An adaptive analysis method for elevator safety risks

Abstract

The present invention relates to the technical field of elevator safety risk analysis, specifically to an adaptive analysis method for elevator safety risks, which is implemented based on a multimodal data acquisition module, an edge computing layer, a risk assessment module, an emergency response module, and a data control layer. The present invention realizes real-time acquisition of elevator operating data and edge warning analysis through the collaboration of a multimodal data acquisition module and an edge computing layer, thereby shortening the data transmission and processing delay; combines short-time energy calculation, current-speed elliptical model, etc. to improve the accuracy of mechanical fault identification, generate primary emergency information, and ensure the timeliness of emergency response; the present invention uses the linkage of the risk assessment module and the emergency response module to output comprehensive status information based on multi-source data, and generates emergency instructions based on the number of consecutive occurrences of warning signals; realizes dynamic adjustment of detection frequency, automatic distribution of maintenance work orders, and emergency braking, forming a multi-level risk prevention and control system.

Description

Technical Field

[0001] The present invention relates to the technical field of elevator safety risk analysis, and in particular to an elevator safety risk adaptive analysis method. Background Art

[0002] With the acceleration of urbanization and the increasing number of high-rise buildings, elevators, as core equipment for vertical transportation, are of paramount importance for their safety. The widespread use of elevators not only solves the problem of commuting in high-rise buildings, significantly improves space utilization efficiency, and promotes rapid urbanization. Furthermore, the development of the elevator industry has driven the coordinated development of multiple upstream and downstream industries, including machinery manufacturing, electronic information, and automatic control.

[0003] An existing adaptive elevator safety risk analysis method integrates the Internet of Things, big data, and AI technologies to achieve real-time monitoring and risk assessment of elevator operating conditions. However, in practical applications, when the system needs to acquire data from multiple modalities to analyze elevator risk status, a series of problems arise. In particular, after acquiring abnormal data, the timeliness of early warning information analysis and emergency response cannot be guaranteed. For example, traditional systems may experience network delays during data transmission, resulting in delayed data arrival. During the data processing phase, complex algorithm calculations consume a significant amount of time, making it impossible to respond promptly to elevator malfunctions, potentially leading to serious safety accidents. Therefore, optimizing existing technologies to improve the timeliness of early warnings has become an urgent issue. Summary of the Invention

[0004] The present invention provides an elevator safety risk adaptive analysis method for solving the above technical problems.

[0005] A first aspect of the present invention provides an elevator safety risk adaptive analysis method, which is implemented based on a multimodal data acquisition module, an edge computing layer, a risk assessment module, an emergency response module, and a data control layer, and includes the following steps:

[0006] Step 1: The multimodal data acquisition module collects data according to the preset acquisition frequency corresponding to each modality acquisition unit, and transmits the data of each modality to the corresponding edge computing layer in real time.

[0007] Step 2: The edge computing layer corresponding to each modality performs edge analysis on the corresponding real-time data to obtain the corresponding warning signal, and executes step 3 on the warning signal.

[0008] Step 3: Conduct primary risk emergency analysis based on early warning signals to obtain corresponding primary emergency instructions.

[0009] Step 4: Each modal acquisition unit performs data acquisition and adjustment on the primary emergency instructions, and the risk control unit performs real-time risk control based on the primary emergency instructions.

[0010] Among them, the multimodal data acquisition module collects data from the elevator through the integrated data perception layer to obtain elevator operating condition data, and sends the elevator operating condition data to the data control layer; the data perception layer collects data from the elevator based on a pre-set data acquisition strategy to obtain elevator operating condition data, and the elevator operating condition data includes mechanical operation data, behavior and environment data, and basic operation data; the acquisition strategy is specifically: based on the preset analysis time interval and the system time axis, the time axis is divided according to the analysis time interval to obtain multiple analysis time periods, the current analysis time period is obtained and the corresponding benchmark data acquisition frequency is set for each type of elevator operating condition data within the analysis time period, and data acquisition is performed according to the benchmark data acquisition frequency corresponding to each type of data to obtain the corresponding data packet.

