A communication system and data transmission method based on elevator Internet of Things

Abstract

The present invention relates to the field of network communication technology, and in particular to a communication system and data transmission method based on the elevator Internet of Things. The method comprises the following steps: using an elevator controller to collect sensor response data streams and generate real-time monitoring response data; performing offset characteristic analysis on the real-time monitoring response data to obtain abnormal event data and normal event data respectively; performing urgency correction processing on the abnormal event data to obtain corrected urgency score data; performing network status monitoring according to the elevator Internet of Things gateway and performing adaptive channel allocation to generate optimal communication routing path data; performing dynamic data transmission processing on the corrected urgency score data and normal event data according to the optimal communication routing path data to obtain a dynamic data transmission strategy. The present invention effectively prevents network congestion and data loss by optimizing the data transmission path, and significantly improves data transmission efficiency.

Description

Technical Field

[0001] The present invention relates to the field of network communication technology, and in particular to a communication system and data sending method based on elevator Internet of Things. Background Art

[0002] Elevators are now an indispensable means of transportation in high-rise buildings, widely used in residential, office, and shopping malls. With the development of IoT technology, elevator IoT systems are becoming a crucial tool for improving elevator efficiency and safety. Using sensors, controllers, and communication modules, these systems monitor elevator operating status, fault information, and passenger status in real time. These systems transmit this data to the cloud for analysis and processing, enabling remote monitoring, preventative maintenance, and emergency rescue. In recent years, with the rapid development of urban construction and the increasing number of high-rise buildings, elevators, as an essential means of transportation for daily commutes, have drawn increasing attention to their safety and efficiency. However, traditional data transmission methods, due to the long distances over which elevator systems are distributed across different floors and areas, result in unstable data transmission and high packet loss rates. Furthermore, the large volume of data generated by elevator systems makes traditional data transmission methods difficult to meet the high reliability, low latency, and high bandwidth requirements. Summary of the Invention

[0003] Based on this, the present invention provides a communication system and data sending method based on elevator Internet of Things to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a data transmission method based on elevator Internet of Things includes the following steps:

[0005] Step S1: Using the elevator controller to perform wireless network access processing, and collect periodic operation status data of the elevator to generate periodic operation status data; performing operation mode matching based on the periodic operation status data to generate elevator operation mode data; collecting sensor response data stream based on the elevator operation mode data to generate real-time monitoring response data;

[0006] Step S2: Perform sensor source response integration processing based on the real-time monitoring response data to generate comprehensive monitoring response data; construct a multi-source monitoring fusion matrix based on the comprehensive monitoring response data; perform elevator state coding and identification processing on the multi-source monitoring fusion matrix to generate real-time elevator state snapshot data;

[0007] Step S3: constructing a time series operation trend prediction model based on the elevator operation mode data; performing expected feedforward calculation on the real-time elevator state snapshot data using the real-time elevator state snapshot data to obtain real-time monitoring feature offset data; performing offset characteristic analysis based on the real-time monitoring feature offset data to obtain abnormal event data and normal event data respectively;

[0008] Step S4: performing abnormal pattern matching on the abnormal event data to obtain abnormal pattern indicator data; performing urgency scoring processing based on the abnormal pattern indicator data to generate urgency score data; performing urgency correction processing on the urgency score data to obtain corrected urgency score data;

[0009] Step S5: Monitor the network status according to the elevator IoT gateway and generate network status monitoring data; perform dynamic bandwidth allocation processing according to the network status monitoring data and generate a dynamic adjustment bandwidth allocation strategy; perform adaptive channel allocation based on the dynamic adjustment bandwidth allocation strategy, and perform intelligent obstacle avoidance optimization processing to generate optimal communication routing path data; perform dynamic data transmission processing on the corrected urgency score data and normal event data according to the optimal communication routing path data to obtain a dynamic data transmission strategy.

[0010] The present invention utilizes an elevator controller for wireless network access processing, enabling seamless connection between the elevator system and the Internet of Things. It collects periodic elevator operating status data, enabling comprehensive real-time information on the elevator's operating status. Operating mode matching based on periodic operating status data accurately identifies the elevator's operating mode, such as normal operation, peak operation, and maintenance operation. Sensor response data streams are collected based on elevator operating mode data, enabling the collection of corresponding sensor data based on different operating modes, improving the targetedness and efficiency of data collection. Sensor source response integration processing based on real-time monitoring response data fuses data from different sensors to comprehensively reflect the elevator's operating status. A multi-source monitoring fusion matrix is constructed based on the integrated monitoring response data, enabling matrix processing of multiple sensor data. Elevator status coding and identification processing is performed on the multi-source monitoring fusion matrix, enabling an encoded representation of the elevator's real-time operating status. Using the multi-source monitoring fusion matrix, the elevator's operating status can be more accurately identified, improving diagnostic accuracy. A time-series operating trend prediction model is constructed using elevator operating mode data, enabling prediction of future operating trends based on historical operating mode data. Utilizing real-time elevator status snapshot data for expected feedforward calculation, the system compares the real-time status with the expected status and promptly identifies abnormal deviations. Analysis of deviation characteristics based on real-time monitoring feature deviation data allows for the separation of abnormal events from normal events. This system enables prediction and early warning of elevator operating status, providing a crucial guarantee for improving elevator operational efficiency and safety. Abnormal pattern matching on abnormal event data allows for specific abnormal patterns, such as failures, jams, and overloads, to be identified based on their characteristics. Urgency scoring based on abnormal pattern indicator data allows for quantitative urgency scores for different types of abnormal situations. Urgency correction processing on urgency score data allows for dynamic adjustment of initial scores based on actual conditions, improving the accuracy of urgency assessments. This system enables intelligent identification and assessment of elevator faults, providing an effective means for timely response and ensuring passenger safety. By monitoring the network status of the elevator IoT gateway, such as connection status, bandwidth availability, and signal strength, network issues, such as disconnection, insufficient bandwidth, and signal interference, can be promptly identified. This system enables real-time monitoring of the elevator IoT system's network status, enabling the timely detection of network problems. Based on network status monitoring data, the bandwidth allocation strategy is dynamically adjusted. For example, when network conditions are good, a larger bandwidth can be allocated, while when conditions are poor, a smaller bandwidth can be allocated to ensure stable and reliable data transmission. Based on the dynamically adjusted bandwidth allocation strategy, the appropriate communication channel is selected, and an intelligent obstacle avoidance algorithm is used to optimize the data transmission path, avoiding network congestion and data loss.By using intelligent obstacle avoidance algorithms, data transmission paths can be prevented from passing through network congestion areas or signal interference areas. For example, paths with higher signal strength can be selected, or network congested nodes can be bypassed to improve data transmission efficiency. Intelligent management of data transmission can adjust data transmission strategies according to the urgency of the event and network conditions, improve data transmission efficiency, ensure the stability and reliability of data transmission, and ensure priority transmission of important data. Therefore, a data transmission method based on the elevator Internet of Things of the present invention comprehensively grasps the real-time status of the network through dynamic network monitoring and hierarchical analysis. Bandwidth resources are flexibly allocated according to the network status to optimize transmission efficiency. Adaptive channel allocation and intelligent obstacle avoidance optimization improve the flexibility and reliability of communication routes. Determine the optimal communication path, avoid potential interference, and solve transmission instability and packet loss problems. Urgent data is given priority in real-time transmission to meet timeliness and real-time requirements. Ordinary transmission data is reasonably cached and transmitted regularly to achieve large-scale data transmission.

[0011] Preferably, step S1 includes the following steps:

[0012] Step S11: Using the elevator controller to perform wireless network access processing on the elevator IoT gateway and generate a wireless network access credential;

[0013] Step S12: performing communication request security authentication processing on the cloud server according to the wireless network access credential to generate communication security authentication token data;

[0014] Step S13: using the elevator controller to collect periodic operation status data of the elevator based on the communication security authentication token data to generate periodic operation status data;

[0015] Step S14: performing time series analysis on the periodic operation status data to generate periodic operation time series data, wherein the periodic operation time series data includes elevator operation status data, elevator load data, elevator operation time data, and elevator floor scheduling data;

[0016] Step S15: performing elevator load analysis based on the periodic operation status data, and performing operation mode matching to generate elevator operation mode data;

[0017] Step S16: extracting operation mode monitoring indicators based on the elevator operation mode data to generate mode monitoring indicator data;

[0018] Step S17: Based on the pattern monitoring index data, the elevator controller controls the sensor to collect the sensor response data stream to generate real-time monitoring response data.

[0019] This invention utilizes the elevator controller to perform wireless network access processing on the elevator IoT gateway, ensuring a secure and reliable wireless connection between the elevator control system and the IoT gateway. Wireless network access credentials, used as the basis for identity authentication, prevent unauthorized devices from accessing the network, improving system security. Secure authentication of the cloud server using the wireless network access credentials ensures secure communication between the elevator system and the cloud server. Communication security authentication token data is used to ensure the security and reliability of the data collection process. Elevator operating status data is organized and analyzed in chronological order to comprehensively reflect the elevator's operating conditions at different points in time. Elevator load analysis is performed based on periodic operating status data, analyzing the elevator's load at different times. This analysis is combined with other operating parameters to accurately identify the elevator's operating mode. By extracting mode monitoring indicator data, sensor response data can be more accurately collected, improving data collection efficiency and accuracy. Based on the extracted mode monitoring indicator data, various sensors (such as temperature sensors, vibration sensors, and load sensors) are specifically controlled to collect data, obtaining real-time monitoring data for the elevator under specific operating modes.

[0020] Preferably, step S2 includes the following steps:

[0021] Step S21: The elevator controller transmits the real-time monitoring response data to the elevator IoT gateway for sensor source preprocessing to generate sensor source preprocessing data;

[0022] Step S22: performing sensor source response integration processing based on the sensor source preprocessing data to generate comprehensive monitoring response data;

[0023] Step S23: constructing a sensor source key feature space for the comprehensive monitoring response data to generate sensor source feature vector data;

[0024] Step S24: constructing a multi-source monitoring fusion matrix based on the comprehensive monitoring response data and the sensor source feature vector data;

[0025] Step S25: performing principal component analysis based on the multi-source monitoring fusion matrix to obtain multi-source key response feature data;

[0026] Step S26: performing elevator status coding and identification processing on the multi-source key response feature data to generate real-time elevator status snapshot data.

