Diagonal elevator fault detection method based on real-time analysis of elevator operation data

Abstract

The application relates to the technical field of elevator operation monitoring, in particular to a fault detection method for inclined elevators based on real-time analysis of elevator operation data, which comprises the following steps: collecting various monitoring data in the operation process of an inclined elevator in real time; obtaining the cumulative change degree and the transient deviation degree of the various monitoring data at each collection time; classifying all collection times before the current time; obtaining the first characteristic value of the instantaneous response state of each class at the current time, and the second characteristic value of the various monitoring data in the instantaneous response state of each class at the current time; obtaining the cumulative deviation response characteristic value of the various monitoring data at the current time; adjusting the time step when a long short-term memory network algorithm is used to perform time series prediction on the various monitoring data, so as to obtain the prediction result of the various monitoring data; and detecting the fault condition of the inclined elevator. The application aims to improve the accuracy and timeliness of fault detection.

Description

Technical Field

[0001] This application relates to the field of elevator operation monitoring technology, specifically to a fault detection method for inclined elevators based on real-time analysis of elevator operation data. Background Technology

[0002] Inclined escalators, as a special type of vertical transportation equipment, are widely used in complex scenarios such as scenic spots, mountainous buildings, and large commercial complexes due to their non-traditional vertical or horizontal operating trajectory. However, their unique operating trajectory makes their operating environment and mechanical load more complex, making timely and accurate fault detection a crucial link in ensuring passenger safety, reducing maintenance costs, and extending equipment lifespan. Effective fault detection can not only provide early warning of potential risks and avoid major safety accidents, but also reduce downtime caused by sudden equipment failures, significantly improving overall operational efficiency.

[0003] Currently, fault detection for inclined elevators based on real-time operational data faces numerous challenges. Traditional fault detection methods struggle to capture subtle anomalies during elevator operation in real time, exhibiting significant detection lag and failing to meet the need for timely fault detection and handling. Although existing automated detection technologies utilize sensors to collect operational data, the significant differences in operational characteristics between inclined and ordinary elevators, particularly the lack of sufficient consideration of the impact of the tilt angle on operational status, make it difficult to accurately analyze the instantaneous changes caused by cable and traction rope sagging. This results in substantial deviations in operational status detection, leading to delayed fault detection and posing a potential risk of safety accidents. Summary of the Invention

[0004] In view of the above, it is necessary to provide a fault detection method for inclined elevators based on real-time analysis of elevator operation data, which improves the accuracy and timeliness of fault detection compared with traditional fault detection methods for inclined elevators.

[0005] The fault detection method for inclined elevators based on real-time analysis of elevator operation data in this application adopts the following technical solution:

[0006] One embodiment of this application provides a fault detection method for inclined elevators based on real-time analysis of elevator operation data. The method includes the following steps:

[0007] Real-time collection of various monitoring data during the operation of the inclined elevator;

[0008] By analyzing the distribution of various monitoring data collected at each acquisition time and before, the cumulative change of various monitoring data at each acquisition time is obtained. By combining the cumulative change with the degree to which various monitoring data at each acquisition time deviates from its preset safety threshold, the transient deviation of various monitoring data at each acquisition time is obtained. By analyzing the similarity of the transient deviation of various monitoring data between different acquisition times before the current time, all acquisition times before the current time are classified.

[0009] By taking the time intervals of all adjacent acquisition times in each category, a first characteristic value of the instantaneous response state of each category at the current moment is obtained; by taking the average level of the transient deviation of various monitoring data at all acquisition times in each category, a second characteristic value of the instantaneous response state of various monitoring data in each category at the current moment is obtained; by taking the first characteristic value and the second characteristic value, a cumulative deviation response characteristic value of various monitoring data at the current moment is obtained; by taking the cumulative deviation response characteristic value, the time step of the Long Short-Term Memory Network algorithm for time-series prediction of various monitoring data is adjusted to obtain the prediction results of various monitoring data;

[0010] The malfunction of the inclined elevator is detected by comparing the predicted results with the preset safety range of various monitoring data.

[0011] In one embodiment, the monitoring data includes: velocity, acceleration, vibration frequency, and tilt angle data.

[0012] In one embodiment, the cumulative degree of change is the normalized value of the dispersion of various monitoring data collected at each acquisition time and prior to that time.

[0013] In one embodiment, the method for obtaining the transient deviation is as follows:

[0014] Calculate the difference between various monitoring data at each acquisition time and their preset safety threshold;

[0015] The transient deviation is the product of the difference value corresponding to various monitoring data at each acquisition time and the cumulative change.

[0016] In one embodiment, the process of classifying all acquisition times prior to the current time is as follows:

[0017] The transient deviations of all types of monitoring data at each acquisition time are combined to form a transient deviation sequence for each acquisition time;

[0018] By using the similarity between the transient deviation sequences at different acquisition times, and combining it with a clustering algorithm, we can obtain the categories of all acquisition times before the current time.

[0019] In one embodiment, the first feature value is the average of the time intervals between all adjacent acquisition times in each class.

[0020] In one embodiment, the second characteristic value is the mean of the transient deviation of various monitoring data at all acquisition times in each category.