[0011] The edge computing layer obtains the elevator operating condition data for the current analysis period, and performs edge warning analysis on the elevator operating condition data collection process corresponding to the current analysis period to obtain primary emergency information, and sends the primary emergency information to the emergency response module. The edge computing layer includes a mechanical analysis unit, a behavioral environment analysis unit, a basic operation analysis unit, and a data scheduling unit. Its specific operation steps are as follows: A1, data scheduling analysis; A2, mechanical analysis unit execution analysis; A3, behavioral environment analysis unit execution analysis; A4, basic operation analysis unit execution analysis; A5, emergency analysis of warning signals;

[0012] As a further improvement of the present invention, data scheduling analysis is specifically as follows:

[0013] The data scheduling unit identifies the elevator operating condition data of the current analysis time period to obtain mechanical operation data, behavior and environment data and basic operation data, and transmits the mechanical operation data, behavior and environment data and basic operation data to the mechanical analysis unit, behavior and environment analysis unit and basic operation analysis unit respectively.

[0014] As a further improvement of the present invention, the mechanical analysis unit performs analysis, and the specific steps are:

[0015] Based on the mechanical operation data of the current analysis time period, the vibration displacement data and motor drive data are obtained; according to the vibration displacement data, the vibration accelerations corresponding to the mechanical components of the car, traction machine and guide rail are obtained, namely, the car vibration acceleration, the traction machine vibration acceleration and the guide rail vibration acceleration; multiple pre-set sampling window lengths are obtained, and the windows are sorted to obtain the window numbers corresponding to the sampling window lengths. A window number is obtained by random generation, and the corresponding number of sampling points is obtained based on the sampling window length corresponding to the generated window number. The car vibration acceleration corresponding to the corresponding number of sampling points is obtained, and the short-time energy calculation formula is used. Calculate the short-time energy E of the car vibration corresponding to the car jx ; Where N is the sampling window length, corresponding to the number of sampling points; k is the window number; n is the sampling point number; a xt Expressed as car vibration acceleration; similarly, the corresponding traction machine vibration short-time energy and guide rail vibration short-time energy are obtained based on the traction machine vibration acceleration and the guide rail vibration acceleration; when the corresponding vibration short-time energy of each mechanical component is greater than the corresponding preset energy threshold, the corresponding energy warning signal is generated;

[0016] According to the motor drive data, the effective value of the current and the traction machine speed are obtained; the traction machine speed corresponding to the effective value of the current is obtained, the effective value of the current and the traction machine speed are normalized and their values ​​are taken, and the corresponding current-speed data point (I yx ,ω), a current-speed ellipse model is established through the current-speed data points ; Among them, I yx and ω represent the normalized values ​​corresponding to the effective value of current and traction speed respectively; a is the preset coupling coefficient corresponding to current and speed; b and c correspond to; obtain (I, ω) of any measured point and substitute the measured point into the current-speed ellipse model The corresponding measured offset D is calculated; when D>1, it means that the measured point is outside the ellipse, and a pre-set standard difference threshold is obtained. When the measured offset exceeds twice the standard difference threshold, a drive abnormality signal is generated.

[0017] As a further improvement of the present invention, the behavior environment analysis unit performs analysis, and the specific steps are:

[0018] Based on the behavior and environment data of the current analysis time period, the car environment data and shaft environment data are obtained; based on the car environment data, the passenger behavior data and car environment parameters are obtained; the passenger behavior data includes the real-time number of passengers and motion image information, and the car area and rated passenger capacity are obtained based on the database. The real-time number of passengers and the rated passenger capacity are respectively calculated with the car area to obtain the real-time passenger density and the rated passenger density. When the real-time passenger density exceeds the rated passenger density, a passenger warning signal is generated; based on the motion image information, adjacent frame images are extracted and recorded as T t 、T t+1 ; and obtain the pixel coordinates corresponding to the action image information as (x, y), and obtain the light loss (u, v), and satisfy the brightness constant equation ;in, All are image gradients; the optical flow field variance calculation formula Calculate the variance of the optical flow field ; Wherein, W and H are the number of horizontal pixels and vertical pixels, respectively, representing the width and height of the optical flow field image; when the variance of the optical flow field is greater than the preset judgment threshold, a corresponding action warning signal is generated.

[0019] Based on the shaft environment data, the shaft temperature, shaft humidity, incident light intensity and transmitted light intensity are obtained; the shaft temperature is calculated by subtracting the corresponding preset temperature reference value to obtain the temperature difference value; similarly, the humidity difference value corresponding to the shaft humidity is obtained; the temperature difference value and the humidity difference value are added together to obtain the temperature and humidity coupling value; the smoke absorbance calculation formula is used The smoke absorbance AG is calculated; where A and A0 represent the transmitted light intensity and the incident light intensity, respectively. When the smoke absorbance exceeds the preset smoke absorbance threshold, a difference calculation is performed to obtain the smoke abnormality value.