[0027] The present invention preprocesses the raw monitoring data from each sensor at the gateway, such as through de-noising, filtering, and standardization. This preliminary processing and normalization improves data quality and consistency. Preprocessed data from different sensors is integrated, fusing monitoring information from different dimensions to achieve integrated processing of real-time data from different sensors. A sensor source key feature space is constructed for the comprehensive monitoring response data, key feature dimensions are extracted from the comprehensive monitoring response data, a sensor source feature space is constructed, and the feature data is vectorized, effectively representing the key features of the elevator's operating status. The comprehensive monitoring response data and sensor source feature vector data are matrix-fused to construct a multi-source monitoring fusion matrix. The multi-source monitoring fusion matrix can provide a unified matrix representation of monitoring data from different dimensions. Principal component analysis is performed on the multi-source monitoring fusion matrix to extract principal component features representative of the elevator's operating status, enabling a high-level summary of the elevator's operating characteristics and reducing data redundancy. The elevator's real-time operating status is identified as a specific state code, and real-time elevator status snapshot data can efficiently and accurately reflect the elevator's current operating status.

[0028] Preferably, step S3 includes the following steps:

[0029] Step S31: extracting state benchmark indicators based on elevator operation mode data to generate state benchmark indicator data;

[0030] Step S32: performing transfer learning processing on the state benchmark indicator data through a preset long short-term memory network model, thereby constructing a time series operation trend prediction model;

[0031] Step S33: Acquire historical elevator monitoring response data; perform expected feedforward calculation on the historical elevator monitoring response data using a time series operation trend prediction model to generate expected elevator operation characteristic data;

[0032] Step S34: using the real-time elevator status snapshot data to compare the elevator expected operation feature data with the feature values of each dimension, and performing distance metric calculation to obtain real-time monitoring feature offset data;

[0033] Step S35: performing an offset characteristic analysis based on the real-time monitoring characteristic offset data to generate elevator state offset characteristic data, wherein the elevator state offset characteristic data includes normal operation characteristic data and abnormal offset characteristic data;

[0034] Step S36: performing anomaly detection processing on the multi-source monitoring fusion matrix using the abnormal offset feature data to obtain abnormal event data and preliminary normal event data;

[0035] Step S37: performing monitoring event data fusion on the normal operation feature data and the preliminary normal event data to generate normal event data.

[0036] Based on the identified elevator operating mode, the present invention extracts key status indicators for the corresponding mode, such as temperature and vibration in normal mode, and load and door opening time in peak mode, reflecting the expected elevator state under different operating modes. Using a deep learning long-short-term memory network, transfer learning is performed on state benchmark indicator data to construct a time series model capable of predicting future elevator operating trends. Real-time elevator state snapshot data is compared with expected operating characteristic data, and the difference in feature values for each dimension is compared one by one to quantify the degree of deviation between the actual elevator operating state and the expected state. Real-time monitoring feature offset data is analyzed, and data with offsets within the normal range is identified as normal operating feature data, while data outside the normal range is identified as abnormal offset feature data. The identified abnormal offset feature data is used to fusion the multi-source monitoring matrix. Furthermore, data not detected as abnormal can be preliminarily identified as normal event data. Abnormal event data reflects faults, anomalies, and other conditions in elevator operation. The normal operation feature data previously obtained from the offset characteristic analysis is fused with the normal event data preliminarily determined from the anomaly detection to obtain the final normal event data.

[0037] Preferably, step S4 includes the following steps:

[0038] Step S41: Identify the abnormal event type of the abnormal event data and generate abnormal type identification data;

[0039] Step S42: performing abnormal pattern matching on the abnormal type identification data using a preset elevator fault model library, and extracting abnormal pattern indicators to generate abnormal pattern indicator data;

[0040] Step S43: performing severity assessment based on the abnormal pattern indicator data to obtain an abnormal type severity index;

[0041] Step S44: performing an urgency assessment on the abnormality type severity index and the abnormality pattern indicator data based on a fuzzy logic inference algorithm to generate an urgency factor;

[0042] Step S45: performing urgency scoring processing on the urgency factor to generate urgency scoring data;

[0043] Step S46: performing event context analysis based on the abnormal event data to generate event context analysis data; performing fault impact analysis on the event context analysis data using the abnormal type severity index to generate an abnormal event impact index;

[0044] Step S47: Using the abnormal event impact index, the urgency score data is subjected to urgency correction processing to obtain corrected urgency score data.

[0045] The present invention analyzes detected abnormal event data to identify the specific type of abnormal event, clarifying its nature. The identified abnormal type is matched with a pre-established elevator fault model library to identify the corresponding abnormal pattern. Key abnormal pattern indicators are extracted from this data, reflecting the specific characteristics of the abnormal event. The severity of the abnormal event is assessed to quantitatively reflect the degree of harm posed by that type of abnormal event. The abnormal type severity index and abnormal pattern indicator data are used as input to generate an urgency factor that reflects the urgency of the abnormal event. The urgency factor comprehensively considers multiple factors influencing the abnormal event. This quantitative assessment of urgency provides a more intuitive reflection of the event's urgency. By analyzing the contextual information surrounding the abnormal event and combining it with the abnormal type severity index, a more comprehensive assessment of the event's impact can be achieved. The initial urgency score is revised; for example, if the event's impact is high, even a low initial score will be revised to a higher score, providing a more accurate assessment of the event's urgency based on its impact.

[0046] Preferably, step S5 includes the following steps:

[0047] Step S51: sorting the abnormal event data by data transmission priority by modifying the urgency score data to obtain priority transmission list data;

[0048] Step S52: performing network status monitoring according to the elevator IoT gateway to generate network status monitoring data, wherein the network status monitoring data includes network topology data, network traffic data, and abnormal network status identification data;

[0049] Step S53: Perform dynamic bandwidth allocation processing based on network status monitoring data to generate a dynamic bandwidth allocation strategy;

[0050] Step S54: performing intelligent communication route planning based on the dynamic bandwidth allocation strategy to obtain intelligent communication route planning data;

[0051] Step S55: Adaptively allocate channels to the priority transmission list data and normal event data using the intelligent communication routing planning data to obtain target communication routing path data;

[0052] Step S56: Using the preset elevator shaft interference source to perform intelligent obstacle avoidance optimization processing on the target communication routing path data to generate optimal communication routing path data;

[0053] Step S57: perform communication capability evaluation on the priority sending list data and normal event data of the optimal communication routing path data to generate communication capability evaluation data; when the communication capability evaluation data satisfies the priority sending list data and the normal event data, it is transmitted to the cloud server in real time through the secure transmission protocol based on the optimal communication routing path data; when the communication capability evaluation data does not satisfy the priority sending list data and the normal event data, it is transmitted to the cloud server in real time through the secure transmission protocol, and the normal event data is temporarily stored in the elevator Internet of Things gateway for scheduled transmission, thereby obtaining a dynamic data sending strategy.

[0054] The present invention prioritizes all abnormal events based on the corrected urgency score data, ensuring that emergency event information is transmitted to the cloud server first. The elevator IoT gateway monitors network conditions in real time, and the network status monitoring data comprehensively reflects the current network operating status. Based on the acquired network status monitoring data, available network bandwidth resources are dynamically allocated, enabling rational allocation of bandwidth resources based on the real-time network status and optimizing network transmission efficiency. Using dynamically adjusted bandwidth allocation strategies as constraints, intelligent algorithms are used to plan communication routes, determining the optimal communication path under current network conditions and improving data transmission reliability and efficiency. Based on the intelligent communication route planning data, combined with the priority transmission list data and normal event data, the appropriate communication channel is selected, enabling intelligent selection of data transmission paths. The optimal transmission path can be selected based on data type and network conditions. Based on preset information about elevator shaft interference sources, such as metal components in the elevator shaft and wireless signal interference, intelligent optimization of data transmission paths is achieved. The optimal data transmission strategy is selected based on network conditions and event urgency, ensuring the priority transmission of important data, improving data transmission efficiency, and ensuring data transmission stability and reliability.

[0055] The present invention also provides a communication system based on the elevator Internet of Things, which executes the data transmission method based on the elevator Internet of Things as described above. The communication system based on the elevator Internet of Things includes:

[0056] The elevator monitoring response module is used to use the elevator controller to perform wireless network access processing, collect periodic operation status data of the elevator, and generate periodic operation status data; match the operation mode based on the periodic operation status data to generate elevator operation mode data; collect sensor response data stream based on the elevator operation mode data to generate real-time monitoring response data;

[0057] The monitoring data fusion module is used to integrate the sensor source responses based on the real-time monitoring response data to generate comprehensive monitoring response data; construct a multi-source monitoring fusion matrix based on the comprehensive monitoring response data; and perform elevator status coding and identification processing on the multi-source monitoring fusion matrix to generate real-time elevator status snapshot data;

[0058] The sending data classification module is used to build a time series operation trend prediction model based on the elevator operation mode data; perform expected feedforward calculation on the real-time elevator status snapshot data using the real-time elevator status snapshot data to obtain real-time monitoring feature offset data; perform offset characteristic analysis based on the real-time monitoring feature offset data to obtain abnormal event data and normal event data respectively;

[0059] The emergency communication assessment module performs abnormal pattern matching on the abnormal event data to obtain abnormal pattern indicator data; performs urgency scoring processing based on the abnormal pattern indicator data to generate urgency score data; performs urgency correction processing on the urgency score data to obtain corrected urgency score data;

[0060] The adaptive communication optimization module is used to monitor the network status of the elevator Internet of Things gateway and generate network status monitoring data; perform dynamic bandwidth allocation processing based on the network status monitoring data to generate a dynamic adjustment bandwidth allocation strategy; perform adaptive channel allocation based on the dynamic adjustment bandwidth allocation strategy, and perform intelligent obstacle avoidance optimization processing to generate optimal communication routing path data; and perform dynamic data transmission processing on the corrected urgency score data and normal event data based on the optimal communication routing path data to obtain a dynamic data transmission strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a schematic flow chart of the steps of a communication system and data transmission method based on elevator Internet of Things of the present invention;

[0062] Figure 2 for Figure 1 Detailed implementation steps of step S2 in FIG.

