance of infrared population, represents the variance of air pressure fluctuation, It represents the operation cycle of the elevator space, and t represents the current sampling time.
5. The elevator passenger monitoring system based on the Internet of Things according to claim 1 is characterized in that: Determination of abnormal retention: Abnormal retention index is obtained by calculating the degree of retention through the integration of the radial velocity of the human body. , where t represents the duration, represents the radial velocity of the human body relative to the radar, Represents the target radial distance r(t) over time The dynamic changes of Indicates the normal movement threshold; the management platform monitors in real time. And infrared temperature entropy When Indicates the preset value of abnormal retention index, Indicates the infrared temperature entropy threshold, which determines abnormal retention and issues an abnormal retention warning signal.
6. The elevator passenger monitoring system based on the Internet of Things according to claim 2 is characterized by: The abnormal fall judgment: combining the micro-Doppler frequency mutation and temperature gradient sudden increase characteristics to obtain the sudden fall index ,in represents the fall feature time window, represents the average temperature gradient within the time window, Indicates different moments The instantaneous temperature gradient, Represents the micro-Doppler frequency over time The management platform monitors the changes in real time. When Indicates the preset value of the sudden fall index, which is judged as a sudden fall and issues a sudden fall warning signal.
7. The elevator passenger monitoring system based on the Internet of Things according to claim 2 is characterized by: The specific method of dynamic trigger infrared verification is to extract the energy proportion of the main frequency band of human motion through FFT. ,in represents the FFT result of the millimeter wave echo signal, Indicates the number of FFT points; the management platform monitors in real time. < When Indicates the energy percentage threshold, starts the infrared thermal imaging secondary verification and increases the sampling frame. If the secondary verification still meets < , issuing a radar signal abnormality warning signal.
8. The elevator passenger monitoring system based on the Internet of Things according to claim 1 is characterized by: The compression ratio , where C represents data redundancy; the dynamic adjustment of transmission delay ,in Indicates the original data volume, B indicates the transmission bandwidth, Indicates edge node processing delay.