[0020] The temperature and humidity coupling values ​​and smoke anomaly values ​​are normalized and their values ​​are taken, and the free energy model is used to The well abnormal value Y is calculated; where, They are represented by the normalized values ​​of smoke anomaly and temperature-humidity coupling values ​​respectively; are respectively preset weight coefficients; a preset well abnormality warning value is obtained, and a well abnormality signal is generated when the well abnormality value exceeds the well abnormality warning value.

[0021] As a further improvement of the present invention, the basic operation analysis unit performs analysis, and the specific steps are:

[0022] Based on the basic operating data of the current analysis time period, the leveling error information and the control response delay information are obtained; based on the leveling error information, the leveling error corresponding to each docking signal in the current analysis time period is obtained, and the preset error tolerance threshold is obtained. When the leveling error corresponding to the docking signal exceeds the error tolerance threshold, the corresponding docking signal is marked as an abnormal docking signal, the number of abnormal docking signals is counted by a counter, and the ratio of the number of abnormal docking signals to the total number of docking signals is calculated to obtain the abnormal docking ratio; when the abnormal docking ratio is greater than the preset abnormal docking ratio threshold, an abnormal docking signal is output.

[0023] Based on the control response delay information, the current response delay time corresponding to each instruction is obtained, and the preset rated response time is obtained based on the database. The difference between the response delay time corresponding to each instruction and the rated response time is calculated to obtain a single delay deviation. The standard deviation of the historical delay time corresponding to the preset historical time period and the allowable drift threshold are obtained through the database. The product of the standard deviation of the historical delay time and the preset confidence coefficient is ratio-calculated with the allowable drift threshold to obtain a calculated value, and the absolute value of the calculated value is taken as the window value; when When , the response warning signal is output; where ΔT yc Expressed as a single delay deviation, P py It represents the allowed drift; M represents the window value, i represents the current single delay deviation, and j represents the local index within the window.

[0024] As a further improvement of the present invention, the early warning signal is subjected to emergency analysis, and the specific steps are as follows:

[0025] The signal detection unit obtains the warning signals of each analysis unit, including energy warning signals, drive abnormality signals, action warning signals, shaft abnormality signals, parking abnormality signals and response warning signals; obtains the number of consecutive occurrences YJ corresponding to each warning signal, and obtains the corresponding preset consecutive number classification values ​​YJ1 and YJ2;

[0026] When YJ>YJ1, the monitoring data corresponding to the warning signal is obtained, and the primary emergency information is generated based on the corresponding monitoring data to increase the detection frequency;

[0027] When YJ1≤YJ<YJ2, obtain the monitoring data corresponding to the warning signal, generate primary emergency information based on the corresponding monitoring data to output warning work orders and increase the detection frequency;

[0028] When YJ>YJ2, the monitoring data corresponding to the warning signal is obtained, and the primary emergency information is generated as an emergency instruction based on the corresponding monitoring data.

[0029] The risk assessment module obtains the elevator operating condition data collected during the current analysis period, inputs the elevator operating condition data into a pre-deployed risk assessment model, outputs the comprehensive status information of the current analysis period through the risk assessment model, and sends the comprehensive status information to the emergency response module.

[0030] The emergency response module receives primary emergency information and comprehensive status information and performs emergency response execution analysis to obtain emergency instructions.

[0031] As a further improvement of the present invention, emergency response execution analysis is specifically as follows: when primary emergency information is monitored, the primary emergency information is identified to obtain the corresponding execution instruction; when the execution instruction corresponds to an increase in the detection frequency, the data collection frequency of the corresponding monitoring data in the current analysis time period is increased; when the execution instruction corresponds to outputting an early warning work order and increasing the detection frequency, the data collection frequency of the corresponding monitoring data is increased, and maintenance and repair reminder information is issued according to the corresponding early warning work order; when the execution instruction corresponds to an emergency instruction, an emergency braking instruction is output and an immediate assessment instruction is sent to the risk assessment module to obtain an immediate risk status; based on the elevator's current corresponding immediate risk status and comprehensive status information, the corresponding emergency instruction is generated as risk warning information.

[0032] As a further improvement of the present invention, the data control layer obtains the elevator operating data and stores it in a database, and obtains the early warning signal and the emergency instruction, and controls and executes the data transmission of the elevator operating data, the early warning signal and the emergency instruction.