[0063] Figure 3 for Figure 1 Detailed implementation steps of step S5 are shown in the flowchart.

[0064] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0065] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are 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 those skilled in the art without making any creative work are within the scope of protection of the present invention.

[0066] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0067] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0068] To achieve this, please refer to Figures 1 to 3 The present invention provides a data transmission method based on elevator Internet of Things, comprising the following steps:

[0069] Step S1: Using the elevator controller to perform wireless network access processing, and collect periodic operation status data of the elevator to generate periodic operation status data; performing operation mode matching based on the periodic operation status data to generate elevator operation mode data; collecting sensor response data stream based on the elevator operation mode data to generate real-time monitoring response data;

[0070] Step S2: Perform sensor source response integration processing based on the real-time monitoring response data to generate comprehensive monitoring response data; construct a multi-source monitoring fusion matrix based on the comprehensive monitoring response data; perform elevator state coding and identification processing on the multi-source monitoring fusion matrix to generate real-time elevator state snapshot data;

[0071] Step S3: constructing a time series operation trend prediction model based on the elevator operation mode data; performing expected feedforward calculation on the real-time elevator state snapshot data using the real-time elevator state snapshot data to obtain real-time monitoring feature offset data; performing offset characteristic analysis based on the real-time monitoring feature offset data to obtain abnormal event data and normal event data respectively;

[0072] Step S4: performing abnormal pattern matching on the abnormal event data to obtain abnormal pattern indicator data; performing urgency scoring processing based on the abnormal pattern indicator data to generate urgency score data; performing urgency correction processing on the urgency score data to obtain corrected urgency score data;

[0073] Step S5: Monitor the network status according to the elevator IoT gateway and generate network status monitoring data; perform dynamic bandwidth allocation processing according to the network status monitoring data and generate a dynamic adjustment bandwidth allocation strategy; perform adaptive channel allocation based on the dynamic adjustment bandwidth allocation strategy, and perform intelligent obstacle avoidance optimization processing to generate optimal communication routing path data; perform dynamic data transmission processing on the corrected urgency score data and normal event data according to the optimal communication routing path data to obtain a dynamic data transmission strategy.

[0074] In the embodiment of the present invention, reference Figure 1 The above is a schematic flow chart of the steps of a data transmission method based on the elevator Internet of Things of the present invention. In this embodiment, the data transmission method based on the elevator Internet of Things includes the following steps:

[0075] Step S1: Using the elevator controller to perform wireless network access processing, and collect periodic operation status data of the elevator to generate periodic operation status data; performing operation mode matching based on the periodic operation status data to generate elevator operation mode data; collecting sensor response data stream based on the elevator operation mode data to generate real-time monitoring response data;

[0076] In an embodiment of the present invention, the elevator controller establishes a wireless network connection with the elevator IoT gateway via a device such as a wireless network card or a Bluetooth module. The gateway sends an access request, and the controller responds and generates a wireless network access credential (such as a WiFi key) to complete access authentication. The elevator controller establishes a wireless network connection with the elevator IoT gateway via a device such as a wireless network card or a Bluetooth module. The gateway sends an access request, and the controller responds and generates a wireless network access credential (such as a WiFi key) to complete access authentication. The controller analyzes the load data based on periodic operation sequence data using a load analysis algorithm (such as cluster analysis) to determine the real-time load level of the elevator, and matches the current load level with various operating modes (such as peak mode and night mode) to generate elevator operating mode data. Based on the extracted mode monitoring indicators, the controller controls various sensors (energy consumption sensors, noise sensors, etc.) to perform targeted data collection and obtain real-time monitoring response data.

[0077] Step S2: Perform sensor source response integration processing based on the real-time monitoring response data to generate comprehensive monitoring response data; construct a multi-source monitoring fusion matrix based on the comprehensive monitoring response data; perform elevator state coding and identification processing on the multi-source monitoring fusion matrix to generate real-time elevator state snapshot data;

[0078] In this embodiment of the present invention, raw data is preprocessed based on sensor source information (sensor type, measurement object, etc.), including methods such as noise removal, normalization, and interpolation, to generate standardized sensor source preprocessed data. The gateway utilizes sensor source response integration algorithms (such as data fusion and wavelet transform) to fuse preprocessed data from multiple heterogeneous sensor sources, eliminating redundancy and noise and generating comprehensive monitoring response data that comprehensively reflects the current operating status of the elevator. Feature engineering techniques are used to construct a key feature space model for the sensor sources, extract representative feature subsets from the raw data, and map them into feature vectors to generate sensor source feature vector data. The comprehensive monitoring response data and its feature vectors are organized into a high-dimensional matrix using a structure (such as parallel or cascaded). State encoding algorithms (such as decision trees and support vector machines) are used to classify and encode these features, mapping the current operating status of the elevator into an identification code. This code is then bound to the response features to generate real-time elevator status snapshot data.

[0079] Step S3: constructing a time series operation trend prediction model based on the elevator operation mode data; performing expected feedforward calculation on the real-time elevator state snapshot data using the real-time elevator state snapshot data to obtain real-time monitoring feature offset data; performing offset characteristic analysis based on the real-time monitoring feature offset data to obtain abnormal event data and normal event data respectively;

[0080] In an embodiment of the present invention, a state coding algorithm (such as a decision tree or support vector machine) is used to classify and encode these features, mapping the current operating state of the elevator to an identification code and binding it to the response feature to generate real-time elevator status snapshot data. With the help of a long short-term memory neural network (LSTM) model, transfer learning is performed on these time series indicator data to construct a time series operation trend prediction model for the current elevator. The model is used to perform expected feedforward calculations on these snapshot data to generate expected elevator operation feature data that can predict the normal operation trend of the elevator. The feature values of the real-time snapshot data and the expected operation feature data are compared dimension by dimension, and the difference between the two in each dimension is calculated as the real-time monitoring feature offset data. The offset within the normal fluctuation range is classified as normal operating feature data, and the abnormal offset is classified as abnormal offset feature data.

[0081] Step S4: performing abnormal pattern matching on the abnormal event data to obtain abnormal pattern indicator data; performing urgency scoring processing based on the abnormal pattern indicator data to generate urgency score data; performing urgency correction processing on the urgency score data to obtain corrected urgency score data;

[0082] In this embodiment of the present invention, an abnormality type recognition model (such as a support vector machine) is used to classify this data, mapping abnormal events to predefined abnormality type identifiers, such as "overload" and "stuck," to generate abnormality type identifier data. Abnormality pattern matching is performed using a preset elevator fault model library, outputting the best-matching abnormality pattern and extracting relevant abnormality pattern indicator data, such as current anomaly and temperature anomaly. The abnormality type severity index and abnormality pattern indicator data are used as input to comprehensively assess the urgency of the abnormal event and generate an urgency factor output. The context of the event (time, location, load status, etc.) is analyzed to generate event context analysis data. This analysis data is combined with the abnormality type severity index to assess the scope and extent of the abnormal event's impact, generating an abnormal event impact index. The abnormal event impact index is used to correct the urgency score data. Abnormal events with a wide impact and high risk are scored higher, while those with less significant impact are scored lower, resulting in corrected urgency score data.

[0083] Step S5: Monitor the network status according to the elevator IoT gateway and generate network status monitoring data; perform dynamic bandwidth allocation processing according to the network status monitoring data and generate a dynamic adjustment bandwidth allocation strategy; perform adaptive channel allocation based on the dynamic adjustment bandwidth allocation strategy, and perform intelligent obstacle avoidance optimization processing to generate optimal communication routing path data; perform dynamic data transmission processing on the corrected urgency score data and normal event data according to the optimal communication routing path data to obtain a dynamic data transmission strategy.

[0084] In an embodiment of the present invention, the gateway acquires network topology data, network traffic data, and abnormal network status identification data through active probing (Ping, Traceroute) and passive capture, generating comprehensive network status monitoring data. A bandwidth allocation algorithm (such as the Max-Min algorithm) is used to dynamically schedule available bandwidth and generate a dynamically adjusted bandwidth allocation strategy. The gateway integrates the network status monitoring data, constructs a flow network model, and classifies network traffic flows. High-priority channels are assigned to the revised urgency score data, while normal event data is assigned to standard channels, generating target communication routing path data. Intelligent optimization algorithms (such as genetic algorithms) are used to plan an optimal routing solution that avoids interference sources, generating optimal communication routing path data. The gateway evaluates the communication capabilities of the optimal route and determines whether it meets the requirements for real-time transmission of all data. If not, only the revised score data is transmitted, while normal data is temporarily stored and transmitted in batches with delays, generating a dynamic data transmission strategy.

[0085] Preferably, step S1 includes the following steps:

[0086] Step S11: Using the elevator controller to perform wireless network access processing on the elevator IoT gateway and generate a wireless network access credential;

[0087] Step S12: performing communication request security authentication processing on the cloud server according to the wireless network access credential to generate communication security authentication token data;

[0088] Step S13: using the elevator controller to collect periodic operation status data of the elevator based on the communication security authentication token data to generate periodic operation status data;

[0089] Step S14: performing time series analysis on the periodic operation status data to generate periodic operation time series data, wherein the periodic operation time series data includes elevator operation status data, elevator load data, elevator operation time data, and elevator floor scheduling data;

[0090] Step S15: performing elevator load analysis based on the periodic operation status data, and performing operation mode matching to generate elevator operation mode data;

[0091] Step S16: extracting operation mode monitoring indicators based on the elevator operation mode data to generate mode monitoring indicator data;

[0092] Step S17: Based on the pattern monitoring index data, the elevator controller controls the sensor to collect the sensor response data stream to generate real-time monitoring response data.

