Elevator internal device fault intelligent inspection management system and management method

Abstract

The invention discloses an elevator internal device fault intelligent inspection management system and management method. The system comprises a sensing layer, a transmission layer, a platform layer and an application layer which communicate in sequence. The sensing layer collects operation data of devices such as an elevator traction machine and a portal crane system through special sensors such as a vibration sensor and a temperature sensor and a camera. The transmission layer encrypts transmission data in a mixed transmission mode after edge calculation preprocessing; the platform layer adopts double databases to store data, and fusion analysis is performed through a convolutional neural network model and a long-short-term memory network model; the application layer provides exclusive functions for multiple subjects. The management method forms a closed loop through data acquisition, processing, diagnosis, operation and maintenance. According to the method, the fault identification accuracy is over 96 percent, the response time is shortened to be within 30 minutes, 45 percent of sudden faults are reduced, accurate decision is supported, the safety is improved, the operation and maintenance cost is reduced, and the applicability is high.

Description

Technical Field

[0001] This invention relates to the field of elevators, specifically to an intelligent inspection and management system and method for faults in internal elevator components. Background Technology

[0002] With the acceleration of urbanization, elevators have become an indispensable vertical transportation tool in high-rise buildings, and their operational safety is directly related to the safety of people's lives and property. Currently, the fault inspection and management of elevator internal components mostly rely on traditional manual methods, which have significant drawbacks: First, the inspection efficiency is low. Maintenance personnel need to check each elevator on-site, and the operating status of core components such as traction machines, door operators, and safety brakes can only be observed by the naked eye and measured by simple instruments, which cannot achieve real-time monitoring. Moreover, manual inspection cycles are long and prone to omissions.

[0003] Secondly, fault identification is lagging and inaccurate. Traditional methods rely on the experience of maintenance personnel to make judgments, making it difficult to predict potential faults such as abnormal vibration and temperature rise in advance. Often, the fault is dealt with passively only after it occurs, leading to frequent problems such as elevator entrapment and expanded damage to components, resulting in a high rate of fault misjudgment.

[0004] Third, the operation and maintenance management is chaotic. The basic elevator information and maintenance files recorded manually are easily lost or made incorrect, making it impossible to form a full life cycle management link. Regulatory authorities also find it difficult to accurately grasp the regional elevator safety situation.

[0005] Meanwhile, the contradiction between the surge in elevator ownership and the shortage of maintenance personnel is becoming increasingly prominent, and traditional models can no longer meet the needs of refined management. Against this backdrop, there is an urgent need for an intelligent inspection and management solution that integrates sensing, transmission, platform, and application layer technologies to solve existing management challenges. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to overcome the defects of the above-mentioned technologies and provide an intelligent inspection and management system and management method for faults in elevator internal components.

[0007] To address the aforementioned technical problems, the present invention provides an intelligent inspection and management system and method for elevator internal component faults: The intelligent inspection and management system for elevator internal component faults includes a sensing layer, a transmission layer, a platform layer, and an application layer. The sensing layer, transmission layer, platform layer, and application layer are sequentially communicatively connected, and each layer collaborates to collect, transmit, analyze, process, and display operational data of elevator internal components. The sensing layer is deployed inside the elevator car and shaft to collect operating parameters and status information of elevator internal components. The transmission layer transmits the data collected by the sensing layer to the platform layer. The platform layer stores, analyzes, and provides business support for the received data. The application layer provides dedicated functional services for different user groups.

[0008] As an improvement, the sensing layer includes multiple dedicated sensors, a camera, and a sound acquisition module. The dedicated sensors include vibration sensors, temperature sensors, displacement sensors, current sensors, and pressure sensors. The vibration and temperature sensors are installed on the elevator traction machine to monitor the operating vibration frequency and winding temperature of the traction machine. The displacement and current sensors are installed on the elevator door operator system to capture the door operator's opening and closing speed and motor current changes. The pressure sensor is installed on the elevator safety brake and buffer to monitor the triggering status of safety devices. The camera is used to acquire image information of the car interior and door body. The sound acquisition module is used to acquire mechanical sound information during elevator operation.

[0009] As an improvement, the transmission layer adopts a hybrid transmission mode, which includes an edge computing module, a fourth-generation mobile communication network, a fifth-generation mobile communication network, and a wireless local area network. The edge computing module is used to perform local preprocessing and redundant data filtering on the data collected by the perception layer. The fourth-generation mobile communication network and the fifth-generation mobile communication network are used to achieve long-distance data transmission. The wireless local area network is used to achieve short-distance interconnection of data between devices in the elevator machine room. The transmission layer also includes an encryption unit and a breakpoint resumption unit. The encryption unit is used to encrypt the transmitted data, and the breakpoint resumption unit is used to avoid data loss caused by network interruption.

[0010] As an improvement, the platform layer includes a data storage module, an intelligent analysis module, and a business support module. The data storage module adopts a combined architecture of relational database and time-series database. The relational database is used to store structured data such as elevator basic information and maintenance records, while the time-series database is used to store real-time data collected by sensors. The intelligent analysis module includes a convolutional neural network model and a long short-term memory network model. The convolutional neural network model is used to process image data to identify visual faults, while the long short-term memory network model is used to analyze time-series data to predict fault trends. The business support module provides access control, log management, and interface management functions.

[0011] As an improvement, the application layer includes a maintenance personnel function module, a user unit function module, and a regulatory department function module; the maintenance personnel function module includes a fault early warning push unit, a work order automatic generation unit, and a maintenance progress tracking unit; the user unit function module includes an operation status display unit, a fault statistics report unit, and a maintenance quality assessment unit; the regulatory department function module includes a regional safety situation analysis unit and a hidden danger investigation and supervision unit; the application layer supports access operations on web and mobile terminals.

[0012] The intelligent inspection and management method for elevator internal component faults includes a data acquisition stage, a data processing stage, a fault diagnosis stage, and an operation and maintenance management stage, with each stage sequentially connected to form a closed-loop management process. The data acquisition stage obtains operating data and status information of the elevator's internal components through a sensing layer device. The data processing stage transmits the acquired data to the platform layer via a transmission layer for storage and preprocessing. The fault diagnosis stage analyzes the data using an intelligent analysis module to identify faults and predict risks. The operation and maintenance management stage conducts fault handling and full-process operation and maintenance work based on the diagnostic results.

[0013] As an improvement, the data processing stage includes a data standardization processing step and a hierarchical storage step; the data standardization processing step is used to clarify the data acquisition range, accuracy requirements and transmission format of each type of sensor to ensure data consistency; the hierarchical storage step is used to adopt differentiated storage strategies for real-time data and historical data, with real-time data stored in a time-series database and historical data compressed and transferred to an archive database, the data storage cycle of which meets the regulatory requirements for special equipment.