[0033] Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects:

[0034] 1. The present invention realizes real-time collection of elevator operating data and edge warning analysis through the collaboration of multimodal data acquisition module and edge computing layer, shortening the data transmission and processing delay; combined with short-time energy calculation, current-speed ellipse model, etc., it improves the accuracy of mechanical fault identification, generates primary emergency information, and ensures the timeliness of emergency response.

[0035] 2. The present invention uses the linkage between the risk assessment module and the emergency response module to output comprehensive status information based on multi-source data, and generates emergency instructions in a hierarchical manner based on the number of consecutive occurrences of early warning signals; it realizes dynamic adjustment of detection frequency, automatic distribution of maintenance work orders and emergency braking, and forms a multi-level risk prevention and control system. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for the description of the embodiments. The following drawings are not intentionally scaled to the actual size, and the focus is on illustrating the main purpose of the present application.

[0037] Figure 1Schematic diagram of each module of an elevator safety risk adaptive analysis method of the present invention;

[0038] Figure 2 This is an analysis flow chart of the edge computing layer of an elevator safety risk adaptive analysis method of the present invention;

[0039] Figure 3 The present invention is a flowchart of an elevator safety risk adaptive analysis method. DETAILED DESCRIPTION

[0040] 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 any creative efforts shall fall within the scope of protection of the present invention.

[0041] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1-3 In an embodiment of the present invention, an elevator safety risk adaptive analysis method is implemented based on a multimodal data acquisition module, an edge computing layer, a risk assessment module, an emergency response module, and a data control layer, and includes the following steps:

[0042] Step 1: Multimodal data acquisition and transmission: The multimodal data acquisition module collects data according to the preset acquisition frequency corresponding to each modality acquisition unit, and transmits the data of each modality to the corresponding edge computing layer in real time.

[0043] Step 2: Edge computing analysis: The edge computing layer corresponding to each modality performs edge analysis on the corresponding real-time data to obtain the corresponding warning signal, and executes step 3 on the warning signal; the edge side completes the feature extraction and anomaly screening of multimodal data in real time, reducing cloud dependence.

[0044] Step 3: Primary risk emergency feedback execution: Conduct primary risk emergency analysis based on early warning signals to obtain corresponding primary emergency instructions; implement a hierarchical response mechanism based on the frequency of early warning signals to achieve precise and dynamic emergency handling.

[0045] Step 4: Primary Emergency Execution: Each modal acquisition unit collects and adjusts data based on the primary emergency instructions, and the risk control unit performs real-time risk control based on these instructions. Upon receiving a command to increase detection frequency, the acquisition unit dynamically adjusts sampling parameters through edge nodes, such as switching the current sensor from periodic sampling to event-triggered sampling. Upon receiving an emergency command, the risk control unit requests an immediate assessment from the risk assessment module. For example, if an elevator suddenly overspeeds, this step completes braking within 50ms and generates a risk report containing 10 sets of real-time data, triggering the fire protection system to shut off unsafe power supplies.

[0046] The multimodal data acquisition module collects elevator data through an integrated data perception layer to obtain elevator operating data, and sends the elevator operating data to the data control layer. The data perception layer includes but is not limited to multi-source acquisition devices such as vibration sensors, current sensors, and video surveillance equipment. The data perception layer collects elevator data based on a pre-set data acquisition strategy to obtain elevator operating data. The elevator operating data includes mechanical operation data, behavior and environment data, and basic operation data.

[0047] The specific collection strategy is as follows: based on the preset analysis time interval and the system's time axis, the time axis is divided according to the analysis time interval to obtain multiple analysis time periods, the current analysis time period is obtained, and the corresponding benchmark data collection frequency is set for each type of elevator operating condition data within the analysis time period, and data collection is performed according to the benchmark data collection frequency corresponding to each type of data to obtain the corresponding data packet; the multimodal data acquisition module collects data at a preset frequency through vibration, current and other sensors and video equipment, and divides the analysis time period to set the benchmark frequency; ensures the synchronous collection of multi-source data to provide comprehensive real-time data support for subsequent analysis.

[0048] The edge computing layer obtains elevator operating data for the current analysis period and performs edge warning analysis on the data collected during the current analysis period to obtain primary emergency information. This information is then sent to the emergency response module. After the edge computing layer obtains the data, it is processed by mechanical and behavioral environment analysis units and the data scheduling unit. This enables data classification and edge analysis, shortening the warning cycle.