[0093] In an embodiment of the present invention, the elevator controller establishes a wireless network connection with the elevator IoT gateway via a wireless communication device such as a wireless network card or Bluetooth module. The gateway sends a wireless network access request. The controller responds by generating wireless network access credentials, such as a WiFi key or Bluetooth pairing code, and transmits them to the gateway for access authentication. Based on the pre-configured cloud server address, port, and other information, the controller initiates a communication connection request to the cloud server. Upon receiving the request, the cloud server performs a security check based on the mutually agreed upon security authentication protocol (e.g., digital certificates, message authentication codes, etc.). Upon successful verification, the cloud server generates and returns a communication security authentication token. Data on the elevator's operating status is collected at predetermined intervals (e.g., hourly or daily). For example, sensors collect real-time data such as the elevator's speed, load, and current, while counters record periodic data such as the number of starts and stops and operating time. This data is analyzed and processed using time series analysis algorithms (e.g., sliding averages and exponential smoothing) to generate periodic operating time series data. This data includes elevator operating status data (such as operating mode), load data (such as number of passengers), operating time data (such as operating duration), and floor scheduling data (such as the order of floors). Statistical indicators such as the mean, variance, and peak value of the load data are calculated to identify overall load trends. The load data is classified into different categories to identify different load patterns, such as peak load and off-peak load. The current load level is matched with multiple preset operating modes (such as peak mode and night mode) to determine the optimal elevator operating mode and generate corresponding elevator operating mode data. Monitoring indicator data related to the current mode is extracted from the pre-set multiple operating mode monitoring indicators. For example, for night mode, indicators such as energy consumption and noise can be extracted; for peak mode, indicators such as waiting time and passenger load factor can be extracted. For example, in night mode, the controller will activate energy consumption sensors and noise sensors to periodically collect energy consumption and noise data during elevator operation as real-time monitoring response data. In peak mode, the controller will activate infrared sensors to monitor passenger flow in the elevator car in real time, obtaining passenger entry and exit data to assess the current passenger load factor and waiting time.

[0094] Preferably, step S15 includes the following steps:

[0095] Step S151: performing elevator motion trajectory analysis based on elevator operation status data to generate elevator motion trajectory data;

[0096] Step S152: performing elevator passenger load flow analysis on the elevator load data using the elevator motion trajectory data to generate passenger load flow distribution data;

[0097] Step S153: performing cluster analysis on the passenger load flow distribution data to generate elevator load cluster data;

[0098] Step S154: performing load-time correlation processing on the elevator running time data using the elevator load clustering data to generate load-time correlation data;

[0099] Step S155: performing elevator load calculation on the elevator load clustering data using the load time correlation data to obtain elevator average load data and elevator peak load data; performing load pattern recognition based on the elevator average load data and elevator peak load data to generate elevator load pattern data;

[0100] Step S156: extracting passenger behavior patterns from the elevator floor dispatch data using the elevator load pattern data to generate passenger behavior pattern data;

[0101] Step S157: performing operation mode matching on a preset elevator operation mode library through the passenger behavior pattern data and the elevator load pattern data to generate elevator operation mode data.

[0102] In embodiments of the present invention, trajectory analysis algorithms (such as Kalman filtering and particle filtering) are used to analyze and reconstruct the elevator's motion trajectory, generating elevator motion trajectory data describing the elevator's motion path in three-dimensional space. For example, the elevator's kinematic data, such as displacement, velocity, and acceleration, are analyzed to construct a two-dimensional coordinate system. The floor and travel direction information corresponding to each time point is mapped into the coordinate system, forming a point. These points are then connected in chronological order to form the elevator's motion trajectory, thereby reconstructing the elevator's round-trip trajectory between different floors. Passenger uplink and downlink traffic is analyzed. For example, pattern recognition algorithms are used to analyze the patterns of passenger boarding and alighting on different floors and during different time periods, generating passenger load flow distribution data describing the distribution of passenger flow between floors. Cluster analysis algorithms (such as K-Means and DBSCAN) are used to perform unsupervised clustering on this data, dividing data points with similar load flow characteristics into multiple clusters to generate elevator load cluster data. This data reflects typical uplink and downlink passenger load patterns that exist at different time periods and on different floors. For example, a peak load pattern involves concentrated and high passenger flow during rush hour. Off-peak load patterns: Passenger flow is dispersed and relatively low. Special load patterns: Passenger flow is concentrated during specific time periods, such as holidays and major events. For example, for daytime commuting peak load clusters, the corresponding time periods are 7:00-9:00 and 17:00-19:00; for nighttime off-peak load clusters, the corresponding time period is 22:00 to 6:00 the following day. The analysis results generate load-time correlation data. The average number of passengers (average load data) and the maximum number of passengers (peak load data) in each load cluster are calculated. Load pattern recognition algorithms (such as pattern matching) are then used to match the average and peak load data with various preset load patterns (such as peak, normal, and off-peak) to determine the current elevator load pattern and generate elevator load pattern data. Passenger behavior pattern extraction analyzes passenger boarding and alighting behavior based on elevator floor scheduling data under different load patterns and extracts corresponding behavior patterns. For example, in a peak load pattern, passenger boarding and alighting behavior is concentrated during commuting peak hours, and passengers primarily board and alight between specific floors. In low-peak load mode: Passenger boarding and alighting behaviors are dispersed, and the frequency of passengers boarding and alighting between different floors is low. The preset elevator operation mode library is matched for operation mode. For example, when passenger behavior pattern data shows that passenger boarding and alighting behaviors are concentrated during rush hour, and passengers mainly board and alight between certain specific floors, the elevator load pattern data will show a peak load mode. When passenger behavior pattern data shows that passenger boarding and alighting behaviors are dispersed, and the frequency of passengers boarding and alighting between different floors is low, the elevator load pattern data will show a low-peak load mode.

[0103] Preferably, step S2 includes the following steps:

[0104] Step S21: The elevator controller transmits the real-time monitoring response data to the elevator IoT gateway for sensor source preprocessing to generate sensor source preprocessing data;

[0105] Step S22: performing sensor source response integration processing based on the sensor source preprocessing data to generate comprehensive monitoring response data;

[0106] Step S23: constructing a sensor source key feature space for the comprehensive monitoring response data to generate sensor source feature vector data;

[0107] Step S24: constructing a multi-source monitoring fusion matrix based on the comprehensive monitoring response data and the sensor source feature vector data;

[0108] Step S25: performing principal component analysis based on the multi-source monitoring fusion matrix to obtain multi-source key response feature data;

[0109] Step S26: performing elevator status coding and identification processing on the multi-source key response feature data to generate real-time elevator status snapshot data.

[0110] As an example of the present invention, refer to Figure 2 As shown, Figure 1 Detailed implementation steps of step S2 are shown in the flowchart. In this example, step S2 includes:

[0111] Step S21: The elevator controller transmits the real-time monitoring response data to the elevator IoT gateway for sensor source preprocessing to generate sensor source preprocessing data;

[0112] In an embodiment of the present invention, the elevator controller transmits the collected real-time monitoring response data to the elevator IoT gateway via a wireless network. After receiving the data, the gateway performs preprocessing on the raw data, such as de-noising, normalization, and interpolation, based on preconfigured sensor source information (such as sensor type and measurement object), to generate standardized sensor source preprocessed data. For example, the analog current value collected by the current sensor is converted into a digital value in standard units. For example, the elevator controller transmits weight sensor data to the gateway, which cleans the data, deletes abnormal data caused by sensor failure, and converts it into a unified JSON format.

[0113] Step S22: performing sensor source response integration processing based on the sensor source preprocessing data to generate comprehensive monitoring response data;

[0114] In this embodiment of the present invention, a sensor source response integration algorithm (such as data fusion and wavelet transform) is used to fuse data from multiple different sensor sources to generate comprehensive monitoring response data that comprehensively reflects the current operating status of the elevator. For example, energy consumption sensor data is integrated with operation sequence data to obtain energy consumption response data of the elevator under different loads. For example, the gateway integrates weight sensor data and speed sensor data to generate comprehensive monitoring response data that includes information such as the elevator's current load, speed, and operating direction.

[0115] Step S23: constructing a sensor source key feature space for the comprehensive monitoring response data to generate sensor source feature vector data;

[0116] In this embodiment of the present invention, feature engineering techniques are used to construct a key feature space model for sensor sources. This model uses methods such as feature selection and extraction to extract the most representative feature subsets describing the elevator's operating status from the raw data. These subsets are then mapped into feature vectors to generate sensor source feature vector data. For example, key features such as average power and maximum power are extracted from energy consumption response data to construct a feature vector.

[0117] Step S24: constructing a multi-source monitoring fusion matrix based on the comprehensive monitoring response data and the sensor source feature vector data;

[0118] In this embodiment of the present invention, the integrated monitoring response data and feature vector data from different sensors are organized, for example, rows representing different sensors and columns representing different monitoring indicators or features. The response data and feature vectors from multiple sensor sources are organized into a high-dimensional matrix using a certain structure (e.g., parallel or cascaded). For example, data from multiple sources, such as current, energy consumption, and noise, can be spliced into a single large matrix.

[0119] Step S25: performing principal component analysis based on the multi-source monitoring fusion matrix to obtain multi-source key response feature data;

[0120] In this embodiment of the present invention, a principal component analysis (PCA) algorithm is used to reduce data dimensionality. A few principal components that significantly influence elevator state changes are extracted from the original high-dimensional matrix and output as multi-source key response feature data. For example, PCA extracts 10 principal component features from a matrix containing over 100 features. The gateway then performs principal component analysis on the fused matrix to obtain two principal components: the first principal component represents changes in elevator load, and the second principal component represents changes in elevator speed.

[0121] Step S26: performing elevator status coding and identification processing on the multi-source key response feature data to generate real-time elevator status snapshot data.

[0122] In this embodiment of the present invention, a state encoding algorithm (such as a decision tree or support vector machine) is used to classify and encode these feature data, mapping the elevator's current operating state to an identification code. This code is then bound to the response feature data to generate real-time elevator status snapshot data that comprehensively describes the elevator's instantaneous operating state. For example, the gateway encodes and identifies the results of principal component analysis, converting the elevator load changes represented by the first principal component into "normal operation" or "overload alarm," and converting the elevator speed changes represented by the second principal component into "normal operation" or "fault shutdown," and generates a real-time elevator status snapshot containing this information.

