Escalator fault diagnosis system and method based on strain pressure and acceleration

Abstract

The invention discloses an escalator fault diagnosis system and method based on strain pressure and acceleration. The system comprises a far-end server and a plurality of step monitoring units. The step monitoring unit is arranged at the bottom of a step pedal surface of the escalator and is used for collecting strain pressure data and acceleration vibration data of a step and carrying out working condition judgment and preliminary fault detection; and the far-end server is in communication connection with each step monitoring unit through a network, and is used for receiving data transmitted by the step monitoring units, carrying out comprehensive analysis by combining load characteristics, vibration characteristics and working condition characteristics, and realizing fault early warning, diagnosis and fault occurrence time prediction of the escalator. According to the scheme, the limitation of traditional escalator monitoring is broken through, field measurement is not needed, the accuracy and timeliness of fault diagnosis and the escalator operation safety are improved, and the maintenance cost is reduced.

Description

Technical Field

[0001] This invention relates to the field of escalator fault diagnosis technology, specifically to an escalator fault diagnosis system and method based on strain pressure and acceleration. Background Technology

[0002] Escalators, as widely used vertical transportation equipment in public places, are directly related to the safety and reliability of public travel. With the increase of service life and changes in operating load, key components such as escalator steps and chains are prone to wear, deformation and other failures. If these are not detected and dealt with in a timely manner, they may lead to safety accidents.

[0003] However, existing technologies have many shortcomings in escalator fault diagnosis:

[0004] Traditional escalator steps are only equipped with targeted weighing devices, which have a single function and can only realize load detection. They cannot provide early warning and diagnosis of potential faults in the steps themselves and the transmission chain, making it difficult to detect hidden safety hazards in a timely manner.

[0005] The existing escalator detection devices are complex in structure and have cumbersome operation procedures. They require staff to frequently go to the site for manual measurement, which not only wastes a lot of manpower but also has low detection efficiency and is not conducive to daily maintenance and routine monitoring.

[0006] Current monitoring systems lack the ability to monitor the deformation of step treads and the operating status of steps in real time and comprehensively. This makes it impossible to accurately assess the overall operating performance of escalators, resulting in a lack of data support for maintenance decisions and affecting maintenance efficiency and escalator operating safety.

[0007] Existing technologies mostly rely solely on load data or vibration data for fault diagnosis, failing to effectively combine load characteristics, vibration characteristics, and operating condition characteristics for comprehensive analysis. This lack of a comprehensive understanding of the escalator's operating status makes it difficult to accurately identify the root cause of the fault.

[0008] When a fault occurs, the existing system cannot quickly locate the cause of the fault, resulting in a delayed maintenance response. Furthermore, it lacks intelligent fault prediction capabilities and cannot anticipate the trend of fault occurrence in advance, further reducing the safety and reliability of escalator operation.

[0009] Therefore, developing an escalator fault diagnosis system capable of real-time monitoring, comprehensive analysis, and intelligent early warning diagnosis has become a pressing technical problem for the industry. Summary of the Invention

[0010] This invention provides an escalator fault diagnosis system and method based on strain pressure and acceleration, aiming to solve the problem that the existing escalator fault diagnosis lacks comprehensive monitoring, integrated analysis capabilities and intelligent early warning and diagnosis functions, so as to realize accurate early warning, rapid diagnosis and early prediction of escalator faults, and improve the safety of escalator operation and maintenance efficiency.

[0011] To achieve the above objectives, the present invention is implemented through the following technical solution:

[0012] In a first aspect, the present invention provides an escalator fault diagnosis system based on strain pressure and acceleration, the system comprising multiple step monitoring units and a remote server;

[0013] The step monitoring unit is installed at the bottom of the step tread surface of the escalator to collect strain pressure data and acceleration vibration data of the steps, and to perform working condition judgment and preliminary fault detection.

[0014] The remote server communicates with each step monitoring unit via a network to receive data transmitted by the step monitoring units, and performs comprehensive analysis by combining load characteristics, vibration characteristics and operating condition characteristics to realize escalator fault early warning, diagnosis and fault occurrence time prediction.

[0015] As a further improvement to the technical solution of the present invention, the cascade monitoring unit includes a strain pressure sensor, an acceleration vibration sensor, a wireless transmission module, and an edge computing module; the strain pressure sensor, the acceleration vibration sensor, and the wireless transmission module are all electrically connected to the edge computing module.

[0016] The strain pressure sensor and the acceleration vibration sensor work in parallel, collecting data from two core dimensions: load deformation and operational vibration, respectively, to provide a comprehensive data source for subsequent analysis. The edge computing module enables local preliminary data processing, reducing data transmission volume and quickly completing condition judgment and preliminary fault screening, while the wireless transmission module ensures the stability and timeliness of remote data interaction.

[0017] As a further improvement to the technical solution of the present invention, the strain pressure sensor is used to monitor the deformation degree of the step tread surface in real time, and converts the monitoring data into pressure value distribution of the step tread surface to characterize the step load and pressure distribution state.