[0049] The edge computing layer includes a mechanical analysis unit, a behavioral environment analysis unit, a basic operation analysis unit, and a data scheduling unit. Its specific operation steps are as follows:

[0050] A1. Data Scheduling and Analysis: The data scheduling unit identifies the elevator operating data for the current analysis period to obtain mechanical operation data, behavior and environment data, and basic operation data. The data scheduling unit transmits the mechanical operation data, behavior and environment data, and basic operation data to the mechanical analysis unit, behavior and environment analysis unit, and basic operation analysis unit, respectively. The data scheduling unit identifies the three types of data and transmits them to the corresponding analysis units, ensuring accurate data diversion and improving the analysis efficiency of each module.

[0051] A2. The mechanical analysis unit is specifically as follows: based on the mechanical operation data of the current analysis time period, vibration displacement data and motor drive data are obtained; based on the vibration displacement data, the vibration accelerations corresponding to the mechanical components of the car, traction machine and guide rail are obtained, namely, the car vibration acceleration, the traction machine vibration acceleration and the guide rail vibration acceleration; multiple pre-set sampling window lengths are obtained, and the windows are sorted to obtain the window numbers corresponding to the sampling window lengths. A window number is obtained by random generation, and the corresponding number of sampling points is obtained based on the sampling window length corresponding to the generated window number. The car vibration acceleration corresponding to the corresponding number of sampling points is obtained, and the short-time energy calculation formula is used. Calculate the short-time energy E of the car vibration corresponding to the car jx ; Where N is the sampling window length, corresponding to the number of sampling points; k is the window number; n is the sampling point number; a xt Expressed as car vibration acceleration; similarly, the corresponding traction machine vibration short-time energy and guide rail vibration short-time energy are obtained based on the traction machine vibration acceleration and the guide rail vibration acceleration; when the corresponding vibration short-time energy of each mechanical component is greater than the corresponding preset energy threshold, the corresponding energy warning signal is generated;

[0052] According to the motor drive data, the effective value of the current and the traction machine speed are obtained; the traction machine speed corresponding to the effective value of the current is obtained, the effective value of the current and the traction machine speed are normalized and their values ​​are taken, and the corresponding current-speed data point (I yx ,ω), a current-speed ellipse model is established through the current-speed data points ; Among them, I yx and ω represent the normalized values ​​corresponding to the effective value of current and traction speed respectively; a is the preset coupling coefficient corresponding to current and speed; b and c correspond to; obtain (I, ω) of any measured point and substitute the measured point into the current-speed ellipse model The corresponding measured offset D is calculated; when D>1, it means that the measured point is outside the ellipse, and a pre-set standard difference threshold is obtained. When the measured offset exceeds twice the standard difference threshold, a drive abnormality signal is generated.

[0053] A3. The behavior and environment analysis unit specifically comprises the following steps: obtaining car environment data and shaft environment data based on the behavior and environment data of the current analysis time period; obtaining passenger behavior data and car environment parameters based on the car environment data; obtaining the car area and rated passenger capacity based on the database for the passenger behavior data including the real-time number of passengers and motion image information; performing density calculations on the real-time number of passengers and the rated passenger capacity with the car area to obtain the real-time passenger density and the rated passenger density; and generating a passenger warning signal when the real-time passenger density exceeds the rated passenger density; extracting adjacent frame images based on the motion image information and recording them as T t 、T t+1 ; and obtain the pixel coordinates corresponding to the action image information as (x, y), and obtain the light loss (u, v), and satisfy the brightness constant equation ;in, All are image gradients; the optical flow field variance calculation formula Calculate the variance of the optical flow field ; Wherein, W and H are the number of horizontal pixels and vertical pixels, respectively, representing the width and height of the optical flow field image; when the variance of the optical flow field is greater than the preset judgment threshold, a corresponding action warning signal is generated.

[0054] Based on the shaft environment data, the shaft temperature, shaft humidity, incident light intensity and transmitted light intensity are obtained; the shaft temperature is calculated by subtracting the corresponding preset temperature reference value to obtain the temperature difference value; similarly, the humidity difference value corresponding to the shaft humidity is obtained; the temperature difference value and the humidity difference value are added together to obtain the temperature and humidity coupling value; the smoke absorbance calculation formula is used The smoke absorbance AG is calculated; where A and A0 represent the transmitted light intensity and the incident light intensity, respectively. When the smoke absorbance exceeds the preset smoke absorbance threshold, a difference calculation is performed to obtain the smoke abnormality value.