[0123] Preferably, step S3 includes the following steps:

[0124] Step S31: extracting state benchmark indicators based on elevator operation mode data to generate state benchmark indicator data;

[0125] Step S32: performing transfer learning processing on the state benchmark indicator data through a preset long short-term memory network model, thereby constructing a time series operation trend prediction model;

[0126] Step S33: Acquire historical elevator monitoring response data; perform expected feedforward calculation on the historical elevator monitoring response data using a time series operation trend prediction model to generate expected elevator operation characteristic data;

[0127] Step S34: using the real-time elevator status snapshot data to compare the elevator expected operation feature data with the feature values of each dimension, and performing distance metric calculation to obtain real-time monitoring feature offset data;

[0128] Step S35: performing an offset characteristic analysis based on the real-time monitoring characteristic offset data to generate elevator state offset characteristic data, wherein the elevator state offset characteristic data includes normal operation characteristic data and abnormal offset characteristic data;

[0129] Step S36: performing anomaly detection processing on the multi-source monitoring fusion matrix using the abnormal offset feature data to obtain abnormal event data and preliminary normal event data;

[0130] Step S37: performing monitoring event data fusion on the normal operation feature data and the preliminary normal event data to generate normal event data.

[0131] In this embodiment of the present invention, a series of key indicators related to the healthy operating status of the elevator are extracted from the operating mode data based on pre-defined status benchmark indicator rules to generate status benchmark indicator data. For example, for "peak commuting mode," indicators such as normal passenger load factor, average waiting time, and energy consumption can be extracted as benchmarks; for "nighttime mode," indicators such as normal noise level and idle power can be extracted. Transfer learning is performed on these time-series indicator data using a pre-trained long short-term memory (LSTM) neural network model to construct a time-series operating trend prediction model for the current elevator. Transfer learning can transfer feature knowledge learned by the LSTM from other domains, accelerating model convergence in the elevator context and improving prediction accuracy. This model is used to perform expected feedforward calculations on these historical response data to generate expected elevator operating characteristic data that can predict the normal operating trend of the elevator. This data describes the expected values of each dimension of characteristic indicators under normal conditions. For example, elevator monitoring response data under peak mode for the past week is obtained and then, using a pre-trained LSTM model, performed expected feedforward calculations on this data to predict the elevator operating trend under peak mode for the next week. The expected elevator operating characteristic data containing the predicted results is then generated. The eigenvalues of the expected elevator operating characteristic data are compared dimension by dimension, and the difference between the two in each dimension is calculated as the real-time monitoring characteristic offset data. For example, the real-time power value is compared with the expected power value to obtain the power offset. This eigenvalue comparison can identify the difference between the actual operating status and the expected operating status. A distance metric is used to calculate the distance between the actual and expected operating statuses, obtaining the real-time monitoring characteristic offset data. These offsets are analyzed using offset characteristic analysis algorithms (such as cluster analysis and anomaly detection). Offsets within the normal fluctuation range are classified as normal operating characteristic data, and abnormal offsets outside the normal range are classified as abnormal offset characteristic data. Anomaly detection processing is performed on the constructed multi-source monitoring fusion matrix, such as reconstructed principal component analysis and out-of-core point detection. Abnormal event data is separated from the original fusion matrix, and event data preliminarily determined to be normal is also obtained. For example, if the average load offset of an event is within the threshold range, but the operating speed offset exceeds the threshold, the event is marked as an abnormal event. If both the average load offset and the operating speed offset of an event are within the threshold range, the event is marked as normal.

[0132] Preferably, step S36 includes the following steps:

[0133] Step S361: performing real-time data stream matching on the multi-source monitoring fusion matrix using abnormal offset feature data to generate a real-time offset data stream;

[0134] Step S362: Divide the real-time offset data stream into time series windows according to a preset time window to generate real-time offset data segments;

[0135] Step S363: performing offset point identification on the abnormal offset feature data based on the real-time offset data segment to generate target offset point index data;

[0136] Step S364: performing feature importance evaluation on the real-time offset data segments and calculating the optimal segmentation value to obtain segmentation decision data;

[0137] Step S365: Using the target offset point index data as a tree child node, an isolation tree is constructed on the real-time offset data segment using an isolation tree algorithm based on the segmentation decision data to generate a feature offset isolation tree;

[0138] Step S366: Calculate the path length based on the feature offset isolation tree to generate offset point path length data; perform time sensitivity adjustment on the offset point path length data to generate time offset weight data;

[0139] Step S367: Using the time offset weight data, perform weighted anomaly scoring on the offset point path length data to generate offset point anomaly score data;

[0140] Step S368: abnormal event judgment is performed on the offset point abnormality score data using a preset abnormality judgment threshold. When the offset point abnormality score data is lower than the offset point abnormality score data, the relevant offset point and the real-time offset data segment are marked as preliminary normal event data; when the offset point abnormality score data is higher than or equal to the offset point abnormality score data, the relevant offset point and the real-time offset data segment are marked as preliminary abnormal event data.

[0141] Step S369: Match the preliminary abnormal event data with the elevator operation log by timestamp, and fuse the event data to obtain abnormal event data.

[0142] In an embodiment of the present invention, abnormal offset feature data is associated with real-time data in a fusion matrix. For example, timestamp matching is used to associate the abnormal offset feature data with fusion matrix data with the same timestamp. Eigenvalue matching can also be used to associate the abnormal offset feature data with fusion matrix data with the same eigenvalue. For example, a constraint-based nearest neighbor search algorithm is used to quickly find data points in the fusion matrix that are most similar to the abnormal offset feature. These points are marked as real-time offset data stream output. This data stream reflects the current abnormal operating state of the elevator. This data stream is segmented according to preset time windows (e.g., 10 seconds, 30 seconds, etc.) to generate multiple real-time offset data segments. These data segments represent snapshots of the abnormal operating state of the elevator at different time periods. An outlier detection algorithm (e.g., density-based local anomaly detection (LODB)) is used to identify offset points in the abnormal offset feature data. Data points in the data segments that are highly correlated with the abnormal offset feature are marked as target offset points, and corresponding target offset point index data is generated. An algorithm (e.g., random forest, SHAP, etc.) is used to automatically calculate the importance of each feature for anomaly detection and select the optimal feature subset based on this information. Based on this feature subset, a decision tree algorithm is then used to calculate the optimal segmentation value for each feature, generating segmentation decision data. The target offset point index data is used as the leaf node label of the anomaly detection tree, and the segmentation decision data is used as the segmentation criteria for internal nodes. The isolation tree algorithm is used to construct the real-time offset data segments, generating a feature offset isolation tree model. For example, using average load offset as the segmentation feature, data points with an average load offset greater than 30 kg are divided into one sub-segment, and data points less than or equal to 30 kg are divided into another sub-segment. The cloud server then uses running speed offset as the segmentation feature to further divide each sub-segment, ultimately constructing an isolation tree. For example, using a linear weighting method, the path length of data points within the last minute is multiplied by 1, the path length of data points older than 1 minute is multiplied by 0.9, the path length of data points older than 2 minutes is multiplied by 0.8, and so on. Using the time offset weight data, the offset point path length data is weighted and summed to obtain the anomaly score for each target offset point. This offset point anomaly score is then compared with the threshold. When the anomaly score data for a particular offset point falls below a threshold, that offset point and the corresponding real-time offset data segment are marked as preliminary normal event data. When the anomaly score data exceeds or equals the threshold, it is marked as preliminary abnormal event data. This anomaly judgment threshold can be statistically learned based on historical data. Multiple thresholds can be set for different anomaly types to achieve hierarchical anomaly detection. Furthermore, other dimensions of contextual information, such as time and location, can be incorporated to dynamically adjust the thresholds and improve the adaptability of anomaly detection. For example, the timestamp of an abnormal event can be compared with the time of the abnormal alarm recorded in the elevator operation log to identify event pairs with highly consistent timestamps.Using multi-source information fusion technology, such as DS evidence theory fusion and wavelet-wavelet fusion, the preliminary abnormal event data is fused with the operation log data to produce abnormal event data output with higher confidence.

[0143] Preferably, step S4 includes the following steps:

[0144] Step S41: Identify the abnormal event type of the abnormal event data and generate abnormal type identification data;

[0145] Step S42: performing abnormal pattern matching on the abnormal type identification data using a preset elevator fault model library, and extracting abnormal pattern indicators to generate abnormal pattern indicator data;

[0146] Step S43: performing severity assessment based on the abnormal pattern indicator data to obtain an abnormal type severity index;

[0147] Step S44: performing an urgency assessment on the abnormality type severity index and the abnormality pattern indicator data based on a fuzzy logic inference algorithm to generate an urgency factor;

[0148] Step S45: performing urgency scoring processing on the urgency factor to generate urgency scoring data;

[0149] Step S46: performing event context analysis based on the abnormal event data to generate event context analysis data; performing fault impact analysis on the event context analysis data using the abnormal type severity index to generate an abnormal event impact index;

[0150] Step S47: Using the abnormal event impact index, the urgency score data is subjected to urgency correction processing to obtain corrected urgency score data.

[0151] In an embodiment of the present invention, an abnormality type recognition model (such as a support vector machine or neural network) is used to classify this data, mapping abnormal events to predefined abnormality type identifiers, such as "overload" and "stuck," to generate corresponding abnormality type identifier data. This model can be trained under the supervision of historical abnormality data to improve recognition accuracy. Abnormality pattern matching is performed using a preset elevator fault model library. This model library contains pattern feature descriptions of various abnormality types, such as fault causes, symptoms, and harmful impacts. The controller matches the current abnormality with the patterns in the model library, outputs the best-matching abnormality pattern, and extracts relevant abnormality pattern indicator data, such as current anomalies and temperature anomalies. The severity of the current abnormality is assessed using expert knowledge or supervised learning methods to generate a quantitative abnormality type severity index. For example, this is calculated by matching the abnormality pattern indicator with a preset scoring rule, or by inputting the indicator into a trained regression model to output a severity score on a scale of 0-10. A fuzzy logic inference algorithm is used to assess the urgency of the abnormality type severity index abnormality pattern indicator data. The resulting urgency factor is stored to generate the urgency factor. For example, the severity index of the abnormality type, the average load offset, and the operating speed offset. The urgency factor is input into a preset scoring model (such as a decision tree, BP neural network, etc.), and the urgency is quantitatively scored to obtain urgency score data of 0-100 points. The context of the event is analyzed to generate event context analysis data. These contexts include information such as the time and place of the event, the current load status of the elevator, and the passenger situation. The urgency score data is corrected. For example, for abnormal events with a wide impact range and a high degree of harm, the urgency score is appropriately increased; for events with a smaller impact, the score is reduced. The score results are weighted and corrected according to different intervals of the impact index. Nonlinear correction functions, such as Gaussian correction, can also be used to reflect the nonlinear effect of the impact index on the score.