[0018] Unlike traditional single-point weighing methods, strain pressure sensors can acquire pressure distribution data across the entire step surface, rather than the deformation at a single monitoring point. This provides a complete description of the stress state of the steps, avoiding monitoring deviations caused by uneven local stress, and offering more accurate data support for load characteristic analysis.

[0019] As a further improvement to the technical solution of the present invention, the acceleration vibration sensor is used to collect the running speed, vibration frequency and vibration data of the steps under different loads in real time when the running speed of the steps reaches a preset value and the tread surface is facing upward.

[0020] The preset value can be flexibly set according to the escalator's designed operating speed, such as 0.5 m / s. By limiting the sensor's operating conditions, it ensures that the collected data are vibration information from the effective operating phase of the steps, avoiding interference from invalid data and improving data quality. Simultaneously, the sensor can record vibration changes under different load conditions, laying the foundation for subsequent correlation analysis.

[0021] As a further improvement to the technical solution of the present invention, the edge computing module is used to receive the monitoring data from the strain pressure sensor and the acceleration vibration sensor, determine the cascade working condition, and extract feature values ​​from the monitoring data to complete the preliminary fault detection.

[0022] The edge computing module has built-in operating condition judgment logic, which can identify in real time whether the ladder is in different operating conditions such as manned operation or no-load operation based on sensor data. During the feature extraction process, key indicators such as peak value, mean, variance, and frequency spectrum characteristics of time series data can be extracted. By comparing with preset normal thresholds, it can quickly determine whether there are any abnormalities, realize preliminary fault screening, and reduce the data processing pressure on the remote server.

[0023] As a further improvement to the technical solution of the present invention, the remote server adopts big data processing technology to integrate and analyze the received monitoring data and feature values. The integration and analysis includes correlation analysis of load characteristics, vibration characteristics and operating condition characteristics.

[0024] Big data processing technology can efficiently integrate massive amounts of time-series data. By mining the inherent correlation between load characteristics, vibration characteristics, and operating condition characteristics, such as the normal vibration frequency range under different loads and the deformation threshold under specific operating conditions, a multi-dimensional fault judgment model can be established, breaking through the limitations of single feature analysis in existing technologies and improving the accuracy of fault identification.

[0025] As a further improvement to the technical solution of the present invention, the stepped monitoring unit processes the collected strain and pressure data as follows:

[0026] If the strain pressure value is greater than the preset range, it is determined that the step has completed the entire process of the tread facing upwards.

[0027] If the strain pressure value changes significantly during operation and sometimes returns to zero, then delete that segment of strain pressure data.

[0028] If the strain pressure value changes significantly but does not return to zero, the strain pressure data will be discarded and interpolated.

[0029] If the strain pressure value suddenly jumps and then remains stable, the strain pressure data should be segmented.

[0030] The aforementioned targeted data processing rules effectively filter out interference data caused by unstable operating conditions such as personnel walking and brief stops, ensuring that all data transmitted to the remote server is valid and stable sample data, providing a high-quality data foundation for model training and fault analysis. The preset range can be calibrated based on the rated load and structural characteristics of the cascade.

[0031] As a further improvement to the technical solution of the present invention, the remote server compares the predicted vibration with the actual vibration using the trained model. If the difference exceeds a preset threshold, a fault alert is issued, and fault diagnosis is performed based on the entire vibration data. The fault occurrence date is predicted by combining historical difference trends.

[0032] The model is generated based on the processed effective sample set and can establish the correspondence between strain changes and acceleration vibration changes. By analyzing the difference between predicted and true values, it can accurately identify abnormal states. The preset threshold can be dynamically adjusted according to factors such as the escalator's operating years and equipment model. During the fault diagnosis process, the frequency, amplitude and other characteristics of vibration data can be combined to locate the fault type (such as chain wear, step deformation, etc.). Historical difference trend analysis can predict the fault development pattern and provide a basis for maintenance plan formulation.

[0033] As a further improvement to the technical solution of the present invention, the wireless transmission module is used to transmit the monitoring data and feature values ​​processed by the edge computing module to the remote server when the escalator enters the low-speed mode, and to receive the parameters and models sent by the remote server.

[0034] Selecting a low-speed mode for escalator data transmission (such as maintenance mode or empty return mode) can avoid data transmission delays or losses during high-speed operation, ensuring data transmission stability. At the same time, the remote server can use this module to send threshold parameters, model update files, etc., to achieve remote system upgrades and parameter calibration, improving system adaptability.

[0035] Secondly, the present invention provides an escalator fault diagnosis method based on strain pressure and acceleration, applied to the aforementioned escalator fault diagnosis system based on strain pressure and acceleration, comprising the following steps:

[0036] S1. The cascade monitoring unit collects strain and pressure data and acceleration and vibration data of the cascade to determine the working condition of the cascade and perform preliminary fault detection.

[0037] S2. When the escalator enters low-speed mode, the step monitoring unit transmits the processed monitoring data and feature values ​​to the remote server.

[0038] S3. The remote server integrates and analyzes the received data, and performs correlation analysis by combining load characteristics, vibration characteristics and working condition characteristics.