[0055] The temperature and humidity coupling values ​​and smoke anomaly values ​​are normalized and their values ​​are taken, and the free energy model is used to The well abnormal value Y is calculated; where, They are represented by the normalized values ​​of smoke anomaly and temperature-humidity coupling values ​​respectively; are respectively preset weight coefficients; a preset well abnormality warning value is obtained, and a well abnormality signal is generated when the well abnormality value exceeds the well abnormality warning value.

[0056] A4. The basic operation analysis unit specifically comprises the following steps: obtaining leveling error information and control response delay information based on the basic operation data of the current analysis time period; obtaining the leveling error corresponding to each docking signal in the current analysis time period based on the leveling error information, and obtaining a preset error tolerance threshold; when the leveling error corresponding to the docking signal exceeds the error tolerance threshold, marking the corresponding docking signal as an abnormal docking signal; counting the number of abnormal docking signals through a counter; and calculating the ratio of the number of abnormal docking signals to the total number of docking signals to obtain an abnormal docking ratio; and outputting an abnormal docking signal when the abnormal docking ratio is greater than the preset abnormal docking ratio threshold.

[0057] Based on the control response delay information, the current response delay time corresponding to each instruction is obtained, and the preset rated response time is obtained based on the database. The difference between the response delay time corresponding to each instruction and the rated response time is calculated to obtain a single delay deviation. The standard deviation of the historical delay time corresponding to the preset historical time period and the allowable drift threshold are obtained through the database. The product of the standard deviation of the historical delay time and the preset confidence coefficient is ratio-calculated with the allowable drift threshold to obtain a calculated value, and the absolute value of the calculated value is taken as the window value; when When , the response warning signal is output; where ΔT yc Expressed as a single delay deviation, P py It represents the allowed drift; M represents the window value, i represents the current single delay deviation, and j represents the local index within the window.

[0058] A5. Perform emergency analysis on warning signals: Obtain warning signals from each analysis unit through the signal detection unit. Warning signals include energy warning signals, drive abnormality signals, motion warning signals, shaft abnormality signals, parking abnormality signals, and response warning signals. Obtain the number of consecutive occurrences YJ corresponding to each warning signal, and obtain the corresponding preset consecutive number classification values ​​YJ1 and YJ2;

[0059] When YJ>YJ1, the monitoring data corresponding to the warning signal is obtained, and the primary emergency information is generated based on the corresponding monitoring data to increase the detection frequency;

[0060] When YJ1≤YJ<YJ2, obtain the monitoring data corresponding to the warning signal, generate primary emergency information based on the corresponding monitoring data to output warning work orders and increase the detection frequency;

[0061] When YJ>YJ2, the monitoring data corresponding to the warning signal is obtained, and the primary emergency information is generated as an emergency instruction based on the corresponding monitoring data.

[0062] The risk assessment module obtains the elevator operating condition data collected during the current analysis period, inputs the elevator operating condition data into a pre-deployed risk assessment model, outputs the comprehensive status information of the current analysis period through the risk assessment model, and sends the comprehensive status information to the emergency response module.

[0063] The emergency response module receives primary emergency information and comprehensive status information and performs emergency response execution analysis to obtain emergency instructions.

[0064] Emergency response execution analysis, which is specifically as follows: when primary emergency information is monitored, the primary emergency information is identified to obtain the corresponding execution instructions; when the execution instruction corresponds to an increase in the detection frequency, the data collection frequency of the corresponding monitoring data in the current analysis time period is increased; when the execution instruction corresponds to outputting an early warning work order and increasing the detection frequency, the data collection frequency of the corresponding monitoring data is increased, and maintenance and repair reminder information is issued according to the corresponding early warning work order; when the execution instruction corresponds to an emergency instruction, an emergency braking instruction is output and an immediate assessment instruction is sent to the risk assessment module to obtain an immediate risk status; based on the elevator's current corresponding immediate risk status and comprehensive status information, the corresponding emergency instruction is generated as risk warning information.

[0065] The data control layer obtains the elevator operating data and stores it in the database, and obtains the early warning signal and emergency instructions, and controls and executes the data transmission of the elevator operating data, early warning signals and emergency instructions.