[0014] As an improvement, the fault diagnosis stage includes a fault classification diagnosis step and a differentiated early warning step. The fault classification diagnosis step divides faults into Level 1, Level 2, and Level 3 early warnings. Level 1 early warnings correspond to minor faults where parameters deviate slightly from the threshold; Level 2 early warnings correspond to serious faults where parameters exceed limits or devices malfunction; and Level 3 early warnings correspond to potential fault trends discovered through time-series analysis. The differentiated early warning step uses different push methods for different levels of early warnings. Level 1 early warnings are pushed via mobile application messages; Level 2 early warnings are pushed via a combination of mobile application messages and telephone notifications; and Level 3 early warnings trigger preventative maintenance by the maintenance unit.

[0015] As an improvement, the operation and maintenance management phase includes a closed-loop fault handling step and a full-process operation and maintenance record step. The closed-loop fault handling step includes early warning triggering, work order dispatching, fault handling, repair verification, and review and optimization. After the early warning is triggered, the system automatically matches the maintenance responsibility unit and personnel and completes the work order dispatching. After the maintenance personnel complete the repair, they upload the voucher. The system verifies the fault resolution by collecting device operation data and regularly reviews the fault handling situation. The full-process operation and maintenance record step is used to record the information of the entire process node such as work order dispatching, response, and completion, forming an elevator full life cycle operation and maintenance file.

[0016] As an improvement, a system security management step is also included, which includes physical security protection, network security protection, and data security protection. The physical security protection adopts anti-tampering and anti-damage design for the perception layer devices, and triggers an alarm when the devices are abnormally disassembled. The network security protection restricts unauthorized access by deploying firewalls and intrusion detection systems and regularly updates security policies. The data security protection adopts data encryption, access control, and operation log auditing measures to ensure that data is only accessed by authorized personnel and that operations are traceable.

[0017] The advantages of this invention compared to existing technologies are as follows: This technical solution achieves the following core benefits through the coordinated operation of the sensing layer, transmission layer, platform layer, and application layer: First, it significantly improves the safety of elevator operation. The sensing layer uses dedicated sensors such as vibration sensors and temperature sensors, along with cameras, to collect comprehensive data. The platform layer integrates and analyzes data using convolutional neural network models and long short-term memory network models, achieving a fault identification accuracy rate of over 96%. The transmission layer transmits data at extremely high speed within 500ms. The application layer provides differentiated early warnings, reducing the fault response time from 2 hours to within 30 minutes, thus significantly reducing the risk of people being trapped and equipment damage.

[0018] Second, the system optimizes operation and maintenance and reduces costs. The system pushes out a three-level early warning 72 hours in advance to enable preventive maintenance and reduce sudden failures by 45%. The maintenance personnel function module enables digital management of work orders, reducing paperwork by 70%. One person can manage more than 50 elevators, significantly reducing labor costs.

[0019] Third, it supports precise decision-making by multiple stakeholders, provides quantitative data such as operational reliability to users, generates safety heat maps for regulatory authorities to achieve precise supervision, and provides fault pattern data for maintenance units to optimize services.

[0020] Fourth, the system is stable and reliable. The advanced encryption standard algorithm at the transport layer and the breakpoint resume function ensure data security. The cloud server cluster is adapted to high concurrency, and the wide voltage sensor is adapted to complex environments, making it suitable for a wide range of applications. Attached Figure Description

[0021] Figure 1 This is a system framework diagram of the present invention.

[0022] Figure 2 This is the process management diagram of the present invention. Detailed Implementation

[0023] To facilitate understanding of this application, a more complete description will be provided below with reference to the accompanying drawings, which illustrate embodiments of the present application. However, the present application can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that the disclosure of this application will be thorough and complete.

[0024] Referring to the attached diagram, an intelligent inspection and management system and method for elevator internal component faults are disclosed. The intelligent inspection and management system includes a sensing layer, a transmission layer, a platform layer, and an application layer. These layers are sequentially connected and communicate with each other, collaboratively collecting, transmitting, analyzing, processing, and displaying operational data of the elevator internal components. The sensing layer is deployed inside the elevator car and shaft to collect operating parameters and status information of the elevator internal components. The transmission layer transmits the data collected by the sensing layer to the platform layer. The platform layer stores, analyzes, and provides business support for the received data. The application layer provides dedicated functional services for different user groups.

[0025] The sensing layer includes multiple dedicated sensors, cameras, and a sound acquisition module. The dedicated sensors include vibration sensors, temperature sensors, displacement sensors, current sensors, and pressure sensors. The vibration and temperature sensors are installed on the elevator traction machine to monitor its operating vibration frequency and winding temperature. The displacement and current sensors are installed on the elevator door operator system to capture the door operator's opening and closing speed and motor current changes. The pressure sensors are installed on the elevator safety brake and buffer to monitor the triggering status of safety devices. The cameras are used to acquire image information of the elevator car interior and doors. The sound acquisition module is used to acquire mechanical sound information during elevator operation.

[0026] The transmission layer adopts a hybrid transmission mode, which includes an edge computing module, a fourth-generation mobile communication network, a fifth-generation mobile communication network, and a wireless local area network. The edge computing module is used to perform local preprocessing and redundant data filtering on the data collected by the sensing layer. The fourth-generation mobile communication network and the fifth-generation mobile communication network are used to achieve long-distance data transmission. The wireless local area network is used to achieve short-distance interconnection of data between devices in the elevator machine room. The transmission layer also includes an encryption unit and a breakpoint resumption unit. The encryption unit is used to encrypt the transmitted data, and the breakpoint resumption unit is used to prevent data loss due to network interruption.

[0027] The platform layer includes a data storage module, an intelligent analysis module, and a business support module. The data storage module adopts a combined architecture of relational database and time-series database. The relational database is used to store structured data such as elevator basic information and maintenance records, while the time-series database is used to store real-time data collected by sensors. The intelligent analysis module includes a convolutional neural network model and a long short-term memory network model. The convolutional neural network model is used to process image data to identify visual faults, while the long short-term memory network model is used to analyze time-series data to achieve fault trend prediction. The business support module provides access control, log management, and interface management functions.

[0028] The application layer includes a maintenance personnel function module, a user unit function module, and a regulatory department function module. The maintenance personnel function module includes a fault early warning push unit, a work order automatic generation unit, and a maintenance progress tracking unit. The user unit function module includes an operation status display unit, a fault statistical report unit, and a maintenance quality assessment unit. The regulatory department function module includes a regional safety situation analysis unit and a hidden danger investigation and supervision unit. The application layer supports access from both web and mobile devices.

[0029] The intelligent inspection and management method for elevator internal component faults includes a data acquisition stage, a data processing stage, a fault diagnosis stage, and an operation and maintenance management stage, with each stage sequentially connected to form a closed-loop management process. The data acquisition stage obtains operating data and status information of the elevator's internal components through a sensing layer device. The data processing stage transmits the acquired data to the platform layer via a transmission layer for storage and preprocessing. The fault diagnosis stage analyzes the data using an intelligent analysis module to identify faults and predict risks. The operation and maintenance management stage conducts fault handling and full-process operation and maintenance work based on the diagnostic results.