[0152] Preferably, step S5 includes the following steps:

[0153] Step S51: sorting the abnormal event data by data transmission priority by modifying the urgency score data to obtain priority transmission list data;

[0154] Step S52: performing network status monitoring according to the elevator IoT gateway to generate network status monitoring data, wherein the network status monitoring data includes network topology data, network traffic data, and abnormal network status identification data;

[0155] Step S53: Perform dynamic bandwidth allocation processing based on network status monitoring data to generate a dynamic bandwidth allocation strategy;

[0156] Step S54: performing intelligent communication route planning based on the dynamic bandwidth allocation strategy to obtain intelligent communication route planning data;

[0157] Step S55: Adaptively allocate channels to the priority transmission list data and normal event data using the intelligent communication routing planning data to obtain target communication routing path data;

[0158] Step S56: Using the preset elevator shaft interference source to perform intelligent obstacle avoidance optimization processing on the target communication routing path data to generate optimal communication routing path data;

[0159] Step S57: perform communication capability evaluation on the priority sending list data and normal event data of the optimal communication routing path data to generate communication capability evaluation data; when the communication capability evaluation data satisfies the priority sending list data and the normal event data, it is transmitted to the cloud server in real time through the secure transmission protocol based on the optimal communication routing path data; when the communication capability evaluation data does not satisfy the priority sending list data and the normal event data, it is transmitted to the cloud server in real time through the secure transmission protocol, and the normal event data is temporarily stored in the elevator Internet of Things gateway for scheduled transmission, thereby obtaining a dynamic data sending strategy.

[0160] As an example of the present invention, refer to Figure 3 As shown, Figure 1 Detailed implementation steps of step S5 are shown in the flowchart. In this example, step S5 includes:

[0161] Step S51: sorting the abnormal event data by data transmission priority by modifying the urgency score data to obtain priority transmission list data;

[0162] In an embodiment of the present invention, abnormal events are sorted from high to low based on their scores. For example, an efficient algorithm such as quick sort can be used to reorder events based on their scores. The sorted abnormal events are added one by one to a priority sending list. For example, the sorted event queue is traversed, and events with a score higher than 90 are added to the sending list first, followed by events with lower scores. When generating the sending list, a maximum list length can be set. When the number of added events reaches the upper limit, the remaining lower-scoring events will not be added to the list temporarily. Events can also be divided into multiple priority levels based on their scores, and corresponding sending lists are generated for each, such as an "emergency event list," "important event list," or "general event list." For multiple events with the same score, additional sorting can be performed based on the time of occurrence, so that earlier events are sent first. The generated priority sending list data contains not only the abnormal event content, but also metadata such as the event ID, score, and time of occurrence.

[0163] Step S52: performing network status monitoring according to the elevator IoT gateway to generate network status monitoring data, wherein the network status monitoring data includes network topology data, network traffic data, and abnormal network status identification data;

[0164] In embodiments of the present invention, the gateway can use active detection methods such as Ping and Traceroute to discover the topology of nodes and links in the network and draw a complete network topology map. Network topology information can also be indirectly obtained through passively captured link layer detection protocol packets such as LLDP and CDP. The gateway can deploy a traffic monitoring agent to capture data flows passing through the gateway and analyze the real-time traffic status, such as bandwidth utilization, latency, and jitter. Traffic counters and other statistical data can also be obtained from network devices such as routers and switches through network management protocols such as SNMP. Based on traffic data, the gateway can set reasonable thresholds for metrics such as latency and packet loss rate. When a monitored metric exceeds the threshold, it determines that the network is in an abnormal state, such as congestion or link failure. Machine learning algorithms can also be applied to automatically learn normal patterns from historical traffic data and mark situations that exceed these patterns as abnormal. The gateway integrates the acquired topology data, traffic data, and abnormal state identification results to generate comprehensive network status monitoring data output. For example, abnormal states can be associated with corresponding topological locations such as network nodes and links. The gateway will continuously update network status monitoring data to reflect the latest network conditions in real time. Once an anomaly is detected, an early warning will be triggered immediately.

[0165] Step S53: Perform dynamic bandwidth allocation processing based on network status monitoring data to generate a dynamic bandwidth allocation strategy;

[0166] In an embodiment of the present invention, the gateway integrates network topology data and traffic data to construct a mathematical model of the current network. For example, a flow network model can be used, representing nodes as sources or sinks and links as bandwidth-constrained edges. A maximum flow algorithm is then used to determine the maximum available bandwidth in the network. Based on abnormal network status identification data, the gateway classifies network traffic flows, distinguishing between critical and standard traffic flows. Critical traffic, such as elevator abnormality event data transmission, requires higher bandwidth and transmission priority. The gateway dynamically schedules available bandwidth resources in the network using a bandwidth allocation algorithm (such as the Max-Min fair allocation algorithm). This algorithm ensures that critical traffic flows receive sufficient bandwidth while preventing standard traffic from overusing bandwidth and causing congestion. When allocating bandwidth, the gateway can set multiple optimization objectives, such as maximizing bandwidth utilization, minimizing latency, and minimizing packet loss rate, and balance these objectives based on priority to arrive at the optimal allocation solution. Genetic algorithms, ant colony algorithms, and other algorithms can be used to solve multi-objective optimization problems. The gateway will continuously monitor changes in network status. Once it discovers new congestion points, link failures, or other abnormal conditions, or changes in the priority of business flows, it will automatically re-execute the above modeling and optimization process, dynamically adjust the bandwidth allocation strategy, and send the new strategy to network nodes for execution.

[0167] Step S54: performing intelligent communication route planning based on the dynamic bandwidth allocation strategy to obtain intelligent communication route planning data;

[0168] In this embodiment of the present invention, based on bandwidth allocation policies, the gateway analyzes communication routing requirements, including source and destination nodes, bandwidth requirements, latency requirements, and priority. For example, prioritized transmission of list data requires transmission routes with higher bandwidth and lower latency. The gateway combines network topology data with bandwidth allocation policies to construct a constraint model for routing selection. For example, a latency-constrained shortest path problem model can be used, with link bandwidth and latency as hard constraints for routing selection. The gateway utilizes intelligent routing algorithms to plan the optimal route for each communication requirement, while satisfying these constraints. Classic algorithms such as Dijkstra and OSPF can be used, as can machine-learning-based intelligent routing algorithms such as reinforcement learning algorithms like Q-Routing, to autonomously discover the optimal routing strategy. For conflicting routing requirements, the gateway constructs a multi-objective optimization model to determine the optimal routing combination within multiple constraints, such as bandwidth, latency, and load balancing, to avoid resource waste and excessive congestion. Intelligent optimization algorithms such as genetic algorithms can be used for this purpose. As network status and service demands change, the gateway continuously optimizes its routing strategy. Once new link failures, bandwidth changes, and other events are discovered, the route will be automatically replanned and the new routing policy will be issued for execution.

[0169] Step S55: Adaptively allocate channels to the priority transmission list data and normal event data using the intelligent communication routing planning data to obtain target communication routing path data;

[0170] In an embodiment of the present invention, the gateway maps the data to different service levels based on its type (priority sending list data or normal event data) and importance, and determines the corresponding transmission priority. For example, priority sending list data is mapped to the high-priority "multicast acceleration" service, and normal event data is mapped to the normal-priority "background download" service. The gateway divides the currently available network channel resources (such as bandwidth, time slots, etc.) into a high-priority channel resource pool and a normal-priority channel resource pool according to the bandwidth allocation strategy to carry data transmission requirements of different priorities. The gateway will monitor the usage of channel resources in real time. Once it is found that the high-priority channel resources are insufficient or the normal channel resources are wasted, the bandwidth strategy will be implemented to dynamically adjust the size of the two resource pools and reallocate channel resources to ensure the transmission quality of high-priority data. Through the above-mentioned adaptive channel allocation process, the gateway will output the target communication routing path data for each data transmission requirement, including detailed information such as the source point, destination point, links and nodes passed through, and allocated channel resources.

[0171] Step S56: Using the preset elevator shaft interference source to perform intelligent obstacle avoidance optimization processing on the target communication routing path data to generate optimal communication routing path data;

[0172] In this embodiment of the present invention, the gateway constructs a three-dimensional elevator shaft environment model based on previously acquired elevator shaft structural drawings, on-site surveys, and other information. This model describes the spatial location and physical properties of obstacles within the elevator shaft (such as reinforced concrete walls and elevator cars). The gateway combines the target communication routing path data with the elevator shaft environment model and, using algorithms such as ray propagation models, simulates and analyzes the propagation of radio signals within the elevator shaft, calculating the effects of each obstacle on signal attenuation, reflection, and diffraction. Based on the signal propagation simulation results, the gateway identifies obstacles that significantly impact communication quality as interference sources and records parameters such as their locations and attenuation characteristics as interference source models. Common interference sources include reinforced concrete walls, metal elevator cars, and pipelines. The gateway uses intelligent optimization algorithms (such as genetic algorithms and ant colony algorithms) to optimize the target routing path for obstacle avoidance. With maximizing signal quality (such as signal-to-noise ratio) as the objective function, the gateway autonomously discovers the optimal routing solution to bypass interference sources while meeting communication requirements (such as bandwidth and latency). For the candidate optimal routes output by the algorithm, the gateway will re-run the signal propagation simulation to evaluate their actual communication quality. If the quality meets the requirements, it will be confirmed as the final optimal communication route data; otherwise, the obstacle avoidance optimization will be re-run until a satisfactory solution is found.

[0173] Step S57: perform communication capability evaluation on the priority sending list data and normal event data of the optimal communication routing path data to generate communication capability evaluation data; when the communication capability evaluation data satisfies the priority sending list data and the normal event data, it is transmitted to the cloud server in real time through the secure transmission protocol based on the optimal communication routing path data; when the communication capability evaluation data does not satisfy the priority sending list data and the normal event data, it is transmitted to the cloud server in real time through the secure transmission protocol, and the normal event data is temporarily stored in the elevator Internet of Things gateway for scheduled transmission, thereby obtaining a dynamic data sending strategy.