[0039] S4. The remote server compares the predicted vibration with the actual vibration using a model. If the difference exceeds a threshold, it issues a fault alert and diagnoses the fault, and combines historical data to predict the time of the fault occurrence.

[0040] This method follows a logical process of data acquisition, local processing, remote analysis, and intelligent diagnosis. Through the collaborative work of tiered monitoring units and remote servers, it achieves fully automated processing from data acquisition to fault prediction without human intervention, thereby improving the real-time performance and intelligence level of fault diagnosis.

[0041] The technical solution of the present invention has the following advantages over the prior art:

[0042] This invention relates to an escalator fault diagnosis system based on strain-pressure and acceleration. By installing a step monitoring unit at the bottom of the step tread, it achieves real-time acquisition, condition judgment, and preliminary fault detection of step strain-pressure data and acceleration vibration data. Combined with comprehensive analysis of load characteristics, vibration characteristics, and condition characteristics by a remote server, it not only breaks through the limitation of traditional escalators that can only achieve a single weighing function, but also can promptly detect potential safety hazards of steps and step chains. Furthermore, it simplifies the structure of the detection device through wireless remote communication, avoiding the waste of manpower costs caused by frequent on-site measurements. At the same time, by using multi-feature correlation analysis and fault prediction mechanisms, it comprehensively improves the completeness of escalator performance evaluation, the accuracy and timeliness of fault diagnosis, provides scientific data support for maintenance decisions, and significantly improves the safety, reliability, and operation and maintenance efficiency of escalators. Attached Figure Description

[0043] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0044] Figure 1 This is a schematic diagram of the structure of an escalator fault diagnosis system based on strain pressure and acceleration, according to some embodiments of the present invention;

[0045] Figure 2 This is a schematic diagram of the internal structure of a cascade monitoring unit according to some embodiments of the present invention;

[0046] Figure 3 This is an exemplary flowchart of an escalator fault diagnosis method based on strain pressure and acceleration, as shown in some embodiments of the present invention. Detailed Implementation

[0047] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0048] The present invention will be further described in detail below with reference to the accompanying drawings.

[0049] Reference Figure 1 In a first aspect, the present invention provides an escalator fault diagnosis system based on strain pressure and acceleration, the system comprising multiple step monitoring units and a remote server;

[0050] The step monitoring unit is installed at the bottom of the step tread surface of the escalator to collect strain pressure data and acceleration vibration data of the steps, and to perform working condition judgment and preliminary fault detection.

[0051] The remote server communicates with each step monitoring unit via a network to receive data transmitted by the step monitoring units, and performs comprehensive analysis by combining load characteristics, vibration characteristics and operating condition characteristics to realize escalator fault early warning, diagnosis and fault occurrence time prediction.

[0052] It should be noted that the step monitoring unit is installed at the bottom of the escalator step tread surface, collecting strain, pressure, acceleration, and vibration data of the steps in real time. It uses built-in logic to determine the step's operating condition and perform preliminary fault detection. The remote server establishes communication with all step monitoring units via the network. After receiving the transmitted data, it integrates load characteristics, vibration characteristics, and operating condition characteristics to perform multi-dimensional correlation analysis, constructing a fault judgment model. It provides fault warnings by comparing the differences between predicted and actual data, completes fault diagnosis by combining vibration data characteristics, and estimates the fault occurrence time based on historical difference trends.

[0053] This invention breaks through the limitations of traditional escalators' single weighing function, realizing full-process management of step and step chain faults, forming a closed loop from early warning, diagnosis to prediction. It eliminates the need for frequent on-site measurements by staff, reducing labor costs and maintenance difficulty. Through multi-feature comprehensive analysis, it improves the accuracy and timeliness of fault diagnosis, providing all-round protection for the safe operation of escalators and significantly improving operation and maintenance efficiency and reliability.

[0054] Reference Figure 2 In some embodiments, the cascade monitoring unit includes a strain pressure sensor, an acceleration vibration sensor, a wireless transmission module, and an edge computing module; the strain pressure sensor, the acceleration vibration sensor, and the wireless transmission module are all electrically connected to the edge computing module.

[0055] The strain pressure sensor and the acceleration vibration sensor work in parallel, collecting data from two core dimensions: load deformation and operational vibration, respectively, to provide a comprehensive data source for subsequent analysis. The edge computing module enables local preliminary data processing, reducing data transmission volume and quickly completing condition judgment and preliminary fault screening, while the wireless transmission module ensures the stability and timeliness of remote data interaction.

[0056] It should be noted that the strain pressure sensor and acceleration vibration sensor are responsible for data acquisition, converting physical signals into electrical signals and transmitting them to the edge computing module. The data processed by the edge computing module interacts with the remote server via a wireless transmission module, enabling data uploading and command reception. The modular design simplifies the structure of the tiered monitoring unit, with each module having a clear division of labor and working together efficiently, ensuring the continuity and stability of data acquisition, processing, and transmission. The electrical connection method ensures the accuracy of signal transmission, providing a high-quality data foundation for subsequent data analysis and fault diagnosis, while also facilitating equipment installation, inspection, and maintenance.