[0066] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An elevator safety risk adaptive analysis method, characterized in that: The steps include: Step 1: The multimodal data acquisition module collects data from each modality acquisition unit according to the preset acquisition frequency, and transmits the data of each modality to the corresponding edge computing layer in real time; Step 2: The edge computing layer corresponding to each modality performs edge analysis on the corresponding real-time data to obtain the corresponding warning signal, and executes step 3 on the warning signal; Step 3: Conduct primary risk emergency analysis based on early warning signals to obtain corresponding primary emergency instructions; Step 4: Each modal acquisition unit collects and adjusts data on the primary emergency instructions, and the risk control unit performs real-time risk control based on the primary emergency instructions; The multimodal data acquisition module collects elevator data through the integrated data perception layer to obtain elevator operating data, and sends the elevator operating data to the data control layer; The edge computing layer obtains the elevator operating condition data for the current analysis period, and performs edge warning analysis on the elevator operating condition data collection process corresponding to the current analysis period to obtain primary emergency information, and sends the primary emergency information to the emergency response module. The edge computing layer includes a mechanical analysis unit, a behavioral environment analysis unit, a basic operation analysis unit, and a data scheduling unit. Its specific operation steps are as follows: A1, data scheduling analysis; A2, mechanical analysis unit execution analysis; A3, behavioral environment analysis unit execution analysis; A4, basic operation analysis unit execution analysis; A5, emergency analysis of warning signals; The mechanical analysis unit performs analysis, and the specific steps are: Based on the mechanical operation data of the current analysis time period, vibration displacement data and motor drive data are obtained; according to the vibration displacement data, the vibration accelerations corresponding to the mechanical components of the car, traction machine and guide rail are obtained, namely, the car vibration acceleration, the traction machine vibration acceleration and the guide rail vibration acceleration; multiple pre-set sampling window lengths are obtained, and the windows are sorted to obtain the window numbers corresponding to the sampling window lengths, a window number is obtained by random generation, and the corresponding number of sampling points is obtained based on the sampling window length corresponding to the generated window number, and the car vibration acceleration corresponding to the corresponding number of sampling points is obtained, and the corresponding car vibration short-time energy of the car is obtained by calculation through the short-time energy calculation formula; similarly, the corresponding traction machine vibration short-time energy and guide rail vibration short-time energy are obtained based on the traction machine vibration acceleration and the guide rail vibration acceleration; when the corresponding vibration short-time energy of each mechanical component is greater than the corresponding preset energy threshold, a corresponding energy warning signal is generated; Get the effective value of current and traction machine speed according to the motor drive data; Obtain the traction speed of the traction machine corresponding to the effective current value, normalize the effective current value and the traction speed and take their numerical values, and obtain the corresponding current-speed data points. Establish a current-speed ellipse model based on the current-speed data points; obtain the data point of any measured point and substitute it into the current-speed ellipse model to calculate the corresponding measured offset; when the measured offset is greater than one, it means that the measured point is outside the ellipse, obtain a pre-set standard difference threshold, and generate a drive abnormality signal when the measured offset exceeds twice the standard difference threshold; The behavior environment analysis unit performs analysis, and the specific steps are: Based on the behavior and environment data of the current analysis time period, the cabin environment data and the shaft environment data are obtained; based on the cabin environment data, the passenger behavior data and cabin environment parameters are obtained; the passenger behavior data includes the real-time number of passengers and motion image information; the cabin area and rated passenger capacity are obtained based on the database; the real-time number of passengers and the rated passenger capacity are respectively calculated with the cabin area to obtain the real-time passenger density and the rated passenger density; when the real-time passenger density exceeds the rated passenger density, a passenger warning signal is generated; based on the motion image information, adjacent frame images are extracted; and the pixel coordinates and light loss corresponding to the motion image information are obtained. The optical flow field variance is calculated using the optical flow field variance calculation formula; when the optical flow field variance is greater than the preset judgment threshold, a corresponding action warning signal is generated; Based on the shaft environment data, the shaft temperature, shaft humidity, incident light intensity, and transmitted light intensity are obtained; the shaft temperature is subtracted from the corresponding preset temperature reference value to obtain a temperature difference value; similarly, the humidity difference value corresponding to the shaft humidity is obtained; the temperature difference value and the humidity difference value are added together to obtain a temperature and humidity coupling value; The smoke absorbance calculation formula is used to calculate the smoke absorbance; when the smoke absorbance exceeds the preset smoke absorbance threshold, a difference calculation is performed to obtain a smoke abnormality value; the temperature and humidity coupling value and the smoke abnormality value are normalized and their values ​​are taken, and the wellbore abnormality value is calculated using the free energy model; a preset wellbore abnormality warning value is obtained, and a wellbore abnormality signal is generated when the wellbore abnormality value exceeds the wellbore abnormality warning value; The risk assessment module obtains the elevator operating condition data collected during the current analysis period, inputs the elevator operating condition data into a pre-deployed risk assessment model, outputs comprehensive status information for the current analysis period through the risk assessment model, and sends the comprehensive status information to the emergency response module; The emergency response module receives primary emergency information and comprehensive status information and performs emergency response execution analysis to obtain emergency instructions; The data control layer obtains the elevator operating data and stores it in the database, and obtains the early warning signal and emergency instructions, and controls and executes the data transmission of the elevator operating data, early warning signals and emergency instructions; The specific process of the data scheduling analysis is as follows: the data scheduling unit identifies the elevator operating condition data of the current analysis time period to obtain mechanical operation data, behavior and environment data and basic operation data, and transmits the mechanical operation data, behavior and environment data and basic operation data to the mechanical analysis unit, behavior and environment analysis unit and basic operation analysis unit respectively.