[0030] The data processing stage includes a data standardization step and a hierarchical storage step. The data standardization step is used to clarify the data acquisition range, accuracy requirements and transmission format of each type of sensor to ensure data consistency. The hierarchical storage step is used to adopt differentiated storage strategies for real-time data and historical data. Real-time data is stored in a time-series database, and historical data is compressed and transferred to an archive database. The data storage period of the archive database meets the regulatory requirements for special equipment.

[0031] The fault diagnosis phase includes a fault classification diagnosis step and a differentiated early warning step. The fault classification diagnosis step divides faults into Level 1, Level 2, and Level 3 early warnings. Level 1 early warnings correspond to minor faults where parameters deviate slightly from the threshold; Level 2 early warnings correspond to serious faults where parameters exceed limits or devices malfunction; and Level 3 early warnings correspond to potential fault trends discovered through time-series analysis. The differentiated early warning step uses different push methods for different levels of early warnings. Level 1 early warnings are pushed via mobile application messages; Level 2 early warnings are pushed via a combination of mobile application messages and telephone notifications; and Level 3 early warnings trigger preventative maintenance by the maintenance unit.

[0032] The operation and maintenance management phase includes closed-loop fault handling steps and full-process operation and maintenance recording steps. The closed-loop fault handling steps include early warning triggering, work order dispatching, fault handling, repair verification, and review and optimization. After the early warning is triggered, the system automatically matches the maintenance responsibility unit and personnel and completes the work order dispatching. After the maintenance personnel complete the repair, they upload the voucher. The system verifies the fault resolution by collecting device operation data and regularly reviews the fault handling situation. The full-process operation and maintenance recording steps are used to record the information of the entire process nodes such as work order dispatching, response, and completion, forming an elevator full life cycle operation and maintenance file.

[0033] It also includes system security management steps, which include physical security protection, network security protection, and data security protection. The physical security protection adopts anti-tampering and anti-damage design for the perception layer devices, and triggers an alarm when the devices are abnormally disassembled. The network security protection restricts unauthorized access by deploying firewalls and intrusion detection systems and regularly updates security policies. The data security protection adopts data encryption, access control, and operation log auditing measures to ensure that data is only accessed by authorized personnel and that operations are traceable.

[0034] Specific implementation of the intelligent inspection and management system for elevator internal component faults:

[0035] The intelligent inspection and management system for elevator internal component faults described in this embodiment includes a perception layer, a transmission layer, a platform layer, and an application layer. Each layer achieves data communication and command interaction through a standardized communication protocol, forming a complete intelligent inspection and management chain for elevator faults.

[0036] Implementation configuration of the perception layer:

[0037] The sensing layer, as the core front-end of the system's data acquisition, is deployed inside the elevator car, shaft, and at the installation locations of key components. Its core function is to accurately capture the operating parameters and status information of the elevator's internal components, providing raw data support for subsequent fault diagnosis. In this embodiment, the specific configuration of the sensing layer is as follows:

[0038] Regarding specialized sensors, vibration and temperature sensors are fixedly installed on the motor end cover and winding positions of the elevator traction machine. The vibration sensor is a piezoelectric type with a measurement range of 0-200Hz, capable of real-time acquisition of vibration signals during traction machine operation. The temperature sensor is a PT100 platinum resistance sensor with a measurement accuracy of ±0.5℃, directly contacting the traction machine windings to obtain accurate temperature data. A current sensor, using a Hall effect current sensor with a measurement range of 0-5A, is connected in series at the output of the elevator door operator controller to monitor current changes during door operator motor operation. A displacement sensor, using a laser displacement sensor with a measurement accuracy of ±0.1mm, is installed next to the door guide rail; by real-time detection of the relative position between the door and the guide rail, the opening and closing speed is calculated. Pressure sensors, using strain gauge type pressure sensors with a measurement range of 0-10MPa, are installed at the contact position between the brake block and the guide rail of the elevator safety clamp and at the top of the piston rod of the buffer; these are used to monitor pressure changes of the safety devices during triggering or operation.

[0039] In addition to dedicated sensors, the perception layer is also equipped with a high-definition camera and a sound acquisition module. The camera is a 2-megapixel network camera, installed on the top of the car near the door, with the lens facing the door and the interior of the car. The frame rate is set to 25 frames per second to ensure clear capture of visual anomalies such as deformation and stuttering during the door opening and closing process. The sound acquisition module uses a high-fidelity microphone array, installed in the shaft near the traction machine, with a sampling rate set to 44.1kHz, to collect the mechanical sounds of the traction machine, door operator, and other devices during operation, filtering out environmental noise interference.

[0040] Implementation configuration of the transport layer:

[0041] The transport layer undertakes the core task of transmitting data from the perception layer to the platform layer. In this embodiment, a hybrid transmission mode of "edge computing module + fourth-generation mobile communication network / fifth-generation mobile communication network + wireless local area network" is adopted to ensure the real-time performance, stability and security of data transmission.

[0042] The edge computing module uses an embedded development board with an integrated ARM Cortex-A9 processor, providing strong data processing capabilities. This module connects to all sensors, cameras, and sound acquisition modules in the perception layer via RS485 or Ethernet interfaces to perform local preprocessing of the acquired raw data. The preprocessing includes data format conversion and redundant data filtering. Data format conversion unifies analog or digital signals from different sensors into JSON format for easier parsing by the platform layer. Redundant data filtering calculates the rate of change of data using the following formula; data with a rate of change below a threshold is considered redundant and removed:

[0043]

[0044] Where R represents the rate of change of data between two adjacent acquisition cycles, and D i D represents the data in the i-th acquisition period. i-1 This represents the data from the (i-1)th acquisition cycle. In this embodiment, the rate of change threshold is set to 0.5%. When R < 0.5%, it indicates that the data has no significant change and is considered redundant data. The purpose of this formula is to reduce data transmission volume. For example, when the traction machine is running stably, the temperature data changes little. If the temperature change rate is below the threshold for several consecutive cycles, only the data from the first cycle is transmitted, and subsequent redundant data is directly filtered out, achieving a data compression ratio of up to 10:1. To avoid missing critical data, when consecutive redundant data reaches 10 cycles, the system will force a data transmission. At this point, a redundant transmission judgment formula is introduced: [Filtering formula based on the number of consecutive redundant transmission cycles]. Where T represents the data transmission judgment result. The purpose of this formula is to ensure that data that has been running stably for a long time can also be monitored in real time while ensuring data compression effect, and to prevent monitoring blind spots caused by data transmission interruption.