[0057] In some embodiments, the strain pressure sensor is used to monitor the deformation of the step tread surface in real time and convert the monitoring data into pressure value distribution of the step tread surface to characterize the step load and pressure distribution state.

[0058] Unlike traditional single-point weighing methods, strain pressure sensors can acquire pressure distribution data across the entire step surface, rather than the deformation at a single monitoring point. This provides a complete description of the stress state of the steps, avoiding monitoring deviations caused by uneven local stress, and offering more accurate data support for load characteristic analysis.

[0059] It should be noted that the strain pressure sensor detects minute deformations of the step tread surface in real time, converting the deformation signal into a quantifiable electrical signal. Through built-in algorithms, this electrical signal is transformed into pressure distribution data covering the entire tread surface, rather than being limited to a single monitoring point. This provides a complete picture of the load magnitude and pressure distribution of the steps. This avoids the force assessment bias caused by traditional single-point monitoring, accurately characterizing the step load and pressure distribution features. It provides comprehensive data support for load characteristic analysis, helping subsequent fault diagnosis models accurately identify potential faults caused by uneven load distribution and improving the system's sensitivity to load-related faults.

[0060] In some embodiments, the acceleration vibration sensor is used to collect the running speed, vibration frequency, and vibration data under different loads of the steps in real time when the running speed of the steps reaches a preset value and the tread surface is facing upwards.

[0061] The preset value can be flexibly set according to the escalator's designed operating speed, such as 0.5 m / s. By limiting the sensor's operating conditions, it ensures that the collected data are vibration information from the effective operating phase of the steps, avoiding interference from invalid data and improving data quality. Simultaneously, the sensor can record vibration changes under different load conditions, laying the foundation for subsequent correlation analysis.

[0062] It should be noted that a preset threshold for the step running speed is established. The speed detection component monitors the step running speed in real time, and the position sensor determines the orientation of the tread surface. When the step running speed reaches the preset threshold and the tread surface is facing upwards, the acceleration vibration sensor automatically activates, continuously collecting step running speed and vibration frequency data, and simultaneously recording detailed data such as vibration amplitude and vibration period under different load conditions. Limiting the sensor's operating conditions ensures that the collected data represents vibration information from the effective running phase of the steps, filtering out invalid data interference and improving data quality. Comprehensive capture of vibration characteristics under different operating conditions provides rich data dimensions for subsequent vibration characteristic analysis and fault diagnosis, enhancing the system's ability to identify vibration-related faults.

[0063] In some embodiments, the edge computing module is used to receive monitoring data from the strain pressure sensor and the acceleration vibration sensor, determine the cascade working condition, and extract feature values ​​from the monitoring data to complete preliminary fault detection.

[0064] The edge computing module has built-in operating condition judgment logic, which can identify in real time whether the ladder is in different operating conditions such as manned operation or no-load operation based on sensor data. During the feature extraction process, key indicators such as peak value, mean, variance, and frequency spectrum characteristics of time series data can be extracted. By comparing with preset normal thresholds, it can quickly determine whether there are any abnormalities, realize preliminary fault screening, and reduce the data processing pressure on the remote server.

[0065] It should be noted that the edge computing module receives raw data transmitted from strain pressure sensors and acceleration vibration sensors, and determines whether the ladder is in an unloaded, passenger-carrying, or pedestrian-walking condition using preset logic rules. Feature value extraction is performed on the raw data, extracting key indicators such as peak value, mean, variance, and frequency spectrum features. These are compared with preset normal operating condition thresholds; if the threshold is exceeded, it is determined to be a preliminary suspected fault state, completing the preliminary fault detection. This invention achieves local data preprocessing and preliminary screening, reducing the amount of data transmitted to the remote server, reducing server computing pressure, and improving the overall system response speed. Early identification of obviously abnormal data provides targeted data for accurate diagnosis by the remote server, shortening the fault judgment cycle and improving the efficiency and accuracy of preliminary fault detection.

[0066] In some embodiments, the remote server employs big data processing technology to integrate and analyze the received monitoring data and feature values. The integrated analysis includes correlation analysis of load characteristics, vibration characteristics, and operating condition characteristics.

[0067] Big data processing technology can efficiently integrate massive amounts of time-series data. By mining the inherent correlation between load characteristics, vibration characteristics, and operating condition characteristics, such as the normal vibration frequency range under different loads and the deformation threshold under specific operating conditions, a multi-dimensional fault judgment model can be established, breaking through the limitations of single feature analysis in existing technologies and improving the accuracy of fault identification.

[0068] It should be noted that the remote server is equipped with a big data processing framework, which classifies, stores, and integrates the received raw monitoring data and feature values ​​extracted by the edge computing module. Through algorithms, it mines the inherent correlations between load characteristics, vibration characteristics, and operating condition characteristics, establishes multi-feature mapping relationships, analyzes the fault occurrence patterns under different feature combinations, and forms a comprehensive operational status assessment system. This invention breaks through the limitations of traditional single-feature analysis, interprets the escalator's operating status from multiple dimensions, deeply explores the potential patterns of fault occurrence, and improves the scientificity and accuracy of fault diagnosis. Big data processing technology ensures efficient processing of massive amounts of data, supports centralized management of multiple escalators and multiple levels, and provides technical support for large-scale applications.