2. The elevator safety risk adaptive analysis method according to claim 1, characterized in that: The basic operation analysis unit performs analysis, and the specific steps are: Based on the basic operating data of the current analysis time period, the leveling error information and control response delay information are obtained; based on the leveling error information, the leveling error corresponding to each docking signal in the current analysis time period is obtained, and a preset error tolerance threshold is obtained. When the leveling error corresponding to the docking signal exceeds the error tolerance threshold, the corresponding docking signal is marked as an abnormal docking signal, the number of abnormal docking signals is counted by a counter, and the abnormal docking ratio is calculated by calculating the ratio of the number of abnormal docking signals to the total number of docking signals; When the abnormal stop ratio is greater than the preset abnormal stop ratio threshold, a stop abnormality signal is output; Based on the control response delay information, the current response delay time corresponding to each instruction is obtained, and the preset rated response time is obtained based on the database. The difference between the response delay time corresponding to each instruction and the rated response time is calculated to obtain a single delay deviation. The standard deviation of the historical delay time corresponding to the preset historical time period and the allowable drift threshold are obtained from the database. The product of the standard deviation of the historical delay time and the preset confidence coefficient is ratio-calculated with the allowable drift threshold to obtain a calculated value, and the absolute value of the calculated value is taken as the window value; A response warning signal will be output.

3. The elevator safety risk adaptive analysis method according to claim 2, characterized in that: Conduct emergency analysis on the warning signal, and the specific steps are as follows: The signal detection unit obtains the warning signals of each analysis unit, including energy warning signals, drive abnormality signals, action warning signals, shaft abnormality signals, parking abnormality signals and response warning signals; obtains the number of consecutive occurrences YJ corresponding to each warning signal, and obtains the corresponding preset consecutive number classification values ​​YJ1 and YJ2; When YJ>YJ1, the monitoring data corresponding to the warning signal is obtained, and the primary emergency information is generated based on the corresponding monitoring data to increase the detection frequency; When YJ1≤YJ<YJ2, obtain the monitoring data corresponding to the warning signal, generate primary emergency information based on the corresponding monitoring data to output warning work orders and increase the detection frequency; When YJ>YJ2, the monitoring data corresponding to the warning signal is obtained, and the primary emergency information is generated as an emergency instruction based on the corresponding monitoring data.

4. The elevator safety risk adaptive analysis method according to claim 3, characterized in that: The specific process of the emergency response execution analysis is as follows: when primary emergency information is monitored, the primary emergency information is identified to obtain the corresponding execution instruction; when the execution instruction corresponds to increasing the detection frequency, the data collection frequency of the corresponding monitoring data in the current analysis time period is increased; when the execution instruction corresponds to outputting an early warning work order and increasing the detection frequency, the data collection frequency of the corresponding monitoring data is increased, and maintenance and repair reminder information is issued according to the corresponding early warning work order; when the execution instruction corresponds to an emergency instruction, an emergency braking instruction is output and an immediate assessment instruction is sent to the risk assessment module to obtain an immediate risk status; Based on the elevator's current immediate risk status and comprehensive status information, corresponding emergency instructions are generated as risk warning information.

5. The elevator safety risk adaptive analysis method according to claim 1, characterized in that: The data perception layer collects data from the elevator based on a pre-set data collection strategy to obtain elevator operating condition data. The collection strategy is specifically: based on a preset analysis time interval and the system time axis, the time axis is divided according to the analysis time interval to obtain multiple analysis time periods, the current analysis time period is obtained and the corresponding benchmark data collection frequency is set for each type of elevator operating condition data within the analysis time period, and data collection is performed according to the benchmark data collection frequency corresponding to each type of data to obtain the corresponding data packet.

Images