[0045] After data preprocessing, the appropriate transmission method is selected based on the transmission scenario: In areas without Wi-Fi coverage within the elevator shaft, long-distance transmission is achieved via 4G or 5G mobile communication networks. Industrial-grade 4G / 5G mobile communication modules are selected, supporting full network compatibility, with data transmission rates exceeding 100Mbps and latency ≤500ms. Equipment within the elevator machine room is interconnected via Wi-Fi, using the IEEE 802.11n standard, achieving transmission rates up to 300Mbps, thus reducing the cost of long-distance transmission. To evaluate transmission quality, this embodiment introduces a transmission quality evaluation formula: Where Q is the transmission quality score (between 0 and 1), S is the transmission success rate (value is 0 or 1, 1 for successful transmission and 0 for failure), and D is the transmission delay (unit: ms). When Q < 0.5, the system automatically switches the transmission mode, for example, switching from a fourth-generation mobile communication network to a fifth-generation mobile communication network. This formula quantifies transmission quality, provides a basis for adaptive switching of transmission modes, and ensures the stability of data transmission. During transmission, the encryption unit uses an advanced encryption standard algorithm to encrypt the data, converting it into ciphertext before transmission; the breakpoint resumption unit records the data transmission progress, and when the network interruption is recovered, it resumes transmission from the breakpoint to avoid data loss.

[0046] Platform layer implementation configuration:

[0047] The platform layer is the core processing center of the system. In this embodiment, the platform layer is deployed on a cloud server cluster and includes a data storage module, an intelligent analysis module, and a business support module. Each module interacts with data through an internal bus.

[0048] The data storage module adopts a combined architecture of relational database and time-series database. The relational database used is MySQL, deployed on two servers in a master-slave architecture. The master server is responsible for data writing, and the slave server is responsible for data reading, ensuring data storage reliability. The MySQL database stores structured data such as basic elevator information (elevator model, installation time, user unit) and maintenance records (e.g., maintenance records, fault handling records). The time-series database used is InfluxDB, deployed on three servers to form a cluster, supporting concurrent data writing and querying from millions of devices. It stores real-time data collected by sensors, and its data storage format is sorted by timestamp for easy subsequent time-series data analysis. To optimize data storage performance, this embodiment calculates data storage priority using the following formula and allocates storage resources according to the priority:

[0049] P = α × W1 + β × W2

[0050] Where P represents data storage priority, α and β are weighting coefficients, and α + β = 1, in this embodiment α = 0.6, β = 0.4; W1 represents the device importance coefficient corresponding to the data (W1 = 1 for core safety devices such as traction machines and safety clamps, W1 = 0.8 for gantry crane systems, and W1 = 0.5 for other auxiliary devices); W2 represents the real-time coefficient of the data (W2 = 1 for real-time data, W2 = 0.3 for historical data). The purpose of this formula is to allocate more storage resources to the real-time data of core devices, ensuring the storage stability and access speed of critical data. Based on storage priority, this embodiment adopts the storage resource allocation formula: R = P × R total Where R is the storage resource allocated to a single type of data, R total This represents the total system storage resources. For example, the priority of real-time data from core devices is P = 0.6 × 1 + 0.4 × 1 = 1, and the allocated storage resources are 10% of the total storage resources. The priority of historical data from auxiliary devices is P = 0.6 × 0.5 + 0.4 × 0.3 = 0.42, and the allocated storage resources are 4.2% of the total storage resources. The purpose of this formula is to scientifically allocate storage resources according to the importance of the data, avoiding resource waste.

[0051] The intelligent analysis module is the core of intelligent fault diagnosis, employing a fusion model of a convolutional neural network (CNN) and a long short-term memory (LSTM) network. The CNN model processes image data captured by a camera, taking a 256×256 pixel image of the door as input. It extracts image features through convolutional layers, compresses feature dimensions through pooling layers, and outputs fault identification results through fully connected layers, recognizing visual faults such as door deformation and excessive door gaps. The LTM network analyzes time-series data such as vibration and temperature, taking 100 consecutive acquisition cycles as input. It memorizes long-term dependencies through gating units and outputs fault trend prediction results. To improve the model's diagnostic accuracy, this embodiment fuses the model outputs using the following formula:

[0052] F = γ × F CNN +(1-γ)×F LSTM

[0053] Where F represents the final fault diagnosis result, and γ is the fusion weight. In this embodiment, γ = 0.5, which can be adjusted according to the actual diagnosis effect; F CNN F represents the output of the convolutional neural network model (0 indicates no fault, 1 indicates visual fault). LSTM This represents the output of the Long Short-Term Memory (LSTM) network model (0 indicates no fault, 1 indicates a fault with temporal data anomalies). The purpose of this formula is to combine the diagnostic results of the two models, avoiding the limitations of a single model and improving the accuracy of fault identification. To further optimize the diagnostic results, this embodiment introduces a model confidence correction formula: F final = F × (0.8 + 0.2 × C), where F final The final corrected diagnostic result is represented by C, which is the model's diagnostic confidence level (between 0 and 1, output by the model based on its training accuracy). When the model's diagnostic confidence level for a certain fault is C = 0.9, the corrected result F... final =F×(0.8+0.2×0.9)=F×0.98, if F=1, the final diagnosis result is closer to 1, indicating that the fault diagnosis is more reliable; if C=0.5, the corrected result F final =F×0.9, indicating that the diagnostic results need to be manually reviewed. This formula improves the reliability of fault diagnosis results through confidence level correction.

[0054] The business support module provides basic support functions. The access control model for permission management is based on a role, which divides users into three roles: maintenance personnel, user unit administrators, and supervisors. Different roles are assigned different operation permissions. The log management records the operation behavior of all users and the system operation status, and the logs are retained for one year. The interface management provides a standardized application programming interface, which supports the connection with the maintenance management system of elevator maintenance units and the supervision platform of special equipment supervision departments to achieve data sharing.

[0055] Application layer implementation configuration:

[0056] At the application level, personalized functional services are provided to different user groups, supporting access from both web (based on B / S architecture) and mobile (based on APP / mini-program), with data synchronized in real time between the two.

[0057] In the maintenance personnel function module, the fault warning push unit pushes fault warning information via mobile application messages. The information includes elevator number, faulty component, fault type, and handling suggestions. The work order automatic generation unit automatically generates maintenance work orders based on the fault information. The work order number adopts the format of "elevator number-date-serial number" to ensure uniqueness. The maintenance progress tracking unit updates the work order progress in real time based on the work status uploaded by maintenance personnel on the mobile terminal (such as "order accepted", "processing", "completed").