[0069] In some embodiments, the cascade monitoring unit processes the collected strain and pressure data as follows:

[0070] If the strain pressure value is greater than the preset range, it is determined that the step has completed the entire process of the tread facing upwards.

[0071] If the strain pressure value changes significantly during operation and sometimes returns to zero, then delete that segment of strain pressure data.

[0072] If the strain pressure value changes significantly but does not return to zero, the strain pressure data will be discarded and interpolated.

[0073] If the strain pressure value suddenly jumps and then remains stable, the strain pressure data should be segmented.

[0074] The aforementioned targeted data processing rules effectively filter out interference data caused by unstable operating conditions such as personnel walking and brief stops, ensuring that all data transmitted to the remote server is valid and stable sample data, providing a high-quality data foundation for model training and fault analysis. The preset range can be calibrated based on the rated load and structural characteristics of the cascade.

[0075] It should be noted that the cascade monitoring unit performs targeted processing logic on the collected strain and pressure data: when the strain and pressure value is greater than the preset range, it is determined that the cascade has completed the entire process of the treads facing upwards, and the complete data is retained; when the data changes significantly and there are instances of zeroing out, it is determined that personnel are walking, and that segment of data is deleted; when the data changes significantly but there are no instances of zeroing out, it performs elimination and interpolation generation processing; when the data suddenly jumps and then stabilizes, it is segmented. This invention effectively filters out interference data generated by unstable working conditions such as personnel walking and short stops, ensuring that the data transmitted to the remote server is all valid and stable sample data, providing a high-quality data foundation for the training and analysis of the fault diagnosis model, and avoiding the impact of noisy data on the accuracy of fault diagnosis.

[0076] In some embodiments, the remote server compares the predicted vibration with the actual vibration using a trained model. If the difference exceeds a preset threshold, a fault alert is issued, and fault diagnosis is performed based on the entire vibration data. The fault occurrence date is predicted by combining historical difference trends.

[0077] The model is generated based on the processed effective sample set and can establish the correspondence between strain changes and acceleration vibration changes. By analyzing the difference between predicted and true values, it can accurately identify abnormal states. The preset threshold can be dynamically adjusted according to factors such as the escalator's operating years and equipment model. During the fault diagnosis process, the frequency, amplitude and other characteristics of vibration data can be combined to locate the fault type (such as chain wear, step deformation, etc.). Historical difference trend analysis can predict the fault development pattern and provide a basis for maintenance plan formulation.

[0078] It should be noted that the remote server uses a fault diagnosis model trained on a valid sample set. It calculates the difference between the predicted vibration data corresponding to the input load characteristics and operating condition characteristics and the actual vibration data collected by the acceleration vibration sensor. A preset difference threshold is set; when the calculation result exceeds the threshold, a fault alert is triggered. The fault type is determined by matching the frequency, amplitude, and other characteristics of the vibration data against a fault feature database. By analyzing the changing trends of historical difference data, regression analysis is used to predict the fault occurrence date. This invention achieves accurate fault location and type identification, providing maintenance personnel with a clear maintenance direction and reducing blind maintenance. Predicting the fault occurrence time in advance facilitates the development of reasonable maintenance plans, avoids losses due to escalator downtime caused by sudden faults, and further improves the safety and reliability of escalator operation.

[0079] In some embodiments, the wireless transmission module is used to transmit the monitoring data and feature values ​​processed by the edge computing module to the remote server when the escalator enters a low-speed mode, and to receive parameters and models sent by the remote server.

[0080] Selecting a low-speed mode for escalator data transmission (such as maintenance mode or empty return mode) can avoid data transmission delays or losses during high-speed operation, ensuring data transmission stability. At the same time, the remote server can use this module to send threshold parameters, model update files, etc., to achieve remote system upgrades and parameter calibration, improving system adaptability.

[0081] It should be noted that the wireless transmission module monitors the escalator's operating speed in real time. When it detects that the escalator has entered low-speed mode, it initiates the data transmission process, sending the monitoring data, feature values, and preliminary fault judgment results processed by the edge computing module to a remote server via the network. Simultaneously, it continuously listens for parameter commands and model update files from the remote server, automatically synchronizing them to the step monitoring unit upon receipt to complete parameter calibration and model upgrades. Selecting low-speed mode for data transmission avoids data transmission delays or loss during high-speed operation, ensuring the stability and integrity of data interaction. The system supports remote upgrades and parameter adjustments without on-site operation, reducing maintenance costs, improving the system's adaptability to different escalator models and operating environments, and extending equipment lifespan.