[0021] In one embodiment, the process of obtaining the cumulative deviation response characteristic value is as follows:

[0022] Calculate the sum of the first feature values ​​of all classes at the current time, and calculate the ratio of the first feature value to the sum;

[0023] Calculate the product of the second characteristic value and the ratio of various monitoring data at the current moment for each category;

[0024] The cumulative deviation response characteristic value is the mean of the product of various monitoring data in all classes at the current time.

[0025] In one embodiment, the method for adjusting the time step is as follows:

[0026] Obtain the result of calculating the difference between 1 and the normalized value of the cumulative deviation response characteristic value;

[0027] The product of the preset upper limit of the calculation time step and the result of the difference calculation is calculated.

[0028] The sum of the product value and the preset initial value of the time step is used as the adjusted time step when using the Long Short-Term Memory network algorithm to perform time-series prediction on various monitoring data.

[0029] In one embodiment, the method for detecting malfunctions in the inclined elevator is as follows:

[0030] If the prediction result of any monitoring data exceeds its preset safety range, the inclined elevator is determined to be faulty; otherwise, the inclined elevator is determined not to be faulty.

[0031] This application has at least the following beneficial effects:

[0032] Addressing the issue that inclined elevators, commonly used in complex installation environments such as outdoor scenic spots and mountains, are prone to cable and traction rope sagging due to gravity caused by their tilt angle. This leads to changes in the elevator car's tilt angle and fluctuations in monitoring data. Furthermore, the differences in instantaneous characteristics at different data collection times result in significant variations in cumulative deviation response characteristics, causing low accuracy in traditional fault detection methods. This application first calculates the cumulative variability, reflecting the state characteristics of various monitoring data during elevator operation up to each data collection time, helping to identify data fluctuation trends. Combining the state characteristics of various monitoring data, the transient deviation is obtained, enabling more accurate capture of the instantaneous characteristics of the monitoring data and providing key features for subsequent fault detection. Through classification... The algorithm calculates a first and a second eigenvalue to quantify the frequency of anomalies and the degree of deviation in various monitoring data during the operation of the inclined elevator. Using these eigenvalues, a cumulative deviation response eigenvalue is calculated, reflecting the cumulative deviation impact of various monitoring data on the inclined elevator's operating status up to the current moment. The time step of the Long Short-Term Memory (LSTM) network algorithm for time-series prediction is dynamically adjusted based on the cumulative deviation response eigenvalue. This adaptive adjustment mechanism automatically focuses on high-risk periods based on real-time deviation characteristics, achieving adaptive predictive analysis of "coarse-grained prediction during stable periods and fine-grained analysis during abnormal periods." This allows for more accurate prediction of the inclined elevator's operating status, improving the accuracy and timeliness of fault detection. Its beneficial effect lies in fully considering the cumulative response characteristics of the deviation caused by the sagging of cables and traction ropes due to the inclined elevator's inclined characteristics, thereby achieving accurate analysis and detection of inclined elevator operating faults and improving the safety of inclined elevator operation. Attached Figure Description

[0033] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 A flowchart illustrating the steps of the inclined elevator fault detection method based on real-time analysis of elevator operation data provided in this application;

[0035] Figure 2 This is a schematic diagram illustrating the process of obtaining the cumulative deviation response characteristic value. Detailed Implementation

[0036] In the description of the embodiments in this application, the words "exemplary," "or," and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary," "or," and "for example" is intended to present the relevant concepts in a specific manner.

[0037] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. It should be understood that, unless otherwise stated, " / " in this application means "or".

[0038] It should also be noted that the terms "first" and "second" in this application are used to distinguish similar objects, rather than to describe a specific order or sequence.

[0039] The following description, in conjunction with the accompanying drawings, details the specific scheme of the inclined elevator fault detection method based on real-time analysis of elevator operation data provided in this application.

[0040] This application provides an embodiment of a fault detection method for inclined elevators based on real-time analysis of elevator operation data. Specifically, the following method for fault detection of inclined elevators based on real-time analysis of elevator operation data is provided. Please refer to [link / reference]. Figure 1 The method includes the following steps:

[0041] Step 1: Collect various monitoring data and location information in real time during the operation of the inclined elevator.

[0042] In this application, in order to achieve comprehensive monitoring of the operating status of the traction inclined elevator, multiple types of sensors are integrated in the elevator car to collect various monitoring data during the operation of the inclined elevator in real time. Specifically, the monitoring data includes: speed, acceleration, vibration frequency and tilt angle data, which are collected by speed sensor, acceleration sensor, vibration sensor and angle sensor respectively. In addition, the position of the inclined elevator is collected in real time using a position sensor.

[0043] Considering that inclined elevators are often used in complex environments such as outdoor scenic spots and mountains, where the safety requirements for elevator operation are extremely high, it is necessary to ensure that the collected monitoring data has high reliability and low noise characteristics to support subsequent accurate fault characteristic analysis. Therefore, a Kalman filter is used to denoise the raw data collected by various types of sensors. A state-space model established by the Kalman filter is used to optimally estimate the data sequence, effectively suppressing interference from environmental noise and sensor errors. The Kalman filter is merely one embodiment of this application. Based on the ability to denoise the raw data collected by various types of sensors, implementers can use other existing feasible technologies; this application does not impose any special limitations.