[0058] Within the functional modules of the user unit, the operating status display unit visually displays elevator operating parameters, such as traction machine temperature and door operator current, in the form of a dashboard. Different colors are used to indicate parameter status (green indicates normal, yellow indicates warning, and red indicates fault). The fault statistics report unit calculates the monthly fault occurrence rate using the following formula and generates a statistical report:

[0059]

[0060] Among them, R fault N represents the monthly failure rate. fault N represents the number of times the elevator malfunctioned in that month. total This indicates the number of days the elevator operated in that month. The purpose of this formula is to quantify elevator malfunctions, providing data for users to assess elevator safety. To comprehensively evaluate the elevator's operational status, this embodiment also includes a supplementary elevator reliability formula: Where R rel For monthly operational reliability (between 0 and 1), m represents the number of failures in that month, and T represents the number of failures in that month. faulti T represents the downtime (in hours) for the i-th failure. total This represents the total elevator operating time for the month (in hours). For example, if an elevator operates for 720 hours in a month and experiences two malfunctions with downtimes of 2 hours and 1 hour respectively, then its operational reliability R is... rel =1-(2+1) / 720≈0.9958, indicating that the elevator has high operational reliability. This formula quantifies elevator reliability over time, compensating for the limitation of only counting the number of failures. The maintenance quality assessment unit introduces a maintenance quality scoring formula based on indicators such as maintenance work order response time, repair time, and user satisfaction. Where S qual For maintenance quality rating (between 0 and 1), Tresp T represents the average response time (in minutes, with a default threshold of 60 minutes). rep S represents the average repair time (in minutes, with a default threshold of 120 minutes). sat User satisfaction (between 0 and 1). The purpose of this formula is to quantify maintenance quality from multiple dimensions, providing a reference for users to select maintenance providers.

[0061] In the regulatory department's functional modules, the regional safety situation analysis unit generates a heat map by summarizing the fault data of all elevators in the region, which intuitively displays areas with high failure rates; the hidden danger investigation and supervision unit marks the faults and hidden dangers that have not been dealt with in a timely manner and sends supervision notices to the corresponding maintenance units and user units to ensure that the hidden dangers are rectified in a timely manner.

[0062] Specific implementation of the intelligent inspection and management method for elevator internal component faults:

[0063] Based on the above system, the intelligent inspection and management method for elevator internal component faults described in this embodiment includes a data acquisition stage, a data processing stage, a fault diagnosis stage, and an operation and maintenance management stage. Each stage is connected sequentially to form a closed-loop management process of "data acquisition-processing-diagnosis-operation and maintenance", ensuring that elevator faults are handled in a timely and effective manner.

[0064] Implementation of the data acquisition phase:

[0065] The core of the data acquisition phase is to comprehensively and accurately obtain the operating data and status information of the internal components of the elevator through the sensing layer devices. In this embodiment, data acquisition adopts a combination of timed acquisition and triggered acquisition: timed acquisition is carried out according to a preset acquisition cycle, and the acquisition cycle is different for different types of sensors. The acquisition cycle of vibration sensors and current sensors is 0.1 seconds, the acquisition cycle of temperature sensors and pressure sensors is 1 second, the acquisition cycle of the camera is 0.04 seconds (i.e., 25 frames / second), and the acquisition cycle of the sound acquisition module is 0.02 seconds; triggered acquisition automatically shortens the acquisition cycle to 1 / 10 of the original cycle when the data collected by the sensor exceeds a preset threshold, so as to achieve intensive acquisition of fault data.

[0066] During the data acquisition process, the sensing layer devices add a timestamp and device identifier to each acquired data point. The timestamp is accurate to the millisecond, and the device identifier is a unique number for the sensor, facilitating subsequent data traceability. For example, vibration data acquired by vibration sensor numbered "VIB-001" at 15:30:20.123 on October 20, 2025, has the identifier "VIB-001_20251020153020123".

[0067] Implementation of the data processing phase:

[0068] The data processing stage includes three parts: data transmission, data standardization, and hierarchical storage. The data transmission part is completed by the transmission layer, which transmits the valid data preprocessed by the perception layer to the platform layer through a hybrid transmission mode. During the transmission process, encryption units and breakpoint resume units ensure data security and integrity.

[0069] The data standardization process is completed by the data storage module at the platform layer. First, the data acquisition range, accuracy requirements, and transmission format for each type of sensor are defined. For example, the acquisition range of a vibration sensor is 0-200Hz, with an accuracy of ±0.1Hz, and the transmission format is "device identifier-timestamp-vibration frequency". Then, the data transmitted to the platform layer is format-validated. If the format is incorrect, an error message is returned, requiring the sensing layer to retransmit. Finally, the data is normalized using the following formula to eliminate the influence of data from different magnitudes:

[0070]

[0071] Among them, D norm D represents the normalized data, with values ​​ranging from 0 to 1. valid D represents valid data. min D represents the minimum value of this data type. max This represents the maximum value for this type of data. The purpose of this formula is to convert data from different types of sensors to the same order of magnitude, facilitating unified processing by the intelligent analysis module. For example, normalized traction machine temperature data (0-200℃) and gantry crane current data (0-5A) can be directly input into the model for analysis. To accommodate the parameter characteristics of different devices, this embodiment uses a standardized formula to process some nonlinear data: Where D std For standardized data, μ represents the historical average of this data type, and σ represents the historical standard deviation. For example, traction machine vibration data follows a normal distribution. After standardization, the data has a mean of 0 and a standard deviation of 1, making it more suitable for time series analysis using long short-term memory network models. This formula optimizes data processing for nonlinear data, improving the accuracy of model analysis.

[0072] The tiered storage mechanism differentiates storage based on data type and importance: real-time data (collected ≤ 24 hours ago) is stored in the memory area of ​​the InfluxDB time-series database to ensure fast access; short-term data (collected > 24 hours ago and ≤ 3 months ago) is stored in the InfluxDB disk area; historical data (collected > 3 months ago) is compressed and transferred to the archive database. The compression algorithm used is LZ4, with a compression ratio of up to 5:1. The archive database has a data storage period of 5 years, meeting the data retention requirements for special equipment supervision.

[0073] Implementation of the fault diagnosis phase:

[0074] The fault diagnosis phase includes two steps: fault classification diagnosis and differentiated early warning, which are completed collaboratively by the intelligent analysis module at the platform layer and the early warning unit at the application layer.

[0075] In the fault classification and diagnosis process, the intelligent analysis module analyzes the standardized data and classifies faults into three levels: Level 1 warning (minor fault) corresponds to parameters slightly deviating from the threshold, i.e., the normalized value of the data is between 0.8 and 0.9 (or 0.1 and 0.2), such as the gantry crane current being slightly higher than the normal range, but not affecting the operation of the gantry crane; Level 2 warning (serious fault) corresponds to parameters severely exceeding the standard or abnormal device operation, i.e., the normalized value of the data is >0.9 (or <0.1), such as the traction machine temperature rising significantly, exceeding the safe operating range; Level 3 warning (predictive fault) corresponds to potential fault trends discovered through time-series analysis, i.e., predicting that the data will exceed the threshold within the next 72 hours using a long short-term memory network model. In this embodiment, the warning level coefficient is calculated using the following formula to determine the fault warning level:

[0076] L=δ×D norm +∈×P predict

[0077] Where L represents the warning level coefficient, δ and ∈ are weighting coefficients, δ = 0.7, ∈ = 0.3; D norm For the normalized data, P predict The probability of fault prediction is between 0 and 1. When 0.8 ≤ L < 0.9, it is judged as a Level 1 warning; when L ≥ 0.9, it is judged as a Level 2 warning; when P predict ≥0.8 and D norm When the threshold is less than 0.8, it is classified as a Level 3 warning. This formula scientifically classifies the fault warning level by comprehensively considering the current data status and future predictions. To achieve dynamic adjustment of the warning threshold, this embodiment introduces an adaptive threshold formula: Among them Th new For the new warning threshold, Th old The original warning threshold, N fault N represents the number of similar malfunctions of this elevator model in the past three months. total This represents the total number of elevators of this model in operation over the past three months. For example, if the original normalized warning threshold for traction machine temperature of a certain elevator model was 0.8, and 8 traction machine temperature failures occurred out of 50 elevators of this model in the past three months, then the new threshold Th... new =0.8×(1+0.1×8 / 50)=0.8128, which achieves stricter early warning by increasing the threshold, reducing the risk of failure. The function of this formula is to dynamically optimize the early warning threshold based on the actual operation of the elevator, thereby improving the pertinence of the early warning.