[0082] Reference Figure 3 Secondly, the present invention provides an escalator fault diagnosis method based on strain pressure and acceleration, applied to the aforementioned escalator fault diagnosis system based on strain pressure and acceleration, comprising the following steps:

[0083] S1. The cascade monitoring unit collects strain and pressure data and acceleration and vibration data of the cascade to determine the working condition of the cascade and perform preliminary fault detection.

[0084] S2. When the escalator enters low-speed mode, the step monitoring unit transmits the processed monitoring data and feature values ​​to the remote server.

[0085] S3. The remote server integrates and analyzes the received data, and performs correlation analysis by combining load characteristics, vibration characteristics and working condition characteristics.

[0086] S4. The remote server compares the predicted vibration with the actual vibration using a model. If the difference exceeds a threshold, it issues a fault alert and diagnoses the fault, and combines historical data to predict the time of the fault occurrence.

[0087] The specific implementation steps are as follows:

[0088] The first step involves the sensors in the step monitoring unit collecting strain, pressure, acceleration, and vibration data. The edge computing module then determines the operating conditions and extracts feature values ​​to complete preliminary fault detection. The second step involves the step monitoring unit uploading the processed data via a wireless transmission module after the escalator enters low-speed mode. The third step involves a remote server integrating the data and analyzing the correlation between load, vibration, and operating condition characteristics. The fourth step involves comparing the predicted data with the actual vibration data using a model. If the difference exceeds a threshold, a fault alert is issued, the fault type is diagnosed, and the timing of the fault is predicted based on historical data.

[0089] This invention establishes a standardized process for data acquisition, processing, analysis, and diagnosis, achieving fully automated escalator fault diagnosis without manual intervention, thus improving the real-time performance and efficiency of fault handling. The process design closely matches the actual operating conditions of escalators, with seamless integration of data transmission and analysis, ensuring the consistency and accuracy of fault diagnosis and providing a clear and operable technical path for escalator maintenance.

[0090] To better understand the technical solution of the present invention, the technical solution of the present invention will be described in detail below:

[0091] refer to Figure 1 The figure is a schematic diagram of the structure of an escalator fault diagnosis system based on strain pressure and acceleration according to some embodiments of the present invention. The system mainly includes multiple step monitoring units and a remote server. Each step monitoring unit is installed at the bottom of the tread surface of one step of the escalator. The remote server establishes a communication connection with all step monitoring units through networks such as the Internet and the Internet of Things to realize data interaction and command issuance.

[0092] refer to Figure 2 The figure is a schematic diagram of the internal structure of the step monitoring unit according to some embodiments of the present invention. The step monitoring unit includes a strain pressure sensor, an acceleration vibration sensor, a wireless transmission module and an edge computing module. The strain pressure sensor, the acceleration vibration sensor and the wireless transmission module are all electrically connected to the edge computing module through wires. The edge computing module adopts a low-power microprocessor to ensure stable operation during the long-term operation of the escalator.

[0093] In practice, the strain pressure sensors are arranged in a distributed manner and evenly installed in the stress area at the bottom of the step tread surface. Resistance strain gauge sensors can be selected, which have high sensitivity and fast response speed. They can capture the small deformation of the tread surface in real time and convert the deformation signal into an electrical signal and transmit it to the edge computing module. The edge computing module processes the electrical signal and converts it into pressure distribution data of the step tread surface, which fully characterizes the load size and pressure distribution state of the step.

[0094] The acceleration vibration sensor is a triaxial accelerometer, installed at the center of the bottom of the step tread. Its measurement range can be set according to the escalator's operating characteristics (e.g., ±10g). When the step speed reaches the preset value (e.g., 0.5m / s) and the position sensor detects that the tread is running upwards, the edge computing module triggers the acceleration vibration sensor to start working, collecting the step speed, vibration frequency, and vibration acceleration data under different loads in real time, and transmitting the data to the edge computing module in real time.

[0095] After receiving monitoring data from strain and pressure sensors and acceleration and vibration sensors, the edge computing module first performs a condition assessment: based on the variation pattern of strain and pressure data and the stability of acceleration data, it identifies whether the ladder is in no-load operation, manned operation, or personnel walking operation; then, it extracts feature values ​​from the monitoring data, extracting key feature indicators such as peak value, mean, variance, kurtosis, and frequency spectrum peak value of the time series data, and compares the extracted feature values ​​with preset normal operating condition thresholds. If the feature values ​​exceed the normal range, it is determined to be a preliminary suspected fault state, the data segment is marked, and further analysis is awaited.

[0096] Meanwhile, the edge computing module cleans the strain and pressure data according to preset data processing rules:

[0097] If the strain pressure value is greater than the preset range (this range is set according to 1.2 times the rated load of the step, for example, when the rated load is 200kg, the preset range is 0-240N), then it is determined that the step has completed the entire process of running with the tread facing upwards, and the complete data of this segment is retained;

[0098] If the strain pressure value changes significantly during operation and sometimes drops to zero (i.e., the pressure value drops to zero), it is determined that someone is walking on the steps. This data segment has poor stability and is therefore deleted.

[0099] If the strain pressure value changes significantly but does not return to zero, it is determined that someone has passed through the step quickly. Abnormal fluctuations in the data are removed, and linear interpolation is used to supplement the missing data to ensure the continuity of the data.