[0174] In an embodiment of the present invention, the gateway calculates the total amount of priority list data and normal event data, including the number and size of packets. Using network simulation tools or analytical models, the gateway calculates the actual data throughput capacity of the path under the constraints of the optimal routing path (such as bandwidth, latency, and signal-to-noise ratio). In addition to throughput, the gateway also evaluates the end-to-end latency of data transmission from the source to the cloud server under the optimal route to ensure that the real-time transmission requirements of the priority list data are met. The gateway also analyzes the reliability of the optimal route, such as link redundancy and node fault tolerance, to assess the success rate and stability of data transmission in the event of link failures or congestion. The evaluation results of multiple dimensions, such as throughput, latency, and reliability, are combined to generate communication capability assessment data to determine whether the current optimal route is capable of simultaneously transmitting both types of data. If the assessment results meet the requirements, the gateway will transmit the priority list and normal event data to the cloud server in real time based on the optimal route using a secure transmission protocol (such as HTTPS or VPN). When the evaluation results are not satisfied, the gateway will only transmit priority list data through the security protocol to ensure real-time reporting of key data; at the same time, normal event data will be temporarily stored in the local cache, and transmitted to the cloud in batches regularly according to changes in network status to obtain a dynamic data sending strategy.

[0175] Preferably, step S53 includes the following steps:

[0176] Step S531: constructing a network topology graph based on the network topology data to obtain a node adjacency matrix;

[0177] Step S532: Perform time series analysis on the network traffic data to obtain time series traffic prediction data;

[0178] Step S533: marking network congested nodes based on the abnormal network status identification data to obtain network congested node data;

[0179] Step S534: performing network hierarchical structure analysis using the node adjacency matrix, the time series traffic prediction data, and the network congestion node data to obtain a network hierarchical structure matrix;

[0180] Step S535: Divide the time axis into multiple time slices, perform bandwidth allocation processing on each time slice through the network hierarchy matrix, and generate bandwidth allocation tokens;

[0181] Step S536: Send the bandwidth allocation token to the corresponding transmission node through the elevator IoT gateway, and continuously monitor changes in network status. When a new congestion point or drastic traffic fluctuation is found, re-execute steps S531-S535 to obtain a dynamically adjusted bandwidth allocation strategy.

[0182] In an embodiment of the present invention, graph theory algorithms (such as breadth-first / depth-first search) are used to discover the topology of nodes and links in the network, constructing a complete network topology map. Based on this topology map, the adjacency relationships between each node are calculated, generating a node adjacency matrix as a mathematical representation of the network model. For example, if node i is adjacent to node j, the adjacency matrix M[i][j] = 1; otherwise, it is 0. Time series prediction algorithms (such as ARIMA and Prophet) are used to predict traffic trends over a period of time. This predicted data can accurately depict the dynamic changes in network traffic over time. Congested nodes in the current network are marked to obtain network congested node data. Congested nodes can be identified by threshold judgment based on indicators such as latency and packet loss rate, or by using machine learning algorithms (such as k-means clustering) to automatically discover congestion patterns. A network hierarchy analysis algorithm (such as K-ray clustering) is used to analyze the network's hierarchical structure. The network is divided into multiple levels, and a network hierarchy matrix describing the relationships between nodes at each level is obtained. Bandwidth allocation is optimized for network traffic within each time slice. For example, the Max-Min fair allocation algorithm is used to ensure that high-level nodes obtain sufficient bandwidth, while low-level nodes are allocated based on the remaining bandwidth to prevent low-priority services from excessively occupying bandwidth resources. The allocation result will generate a bandwidth allocation token. The bandwidth allocation token is sent to the corresponding network transmission node, and each node executes the corresponding bandwidth control strategy based on the token. At the same time, the gateway will continuously monitor the real-time status of the network. Once a new congestion point is found or the network traffic fluctuates sharply, steps S531-S535 will be re-executed to obtain a dynamically adjusted bandwidth allocation strategy. For example, if the gateway detects a sudden increase in network traffic and identifies a new congested node, steps S531-S535 will be re-executed to update the network topology data, time-series traffic forecast data, and network congested node data, and regenerate bandwidth allocation tokens to allocate more bandwidth to the network layer where the congested node is located.

[0183] The present invention also provides a communication system based on the elevator Internet of Things, which executes the data transmission method based on the elevator Internet of Things as described above. The communication system based on the elevator Internet of Things includes:

[0184] The elevator monitoring response module is used to use the elevator controller to perform wireless network access processing, collect periodic operation status data of the elevator, and generate periodic operation status data; match the operation mode based on the periodic operation status data to generate elevator operation mode data; collect sensor response data stream based on the elevator operation mode data to generate real-time monitoring response data;

[0185] The monitoring data fusion module is used to integrate the sensor source responses based on the real-time monitoring response data to generate comprehensive monitoring response data; construct a multi-source monitoring fusion matrix based on the comprehensive monitoring response data; and perform elevator status coding and identification processing on the multi-source monitoring fusion matrix to generate real-time elevator status snapshot data;

[0186] The sending data classification module is used to build a time series operation trend prediction model based on the elevator operation mode data; perform expected feedforward calculation on the real-time elevator status snapshot data using the real-time elevator status snapshot data to obtain real-time monitoring feature offset data; perform offset characteristic analysis based on the real-time monitoring feature offset data to obtain abnormal event data and normal event data respectively;

[0187] The emergency communication assessment module performs abnormal pattern matching on the abnormal event data to obtain abnormal pattern indicator data; performs urgency scoring processing based on the abnormal pattern indicator data to generate urgency score data; performs urgency correction processing on the urgency score data to obtain corrected urgency score data;

[0188] The adaptive communication optimization module is used to monitor the network status of the elevator Internet of Things gateway and generate network status monitoring data; perform dynamic bandwidth allocation processing based on the network status monitoring data to generate a dynamic adjustment bandwidth allocation strategy; perform adaptive channel allocation based on the dynamic adjustment bandwidth allocation strategy, and perform intelligent obstacle avoidance optimization processing to generate optimal communication routing path data; and perform dynamic data transmission processing on the corrected urgency score data and normal event data based on the optimal communication routing path data to obtain a dynamic data transmission strategy.

[0189] The beneficial effects of this application lie in utilizing the elevator controller for wireless network access processing, achieving seamless connectivity between the elevator system and the Internet of Things (IoT), enabling comprehensive real-time elevator operation status. Operational mode matching based on periodic operating status data accurately identifies elevator operating modes, such as normal operation, peak operation, and maintenance operation. A multi-source monitoring fusion matrix is constructed based on real-time monitoring response data, enabling matrix processing of multiple sensor data to more accurately identify the elevator's operating status and improve diagnostic accuracy. Expected feedforward calculations using real-time elevator status snapshot data compare the real-time status with the expected state, enabling timely detection of abnormal deviations. Analysis of deviation characteristics based on real-time monitoring feature deviation data separates abnormal events from normal events, enabling prediction and early warning of elevator operating status. Abnormal pattern matching of abnormal event data allows specific abnormal patterns, such as failure, jam, and overload, to be matched based on abnormal event characteristics. Urgency scoring based on abnormal pattern indicator data and urgency correction processing of the urgency score data allow dynamic adjustment of the initial score based on actual conditions, improving the accuracy of urgency assessment. This achieves intelligent identification and assessment of elevator faults. By monitoring the network status of the elevator IoT gateway, network issues such as disconnected connections, insufficient bandwidth, and signal interference can be promptly identified, enabling real-time monitoring of the elevator IoT system's network status. Based on this network status monitoring data, bandwidth allocation strategies are dynamically adjusted, and intelligent obstacle avoidance algorithms are used to optimize data transmission paths, avoiding network congestion and data loss, and improving data transmission efficiency. Intelligent data transmission management adjusts data transmission strategies based on event urgency and network conditions, improving data transmission efficiency, ensuring data transmission stability and reliability, and prioritizing the transmission of important data.

[0190] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0191] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A data transmission method based on elevator Internet of Things, characterized in that: The following steps are involved: Step S1: Using the elevator controller to perform wireless network access processing, and collect periodic operation status data of the elevator to generate periodic operation status data; Perform operation mode matching based on periodic operation status data to generate elevator operation mode data; Collect sensor response data streams based on elevator operation mode data to generate real-time monitoring response data; Step S2: performing sensor source response integration processing based on the real-time monitoring response data to generate comprehensive monitoring response data; Build a multi-source monitoring fusion matrix based on comprehensive monitoring response data; perform elevator status coding and identification processing on the multi-source monitoring fusion matrix to generate real-time elevator status snapshot data; Step S3: constructing a time series operation trend prediction model based on the elevator operation mode data; performing expected feedforward calculation on the real-time elevator state snapshot data using the real-time elevator state snapshot data to obtain real-time monitoring feature offset data; Perform offset characteristic analysis based on real-time monitoring feature offset data to obtain abnormal event data and normal event data respectively; Step S4: performing abnormal pattern matching on the abnormal event data to obtain abnormal pattern indicator data; Perform urgency scoring processing based on abnormal pattern indicator data to generate urgency scoring data; Performing urgency correction processing on the urgency score data to obtain corrected urgency score data; Step S5: Monitor the network status according to the elevator IoT gateway and generate network status monitoring data; Perform dynamic bandwidth allocation processing based on network status monitoring data and generate dynamic bandwidth allocation strategies; Based on the dynamic adjustment of bandwidth allocation strategy, adaptive channel allocation is performed, and intelligent obstacle avoidance optimization processing is carried out to generate optimal communication routing path data; according to the optimal communication routing path data, dynamic data transmission processing is performed on the corrected urgency score data and normal event data to obtain a dynamic data transmission strategy.

2. The data transmission method based on elevator Internet of Things according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: Using the elevator controller to perform wireless network access processing on the elevator IoT gateway and generate a wireless network access credential; Step S12: performing communication request security authentication processing on the cloud server according to the wireless network access credential to generate communication security authentication token data; Step S13: using the elevator controller to collect periodic operation status data of the elevator based on the communication security authentication token data to generate periodic operation status data; Step S14: performing time series analysis on the periodic operation status data to generate periodic operation time series data, wherein the periodic operation time series data includes elevator operation status data, elevator load data, and elevator floor scheduling data; Step S15: performing elevator load analysis based on the periodic operation status data, and performing operation mode matching to generate elevator operation mode data; Step S16: extracting operation mode monitoring indicators based on the elevator operation mode data to generate mode monitoring indicator data; Step S17: Based on the pattern monitoring index data, the elevator controller controls the sensor to collect the sensor response data stream to generate real-time monitoring response data.