[0078] Differentiated early warning procedures employ different push methods for different levels of early warning: Level 1 early warnings send text messages to maintenance personnel via mobile application; Level 2 early warnings, while simultaneously sending messages via mobile application, notify maintenance personnel and user unit administrators via voice call; Level 3 early warnings are linked to the maintenance unit's dispatch system, automatically sending preventative maintenance notices and synchronizing early warning information to the regulatory department's platform.

[0079] Implementation of the operation and maintenance management phase:

[0080] The operation and maintenance management phase includes two steps: closed-loop fault handling and full-process operation and maintenance recording, which realizes full-process management of faults from discovery to resolution and the traceability of operation and maintenance data.

[0081] The closed-loop fault handling process includes five stages: early warning triggering, work order dispatch, fault handling, repair verification, and post-mortem optimization. After an early warning is triggered, the system automatically matches the elevator's maintenance responsibility unit with the corresponding maintenance personnel, completing work order dispatch within 10 minutes. Maintenance personnel receive the work order via mobile device, confirm it, and proceed to the site for handling. During the handling process, they can upload site photos and repair parts information via their mobile device. After the fault is repaired, the maintenance personnel upload the repair certificate, and the system automatically collects the operating data of the elevator components, verifying whether the fault has been resolved using the following formula:

[0082]

[0083] Where C represents the normalized average of the data collected over n collection cycles after repair, n = 100, D i,norm This represents the normalized value of the data in the i-th acquisition cycle. When 0.3 ≤ C ≤ 0.7, the fault is considered resolved, and the work order is marked as "completed"; otherwise, the repair is considered unsatisfactory, and maintenance personnel are notified to handle it again. To further verify the repair effect, this embodiment supplements the data stability verification formula: Where S represents the variance of the repaired data; the smaller the variance, the more stable the data. When C is within the normal range and S < 0.01, the repair quality is considered excellent; when C is within the normal range but 0.01 ≤ S ≤ 0.05, the repair is considered acceptable but requires enhanced monitoring; when S > 0.05, even if C is normal, the repair is considered to have potential problems and needs to be re-examined. This formula assesses the repair quality from the perspective of data stability, avoiding situations where the repair appears acceptable but has potential problems. A weekly review of the week's fault handling is conducted, introducing a review optimization coefficient formula: Where O is the post-mortem optimization coefficient (between 0 and 1), N success N represents the number of faults successfully handled this week. total T represents the total number of faults this week. avgThis represents the average fault response time for the week (in minutes). When O < 0.8, the system automatically generates an optimization report, prompting adjustments to fault diagnosis parameters or maintenance dispatch rules. This formula quantifies the effectiveness of fault handling, providing data support for process optimization.

[0084] The entire operation and maintenance process is recorded by the platform-level business support module, documenting key milestones such as work order dispatch time, maintenance personnel response time, fault handling start time, fault repair time, repair content, and parts used, forming a full lifecycle operation and maintenance file for the elevator. This file is linked to the elevator's serial number and can be searched using keywords such as elevator number and time, providing a basis for subsequent elevator maintenance, fault tracing, and quality assessment.

[0085] Implementation of system security management:

[0086] The management method in this embodiment also includes system security management steps, ensuring stable system operation from three dimensions: physical security, network security, and data security. Physical security protection employs anti-tampering and anti-damage designs for the perception layer devices. Anti-tamper switches are installed on the casings of sensors and edge computing modules. When a device is abnormally disassembled, the anti-tamper switch is triggered, immediately sending an alarm message to the platform layer. Network security protection involves deploying firewalls and intrusion detection systems to restrict access from unauthorized IP addresses, regularly (weekly) updating security policies, and introducing a network security risk assessment formula: Risk = 0.5 × A + 0.3 × B + 0.2 × C, where Risk is the network security risk value (between 0 and 1), A is the frequency coefficient of unauthorized access attempts (between 0 and 1, the higher the number of attempts), B is the coefficient for security policy update lag days (between 0 and 1, the higher the lag days), and C is the coefficient for the number of network vulnerabilities (between 0 and 1, the higher the number of vulnerabilities). When Risk > 0.7, the system automatically initiates emergency protection measures, such as disconnecting unnecessary network connections and notifying the administrator. This formula quantifies cybersecurity risks, enabling early warning and mitigation. Data security protection employs data encryption, access control, and operation log auditing measures, introducing the data security compliance formula: Compliance = 0.4 × E + 0.3 × P + 0.3 × L, where Compliance is the data security compliance score (between 0 and 1), E is the data encryption coverage (between 0 and 1), P is the access control accuracy (between 0 and 1), and L is the log audit integrity (between 0 and 1). This formula assesses data security status from multiple dimensions, ensuring compliance with special equipment data security management requirements. Data storage and transmission are encrypted; users with different roles can only access data within their authorized scope; all data operations are logged to ensure data security and operational traceability.

[0087] Implementation Results Description:

[0088] The intelligent fault detection and management system and method for elevator internal components described in this embodiment were tested for six months on 100 elevators in a residential community. To quantify the test results, in addition to the original indicators, a comprehensive system performance formula was introduced: Where Eff is the overall system performance score (between 0 and 1), and R... acc T represents the fault identification accuracy (between 0 and 1). resp R is the mean fault response time (in minutes). fault R represents the monthly failure rate. rel To ensure the reliability of elevator operation. Trial operation results show that R... acc =0.96, T resp =30 minutes, R fault =0.05, R rel =0.995, then the overall system efficiency Eff = 0.3 × 0.96 + 0.25 × (1 - 30 / 120) + 0.25 × (1 - 0.05) + 0.2 × 0.995 ≈ 0.92, indicating excellent overall system efficiency. Specific data shows that the system can collect real-time and accurate operating data of elevator internal components, achieving a fault identification accuracy rate of over 96%, a 40% improvement over traditional manual inspection methods; fault response time has been reduced from the traditional 2 hours to within 30 minutes, improving fault repair efficiency by 60%; through a three-level early warning mechanism and preventative maintenance, the incidence of sudden elevator failures has decreased by 45%, significantly improving the safety and reliability of elevator operation. Simultaneously, the system generates full-process operation and maintenance records and statistical reports, providing strong data support for maintenance units to optimize services, user units to strengthen management, and regulatory departments to conduct precise supervision, demonstrating high practical application value.