[0100] If the strain pressure value suddenly jumps and then remains stable (the jump amplitude exceeds the preset value, such as 50N, and the stabilization time exceeds 3 seconds), it is determined that someone is standing on the step. The data segment is then processed according to the jump time point, and the valid data before and after the standing are retained respectively.

[0101] When the escalator enters a low-speed mode (e.g., the operating speed is below 0.2m / s), the edge computing module controls the wireless transmission module to start data transmission, transmitting the processed monitoring raw data, extracted feature values, and preliminary fault judgment results to the remote server via 4G / 5G or Wi-Fi network; at the same time, the wireless transmission module receives parameters (such as strain pressure preset range, fault difference threshold, etc.) and model update files sent by the remote server, realizing parameter calibration and function upgrade of the escalator monitoring unit.

[0102] The remote server is equipped with a high-performance processor and large-capacity storage device, and uses big data processing frameworks such as Hadoop and Spark to integrate, store, and analyze the received multi-level and multi-time period data. First, the data transmitted by each monitoring unit is classified and organized to establish a database indexed by level number and timestamp. Then, correlation analysis is performed on load characteristics (pressure distribution, load size), vibration characteristics (vibration frequency, acceleration amplitude), and operating condition characteristics (empty load, manned load) to explore the mapping relationship between different characteristics, such as establishing a comparison table of normal vibration frequencies under different loads.

[0103] A fault diagnosis model is pre-trained on the remote server. This model is generated based on a historical valid sample set, which includes multi-dimensional feature data under normal operating conditions and feature data under known fault conditions. The model uses machine learning algorithms (such as random forest and support vector machine) to predict the corresponding vibration characteristics based on the input load characteristics and operating conditions. The difference between the real-time collected vibration characteristics and the vibration characteristics predicted by the model is calculated. If the difference exceeds a preset threshold (this threshold is set according to the escalator equipment model and operating years, for example, 5%), the remote server issues a fault alert signal (which can notify management personnel via SMS, platform push, etc.). At the same time, based on the frequency spectrum and time domain characteristics of the entire vibration data, combined with the fault feature database, the fault type is diagnosed (such as chain wear corresponding to vibration anomalies in a specific frequency range, and step deformation corresponding to uneven pressure distribution and increased vibration amplitude). Finally, by analyzing the historical difference trend (such as the curve of difference value changing over time), regression analysis is used to predict the specific date of the fault, providing maintenance personnel with a clear maintenance time node.

[0104] refer to Figure 3 The figure is an exemplary flowchart of an escalator fault diagnosis method based on strain pressure and acceleration according to some embodiments of the present invention, the method comprising the following steps:

[0105] S1. The strain and pressure sensor and the acceleration and vibration sensor in the cascade monitoring unit collect strain and pressure data and acceleration and vibration data of the cascade, respectively. After receiving the data, the edge computing module judges the working condition of the cascade and extracts the data feature values ​​and compares them with the normal threshold to complete the preliminary fault detection.

[0106] S2. When the escalator's operating speed drops to the low-speed mode threshold, the wireless transmission module transmits the monitoring data, feature values, and preliminary fault judgment results processed by the edge computing module to the remote server.

[0107] S3. The remote server classifies and integrates the received data, and uses big data processing technology to analyze the correlation between load characteristics, vibration characteristics and operating condition characteristics.

[0108] S4. The remote server compares the predicted vibration characteristics with the actual vibration characteristics using the fault diagnosis model and calculates the difference value. If the difference value does not exceed the preset threshold, the escalator is determined to be operating normally, and the historical database is updated. If the difference value exceeds the preset threshold, a fault reminder is issued, the fault type is diagnosed in combination with the vibration data, and the fault occurrence time is predicted based on the historical difference trend. Maintenance suggestions are generated and pushed to the management personnel terminal.

[0109] This invention, through the collaborative work of the step monitoring unit and the remote server, achieves comprehensive monitoring, accurate diagnosis, and early prediction of escalator faults, effectively addressing the shortcomings of existing technologies and providing reliable technical support for the safe operation and efficient maintenance of escalators.

[0110] The technical solutions provided by the embodiments of the present invention have the following beneficial effects:

[0111] This system integrates strain pressure sensors and acceleration vibration sensors at the bottom of the step tread surface, combined with edge computing modules and remote servers, to achieve full-dimensional monitoring and fault early warning diagnosis of steps and step chains. It breaks through the limitation of traditional escalators that can only achieve a single weighing function, can detect potential safety hazards in time, and significantly improve the safety of escalator operation.

[0112] Strain pressure sensors monitor the deformation and pressure distribution of the tread surface in real time, while acceleration vibration sensors capture operating status parameters. The combination of the two enables comprehensive collection of multi-dimensional characteristics of escalator load, vibration, and operating conditions, providing complete data support for escalator performance evaluation and improving the scientific nature and efficiency of maintenance decisions.