3. The data transmission method based on elevator Internet of Things according to claim 2 is characterized in that: Step S15 includes the following steps: Step S151: performing elevator motion trajectory analysis based on elevator operation status data to generate elevator motion trajectory data; Step S152: performing elevator passenger load flow analysis on the elevator load data using the elevator motion trajectory data to generate passenger load flow distribution data; Step S153: performing cluster analysis on the passenger load flow distribution data to generate elevator load cluster data; Step S154: performing load-time correlation processing on the elevator running time in the elevator running status data using the elevator load clustering data to generate load-time correlation data; Step S155: performing elevator load calculation on the elevator load clustering data using the load time correlation data to obtain elevator average load data and elevator peak load data; performing load pattern recognition based on the elevator average load data and elevator peak load data to generate elevator load pattern data; Step S156: extracting passenger behavior patterns from the elevator floor dispatch data using the elevator load pattern data to generate passenger behavior pattern data; Step S157: performing operation mode matching on a preset elevator operation mode library through the passenger behavior pattern data and the elevator load pattern data to generate elevator operation mode data.

4. The data transmission method based on elevator Internet of Things according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: The elevator controller transmits the real-time monitoring response data to the elevator IoT gateway for sensor source preprocessing to generate sensor source preprocessing data; Step S22: performing sensor source response integration processing based on the sensor source preprocessing data to generate comprehensive monitoring response data; Step S23: constructing a sensor source key feature space for the comprehensive monitoring response data to generate sensor source feature vector data; Step S24: constructing a multi-source monitoring fusion matrix based on the comprehensive monitoring response data and the sensor source feature vector data; Step S25: performing principal component analysis based on the multi-source monitoring fusion matrix to obtain multi-source key response feature data; Step S26: performing elevator status coding and identification processing on the multi-source key response feature data to generate real-time elevator status snapshot data.

5. The data transmission method based on elevator Internet of Things according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: extracting state benchmark indicators based on elevator operation mode data to generate state benchmark indicator data; Step S32: performing transfer learning processing on the state benchmark indicator data through a preset long short-term memory network model, thereby constructing a time series operation trend prediction model; Step S33: Acquire historical elevator monitoring response data; perform expected feedforward calculation on the historical elevator monitoring response data using a time series operation trend prediction model to generate expected elevator operation characteristic data; Step S34: using the real-time elevator status snapshot data to compare the elevator expected operation feature data with the feature values of each dimension, and performing distance metric calculation to obtain real-time monitoring feature offset data; Step S35: performing an offset characteristic analysis based on the real-time monitoring characteristic offset data to generate elevator state offset characteristic data, wherein the elevator state offset characteristic data includes normal operation characteristic data and abnormal offset characteristic data; Step S36: performing anomaly detection processing on the multi-source monitoring fusion matrix using the abnormal offset feature data to obtain abnormal event data and preliminary normal event data; Step S37: performing monitoring event data fusion on the normal operation feature data and the preliminary normal event data to generate normal event data.

6. The data transmission method based on elevator Internet of Things according to claim 5 is characterized in that: Step S36 includes the following steps: Step S361: performing real-time data stream matching on the multi-source monitoring fusion matrix using abnormal offset feature data to generate a real-time offset data stream; Step S362: Divide the real-time offset data stream into time series windows according to a preset time window to generate real-time offset data segments; Step S363: performing offset point identification on the abnormal offset feature data based on the real-time offset data segment to generate target offset point index data; Step S364: performing feature importance evaluation on the real-time offset data segments and calculating the optimal segmentation value to obtain segmentation decision data; Step S365: Using the target offset point index data as a tree child node, an isolation tree is constructed on the real-time offset data segment using an isolation tree algorithm based on the segmentation decision data to generate a feature offset isolation tree; Step S366: Calculate the path length based on the feature offset isolation tree to generate offset point path length data; perform time sensitivity adjustment on the offset point path length data to generate time offset weight data; Step S367: Using the time offset weight data, perform weighted anomaly scoring on the offset point path length data to generate offset point anomaly score data; Step S368: abnormal event judgment is performed on the offset point abnormality score data using a preset abnormality judgment threshold. When the offset point abnormality score data is lower than the offset point abnormality score data, the relevant offset point and the real-time offset data segment are marked as preliminary normal event data; when the offset point abnormality score data is higher than or equal to the offset point abnormality score data, the relevant offset point and the real-time offset data segment are marked as preliminary abnormal event data. Step S369: Match the preliminary abnormal event data with the elevator operation log by timestamp, and fuse the event data to obtain abnormal event data.

7. The data transmission method based on elevator Internet of Things according to claim 1 is characterized in that: Step S4 includes the following steps: Step S41: Identify the abnormal event type of the abnormal event data and generate abnormal type identification data; Step S42: performing abnormal pattern matching on the abnormal type identification data using a preset elevator fault model library, and extracting abnormal pattern indicators to generate abnormal pattern indicator data; Step S43: performing severity assessment based on the abnormal pattern indicator data to obtain an abnormal type severity index; Step S44: performing an urgency assessment on the abnormality type severity index and the abnormality pattern indicator data based on a fuzzy logic inference algorithm to generate an urgency factor; Step S45: performing urgency scoring processing on the urgency factor to generate urgency scoring data; Step S46: performing event context analysis based on the abnormal event data to generate event context analysis data; performing fault impact analysis on the event context analysis data using the abnormal type severity index to generate an abnormal event impact index; Step S47: Using the abnormal event impact index, the urgency score data is subjected to urgency correction processing to obtain corrected urgency score data.

8. The data transmission method based on elevator Internet of Things according to claim 1 is characterized in that: Step S5 includes the following steps: Step S51: sorting the abnormal event data by data transmission priority by modifying the urgency score data to obtain priority transmission list data; Step S52: performing network status monitoring according to the elevator IoT gateway to generate network status monitoring data, wherein the network status monitoring data includes network topology data, network traffic data, and abnormal network status identification data; Step S53: Perform dynamic bandwidth allocation processing based on network status monitoring data to generate a dynamic bandwidth allocation strategy; Step S54: performing intelligent communication route planning based on the dynamic bandwidth allocation strategy to obtain intelligent communication route planning data; Step S55: Adaptively allocate channels to the priority transmission list data and normal event data using the intelligent communication routing planning data to obtain target communication routing path data; Step S56: Using the preset elevator shaft interference source to perform intelligent obstacle avoidance optimization processing on the target communication routing path data to generate optimal communication routing path data; Step S57: perform communication capability evaluation on the priority sending list data and normal event data of the optimal communication routing path data to generate communication capability evaluation data; when the communication capability evaluation data satisfies the priority sending list data and the normal event data, it is transmitted to the cloud server in real time through the secure transmission protocol based on the optimal communication routing path data; when the communication capability evaluation data does not satisfy the priority sending list data and the normal event data, it is transmitted to the cloud server in real time through the secure transmission protocol, and the normal event data is temporarily stored in the elevator Internet of Things gateway for scheduled transmission, thereby obtaining a dynamic data sending strategy.

9. The data transmission method based on elevator Internet of Things according to claim 8, characterized in that: Step S53 includes the following steps: Step S531: constructing a network topology graph based on the network topology data to obtain a node adjacency matrix; Step S532: Perform time series analysis on the network traffic data to obtain time series traffic prediction data; Step S533: marking network congested nodes based on the abnormal network status identification data to obtain network congested node data; Step S534: performing network hierarchical structure analysis using the node adjacency matrix, the time series traffic prediction data, and the network congestion node data to obtain a network hierarchical structure matrix; Step S535: Divide the time axis into multiple time slices, perform bandwidth allocation processing on each time slice through the network hierarchy matrix, and generate bandwidth allocation tokens; Step S536: Send the bandwidth allocation token to the corresponding transmission node through the elevator IoT gateway, and continuously monitor changes in network status. When a new congestion point or drastic traffic fluctuation is found, re-execute steps S531-S535 to obtain a dynamically adjusted bandwidth allocation strategy.

10. A communication system based on elevator Internet of Things, characterized in that: For executing the data transmission method based on elevator Internet of Things according to claim 1, the communication system based on elevator Internet of Things includes: The elevator monitoring response module is used to use the elevator controller to perform wireless network access processing, collect periodic operation status data of the elevator, and generate periodic operation status data; match the operation mode based on the periodic operation status data to generate elevator operation mode data; collect sensor response data stream based on the elevator operation mode data to generate real-time monitoring response data; The monitoring data fusion module is used to integrate the sensor source responses based on the real-time monitoring response data to generate comprehensive monitoring response data; construct a multi-source monitoring fusion matrix based on the comprehensive monitoring response data; and perform elevator status coding and identification processing on the multi-source monitoring fusion matrix to generate real-time elevator status snapshot data; The sending data classification module is used to build a time series operation trend prediction model based on the elevator operation mode data; perform expected feedforward calculation on the real-time elevator status snapshot data using the real-time elevator status snapshot data to obtain real-time monitoring feature offset data; perform offset characteristic analysis based on the real-time monitoring feature offset data to obtain abnormal event data and normal event data respectively; The emergency communication assessment module performs abnormal pattern matching on the abnormal event data to obtain abnormal pattern indicator data; performs urgency scoring processing based on the abnormal pattern indicator data to generate urgency score data; performs urgency correction processing on the urgency score data to obtain corrected urgency score data; The adaptive communication optimization module is used to monitor the network status of the elevator Internet of Things gateway and generate network status monitoring data; perform dynamic bandwidth allocation processing based on the network status monitoring data to generate a dynamic adjustment bandwidth allocation strategy; perform adaptive channel allocation based on the dynamic adjustment bandwidth allocation strategy, and perform intelligent obstacle avoidance optimization processing to generate optimal communication routing path data; and perform dynamic data transmission processing on the corrected urgency score data and normal event data based on the optimal communication routing path data to obtain a dynamic data transmission strategy.

Images