[0089] The intelligent inspection and management system and method for elevator internal component faults described in this invention constructs a closed-loop management system of "data acquisition-processing-diagnosis-maintenance" through the collaborative design of the perception layer, transmission layer, platform layer and application layer. Compared with the traditional elevator inspection and management model, it has the following multi-dimensional beneficial effects, and all technical terms correspond completely to the descriptions in the aforementioned system and method, without the use of abbreviations.

[0090] Improve the accuracy and timeliness of fault identification to enhance elevator operational safety.

[0091] Significantly Improved Fault Identification Accuracy: The perception layer of this invention deploys specialized sensors such as vibration sensors, temperature sensors, current sensors, and pressure sensors on core components like elevator traction machines, door operating systems, safety clamps, and buffers. Combined with high-definition cameras and sound acquisition modules, this enables comprehensive collection of operating parameters and status information of internal elevator components. The intelligent analysis module at the platform layer employs a fusion model of convolutional neural network (CNN) and long short-term memory (LSTM) networks. The CNN model accurately processes image data to identify visual faults such as door deformation, while the LTM model deeply analyzes temporal data such as vibration and temperature to capture potential fault trends. The diagnostic results are then optimized through model result fusion formulas and confidence level correction formulas, achieving a fault identification accuracy rate of over 96%. This reduces the false positive rate by more than 40% compared to traditional manual judgment based on experience, effectively avoiding missed or incorrect detections.

[0092] The fault response time is significantly shortened: The system collects data in real time through the perception layer, and the transmission layer adopts a hybrid transmission mode of "edge computing module + fourth-generation mobile communication network / fifth-generation mobile communication network + wireless LAN" to achieve a data transmission latency of ≤500ms, ensuring that operational anomalies are detected immediately. The platform layer scientifically classifies warnings into Level 1, Level 2, and Level 3 warnings using a warning level coefficient formula. The application layer adopts differentiated push methods for different warning levels. Level 1 warnings are pushed via mobile application messages, Level 2 warnings are superimposed with voice telephone notifications, and Level 3 warnings are linked to the maintenance unit's dispatch system. This reduces the fault response time from the traditional 2 hours of manual inspection to within 30 minutes, gaining critical time for rapid fault handling and significantly reducing the risks of elevator entrapment and the spread of component damage.

[0093] Optimize operation and maintenance management processes to reduce elevator operation and maintenance costs:

[0094] Achieving preventative maintenance and reducing sudden malfunctions: The system's intelligent analysis module uses a long short-term memory network model to predict trends in time-series data and dynamically adjusts warning thresholds using an adaptive threshold formula. This enables it to issue a three-level warning 72 hours before a significant malfunction occurs in internal elevator components, driving a shift in maintenance from "reactive repair" to "preventative maintenance." Trial operation results show that preventative maintenance can reduce sudden elevator malfunctions by 45%, avoiding high repair costs caused by escalating component failures and minimizing inconvenience caused by elevator downtime.

[0095] Streamlining operation and maintenance management and improving work efficiency: The application-layer maintenance personnel module can automatically generate maintenance work orders containing elevator number, faulty components, fault type, and handling suggestions. Through mobile devices, the entire process of work order dispatch, progress tracking, and voucher uploading is digitally managed, eliminating the need for manual paperwork and reducing paperwork time by 70%. The platform-layer's full-process operation and maintenance record function automatically stores information on work order dispatch, response, and repair, forming a full lifecycle operation and maintenance archive for the elevator. This provides a clear basis for maintenance quality traceability and evaluation, avoiding issues of shirking responsibility due to unclear accountability.

[0096] Reduce labor costs: Traditional elevator inspection requires maintenance personnel to conduct regular on-site checks of each elevator, which is time-consuming, labor-intensive, and inefficient. This invention achieves centralized remote monitoring of multiple elevators through automatic data collection at the sensing layer and intelligent analysis and diagnosis at the platform layer. One maintenance personnel can manage more than 50 elevators simultaneously through the system, increasing the number of elevators managed per person by four times compared to the traditional model, and significantly reducing the cost of manual inspection.

[0097] To achieve data-driven, refined management and support multi-stakeholder decision-making needs:

[0098] Providing comprehensive elevator management data for users: The user-level functional modules at the application layer generate multi-dimensional statistical reports using quantitative indicators such as monthly failure rate formulas, elevator operational reliability formulas, and maintenance quality scoring formulas. These reports visually display elevator operating status, failure distribution, and maintenance service quality. Users can use this data to accurately grasp the safety and reliability of each elevator, scientifically formulate elevator upgrade and renovation plans, and select high-quality maintenance companies through maintenance quality scoring, thereby enhancing proactive elevator management.

[0099] Providing precise regulatory support to regulatory authorities: The application-layer regulatory module generates a safety situation heatmap by aggregating elevator malfunction data within a region. Combined with assessment indicators such as cybersecurity risk assessment formulas and data security compliance formulas, it enables precise location and supervision of elevator safety hazards. Regulatory authorities can grasp the overall elevator safety status of a region without conducting on-site inspections, shifting regulatory work from a "broad-based" to a "precision-targeted" approach, improving regulatory efficiency and targeting, and contributing to the improvement of the special equipment safety regulatory system.

[0100] The system offers optimized services to maintenance companies by periodically analyzing fault handling effectiveness using a review and optimization coefficient formula. It generates optimization reports for issues such as long response times and poor repair quality, allowing maintenance companies to adjust dispatch rules, enhance staff training, and improve service levels. Simultaneously, the system's vast accumulation of fault data provides support for maintenance companies to summarize fault patterns across different elevator brands and models, enabling them to optimize maintenance plans.

[0101] To ensure the stability and security of system operation and improve the reliability of technology applications:

[0102] Data transmission and storage are secure and reliable: The transmission layer employs advanced encryption standards to encrypt data, and a breakpoint resume unit prevents data loss due to network interruptions. A transmission quality evaluation formula ensures adaptive switching of data transmission methods, guaranteeing the security and stability of data transmission. The platform layer adopts a combined architecture of relational database and time-series database, scientifically allocating storage resources using data storage priority and resource allocation formulas. This enables stable storage and rapid retrieval of millions of pieces of equipment data, meeting the requirement for 5-year data retention for special equipment supervision.

[0103] The system boasts high adaptability: the sensors in the perception layer support wide voltage input, adapting to the complex power supply environment of elevators; the hybrid transmission mode in the transmission layer can automatically switch transmission methods according to different scenarios such as elevator shafts and machine rooms, without being limited by regional network conditions; the cloud server cluster deployment in the platform layer ensures that the system can still operate stably under high concurrency access, adapting to the elevator management needs of different sized communities, office buildings and other places, and has a wide range of application scenarios.