[0113] By using a wireless transmission module to establish remote communication with a remote server, the absence of on-site measurement by staff simplifies the structure of the detection device, reduces labor costs and maintenance difficulty, and enables centralized monitoring of multiple steps and escalators, thereby improving operational management efficiency.

[0114] By integrating load characteristics, vibration characteristics, and operating condition characteristics for correlation analysis, and combining big data processing technology and intelligent models, a deep understanding of the escalator's operating status is achieved. Compared with existing single-feature analysis methods, the accuracy and timeliness of fault diagnosis are significantly improved.

[0115] The system has real-time monitoring, preliminary fault detection, remote accurate diagnosis and fault prediction functions, forming a full-chain fault handling mechanism of early warning-diagnosis-prediction. It can predict the occurrence time and development trend of faults in advance, making it easier to formulate maintenance plans in advance, reduce downtime losses caused by sudden faults, and further improve the reliability and safety of escalator operation.

[0116] The technical solutions provided by the embodiments of the present invention have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of the embodiments of the present invention. The descriptions of the embodiments above are only for helping to understand the principles of the embodiments of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the embodiments of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. An escalator fault diagnosis system based on strain pressure and acceleration, characterized in that, include: Remote server and multiple tiered monitoring units; The step monitoring unit is installed at the bottom of the step tread surface of the escalator to collect strain pressure data and acceleration vibration data of the steps, and to perform working condition judgment and preliminary fault detection. The remote server communicates with each step monitoring unit via a network to receive data transmitted by the step monitoring units, and performs comprehensive analysis by combining load characteristics, vibration characteristics and operating condition characteristics to realize escalator fault early warning, diagnosis and fault occurrence time prediction.

2. The escalator fault diagnosis system based on strain pressure and acceleration according to claim 1, characterized in that, The cascade monitoring unit includes a strain pressure sensor, an acceleration vibration sensor, a wireless transmission module, and an edge computing module; the strain pressure sensor, the acceleration vibration sensor, and the wireless transmission module are all electrically connected to the edge computing module.

3. The escalator fault diagnosis system based on strain pressure and acceleration according to claim 2, characterized in that, The strain pressure sensor is used to monitor the deformation of the step tread surface in real time, and converts the monitoring data into pressure value distribution of the step tread surface to characterize the step load and pressure distribution state.

4. The escalator fault diagnosis system based on strain pressure and acceleration according to claim 2, characterized in that, The acceleration vibration sensor is used to collect the running speed, vibration frequency, and vibration data under different loads of the steps in real time when the running speed of the steps reaches a preset value and the tread surface is facing upwards.

5. The escalator fault diagnosis system based on strain pressure and acceleration according to claim 2, characterized in that, The edge computing module is used to receive monitoring data from the strain pressure sensor and the acceleration vibration sensor, determine the cascade working condition, and extract feature values ​​from the monitoring data to complete preliminary fault detection.

6. The escalator fault diagnosis system based on strain pressure and acceleration according to claim 1, characterized in that, The remote server uses big data processing technology to integrate and analyze the received monitoring data and feature values. The integration and analysis includes correlation analysis of load characteristics, vibration characteristics and operating condition characteristics.

7. The escalator fault diagnosis system based on strain pressure and acceleration according to claim 1, characterized in that, The cascade monitoring unit processes the collected strain and pressure data as follows: If the strain pressure value is greater than the preset range, it is determined that the step has completed the entire process of the tread facing upwards. If the strain pressure value changes significantly during operation and sometimes returns to zero, then delete that segment of strain pressure data. If the strain pressure value changes significantly but does not return to zero, the strain pressure data will be discarded and interpolated. If the strain pressure value suddenly jumps and then remains stable, the strain pressure data should be segmented.

8. The escalator fault diagnosis system based on strain pressure and acceleration according to claim 1, characterized in that, The remote server compares the predicted vibration with the actual vibration using the trained model. If the difference exceeds a preset threshold, it issues a fault alert and performs fault diagnosis based on the entire vibration data. It also predicts the fault occurrence date by combining historical difference trends.

9. The escalator fault diagnosis system based on strain pressure and acceleration according to claim 2, characterized in that, The wireless transmission module is used to transmit the monitoring data and feature values ​​processed by the edge computing module to the remote server when the escalator enters low-speed mode, and to receive the parameters and models sent by the remote server.

10. A method for escalator fault diagnosis based on strain pressure and acceleration, characterized in that, The escalator fault diagnosis system based on strain pressure and acceleration as described in any one of claims 1-9 includes the following steps: S1. The cascade monitoring unit collects strain and pressure data and acceleration and vibration data of the cascade to determine the working condition of the cascade and perform preliminary fault detection. S2. When the escalator enters low-speed mode, the step monitoring unit transmits the processed monitoring data and feature values ​​to the remote server. S3. The remote server integrates and analyzes the received data, and performs correlation analysis by combining load characteristics, vibration characteristics and working condition characteristics. S4. The remote server compares the predicted vibration with the actual vibration using a model. If the difference exceeds a threshold, it issues a fault alert and diagnoses the fault, and combines historical data to predict the time of the fault occurrence.

Images