[0104] In summary, the intelligent inspection and management system and method for elevator internal component faults of the present invention realizes intelligent elevator fault inspection, refined operation and maintenance management, and data-driven safety supervision through technological innovation. It not only significantly improves the safety and reliability of elevator operation, but also effectively reduces operation and maintenance costs. It provides comprehensive technical support for elevator users, maintenance units, and regulatory departments, and has extremely high practical value and promotion significance.

[0105] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. An intelligent inspection and management system for faults in elevator internal components, characterized by: It includes a sensing layer, a transmission layer, a platform layer, and an application layer. The sensing layer, the transmission layer, the platform layer, and the application layer are connected in sequence and communicate with each other. Each layer works together to collect, transmit, analyze, process, and display the operating data of the elevator's internal components. The sensing layer is deployed inside the elevator car and shaft to collect the operating parameters and status information of the elevator's internal components. The transmission layer is used to transmit the data collected by the perception layer to the platform layer; the platform layer is used to store, analyze and provide business support for the received data; the application layer is used to provide dedicated functional services for different user groups.

2. The intelligent inspection and management system for elevator internal component faults according to claim 1, characterized in that: The sensing layer includes multiple dedicated sensors, cameras, and a sound acquisition module. The dedicated sensors include vibration sensors, temperature sensors, displacement sensors, current sensors, and pressure sensors. The vibration and temperature sensors are installed in the elevator traction machine to monitor the operating vibration frequency and winding temperature of the traction machine. The displacement and current sensors are installed in the elevator door operator system to capture the door operator's opening and closing speed and motor current changes. The pressure sensor is installed in the elevator safety brake and buffer to monitor the trigger status of the safety devices. The camera is used to collect image information of the inside of the car and the door. The sound acquisition module is used to collect mechanical sound information during elevator operation.

3. The intelligent inspection and management system for elevator internal component faults according to claim 1, characterized in that: The transmission layer adopts a hybrid transmission mode, which includes an edge computing module, a fourth-generation mobile communication network, a fifth-generation mobile communication network, and a wireless local area network. The edge computing module is used to perform local preprocessing and redundant data filtering on the data collected by the sensing layer. The fourth-generation mobile communication network and the fifth-generation mobile communication network are used to achieve long-distance data transmission. The wireless local area network is used to achieve short-distance interconnection of data between devices in the elevator machine room. The transmission layer also includes an encryption unit and a breakpoint resumption unit. The encryption unit is used to encrypt the transmitted data, and the breakpoint resumption unit is used to prevent data loss due to network interruption.

4. The intelligent inspection and management system for elevator internal component faults according to claim 1, characterized in that: The platform layer includes a data storage module, an intelligent analysis module, and a business support module. The data storage module adopts a combined architecture of relational database and time-series database. The relational database is used to store structured data such as elevator basic information and maintenance records, while the time-series database is used to store real-time data collected by sensors. The intelligent analysis module includes a convolutional neural network model and a long short-term memory network model. The convolutional neural network model is used to process image data to identify visual faults, while the long short-term memory network model is used to analyze time-series data to achieve fault trend prediction. The business support module provides access control, log management, and interface management functions.

5. The intelligent inspection and management system for elevator internal component faults according to claim 1, characterized in that: The application layer includes a maintenance personnel function module, a user unit function module, and a regulatory department function module. The maintenance personnel function module includes a fault early warning push unit, a work order automatic generation unit, and a maintenance progress tracking unit. The user unit function module includes an operation status display unit, a fault statistical report unit, and a maintenance quality assessment unit. The regulatory department function module includes a regional safety situation analysis unit and a hidden danger investigation and supervision unit. The application layer supports access from both web and mobile devices.

6. A method for intelligent inspection and management of elevator internal component faults based on the system described in any one of claims 1 to 5, characterized in that: The system includes a data acquisition phase, a data processing phase, a fault diagnosis phase, and an operation and maintenance management phase, with each phase sequentially linked to form a closed-loop management process. The data acquisition phase obtains the operating data and status information of the elevator's internal components through the sensing layer equipment. The data processing phase transmits the acquired data to the platform layer through the transmission layer and completes storage and preprocessing. The fault diagnosis phase uses an intelligent analysis module to analyze data to identify faults and predict risks; the operation and maintenance management phase carries out fault handling and full-process operation and maintenance work based on the diagnosis results.

7. The intelligent inspection and management method for elevator internal component faults according to claim 6, characterized in that: The data processing stage includes a data standardization step and a hierarchical storage step. The data standardization step is used to clarify the data acquisition range, accuracy requirements and transmission format of each type of sensor to ensure data consistency. The hierarchical storage step is used to adopt differentiated storage strategies for real-time data and historical data. Real-time data is stored in a time-series database, and historical data is compressed and transferred to an archive database. The data storage period of the archive database meets the regulatory requirements for special equipment.

8. The intelligent inspection and management method for elevator internal component faults according to claim 6, characterized in that: The fault diagnosis phase includes a fault classification diagnosis step and a differentiated early warning step. The fault classification diagnosis step divides faults into Level 1, Level 2, and Level 3 early warnings. Level 1 early warnings correspond to minor faults where parameters deviate slightly from the threshold; Level 2 early warnings correspond to serious faults where parameters exceed limits or devices malfunction; and Level 3 early warnings correspond to potential fault trends discovered through time-series analysis. The differentiated early warning step uses different push methods for different levels of early warnings. Level 1 early warnings are pushed via mobile application messages; Level 2 early warnings are pushed via a combination of mobile application messages and telephone notifications; and Level 3 early warnings trigger preventative maintenance by the maintenance unit.

9. The intelligent inspection and management method for elevator internal component faults according to claim 6, characterized in that: The operation and maintenance management phase includes closed-loop fault handling steps and full-process operation and maintenance recording steps. The closed-loop fault handling steps include early warning triggering, work order dispatching, fault handling, repair verification, and review and optimization. After the early warning is triggered, the system automatically matches the maintenance responsibility unit and personnel and completes the work order dispatching. After the maintenance personnel complete the repair, they upload the voucher. The system verifies the fault resolution by collecting device operation data and regularly reviews the fault handling situation. The full-process operation and maintenance recording steps are used to record the information of the entire process nodes such as work order dispatching, response, and completion, forming an elevator full life cycle operation and maintenance file.

10. The intelligent inspection and management method for faults in elevator internal components according to claim 6, characterized in that: It also includes system security management steps, which include physical security protection, network security protection, and data security protection. The physical security protection adopts anti-tampering and anti-damage design for the perception layer devices, and triggers an alarm when the devices are abnormally disassembled. The network security protection restricts unauthorized access by deploying firewalls and intrusion detection systems and regularly updates security policies. The data security protection adopts data encryption, access control, and operation log auditing measures to ensure that data is only accessed by authorized personnel and that operations are traceable.

Images