[0044] Step 2: Analyze various monitoring data during the operation of the inclined elevator, obtain the cumulative deviation response characteristic value of various monitoring data at the current moment, and use it to adjust the time step when using the Long Short-Term Memory Network algorithm to make time-series predictions of various monitoring data, so as to obtain the prediction results of various monitoring data.

[0045] Due to the complexity of the installation scenarios and the high safety requirements of inclined elevators, precise monitoring and analysis of their operation are necessary to improve the accuracy of fault detection. Considering the inherent tilt angle of inclined elevators, gravity can cause cables and traction ropes to sag during operation, leading to changes in the elevator car's tilt angle. Changes in the elevator car's angle can further cause wear on the guide rails and traction sheaves, affecting the elevator's operational safety.

[0046] Step 2.1: Obtain the cumulative change of various monitoring data at each acquisition time by analyzing the distribution of various monitoring data acquired at each acquisition time and before; obtain the transient deviation of various monitoring data at each acquisition time by combining the cumulative change with the degree to which various monitoring data at each acquisition time deviates from its preset safety threshold; classify all acquisition times before the current time by analyzing the similarity of the transient deviation of various monitoring data between different acquisition times before the current time.

[0047] Based on the above analysis, during the operation of an inclined elevator, when fault detection is performed using real-time collected monitoring data, due to the elevator's tilt angle, there are differences in the instantaneous response characteristics of different monitoring data when an operational anomaly occurs. That is, different monitoring data will exhibit abnormal changes sequentially before and after an elevator malfunction. Therefore, the dispersion of various monitoring data collected at each acquisition time and all preceding acquisition times is calculated. This dispersion reflects the state characteristics of various monitoring data during elevator operation up to each acquisition time. The normalized value of this dispersion is then used as the cumulative change rate of various monitoring data at each acquisition time.

[0048] In this embodiment, the dispersion is the coefficient of variation. The calculation of the coefficient of variation is a well-known technique and will not be described in detail here. As other implementation methods, based on the ability to measure the uneven distribution of various monitoring data, implementers may use other existing techniques, such as variance, standard deviation, etc. This application does not impose any special restrictions.

[0049] In this embodiment, the Softmax function is used to obtain the normalized value of the discreteness. The Softmax function is a well-known technique and will not be described in detail in this application.

[0050] Furthermore, the difference between various monitoring data at each acquisition time and their preset safety threshold is calculated. The product of the difference value and the cumulative degree of change for each monitoring data at each acquisition time is used as the transient deviation of the various monitoring data at each acquisition time. The reason for this operation is that, for the same type of monitoring data, the more significant its state characteristics up to each acquisition time, the greater the deviation impact on the monitoring data at each acquisition time, and the more significant the transient characteristics of various monitoring data. To comprehensively analyze the differences in transient characteristics at different acquisition times, all acquisition times before the current time are classified based on the similarity of the transient deviation of various monitoring data between different acquisition times before the current time. Specifically:

[0051] The transient deviations of all types of monitoring data at each acquisition time are used to form a transient deviation sequence for each acquisition time. By combining the similarity between the transient deviation sequences of different acquisition times before the current time with a clustering algorithm, the categories of all acquisition times before the current time are obtained.

[0052] In this embodiment, the difference between the monitoring data and its preset safety threshold is the absolute value of the difference.

[0053] In this embodiment, the safety thresholds for velocity, acceleration, tilt angle, and vibration frequency are respectively... , , The safety thresholds for speed, acceleration, tilt angle, and vibration frequency are preset by humans and can be set by the implementer according to the actual application scenario, mechanical characteristics, and industry standards. This application does not impose any special restrictions.

[0054] In this embodiment, the clustering algorithm is a cohesive hierarchical clustering algorithm, wherein the Manhattan distance between transient deviation sequences is used as the metric distance in the cohesive hierarchical clustering algorithm. The cohesive hierarchical clustering algorithm is a well-known technology and will not be described in detail in this application. As other implementation methods, implementers may adopt other existing feasible technologies, such as the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm, etc. This application does not impose any special restrictions.

[0055] Step 2.2: Obtain the first feature value of the instantaneous response state of each category at the current moment by using the time interval of all adjacent acquisition times in each category; obtain the second feature value of the instantaneous response state of each monitoring data in each category at the current moment by using the average level of the transient deviation of each monitoring data at all acquisition times in each category; obtain the cumulative deviation response feature value of each monitoring data at the current moment by using the first feature value and the second feature value.

[0056] Based on the classification results, the sensitivity characteristics of abnormal responses of different types of monitoring data up to the current moment during the operation of the inclined elevator are analyzed. Specifically, the average time interval of all adjacent collection moments in each category is used as the first characteristic value of the instantaneous response state of each category at the current moment. The smaller the first characteristic value, the shorter the time interval of abnormal responses in the inclined elevator operation and the higher the frequency of abnormalities.

[0057] Furthermore, based on the analysis of the transient response persistence characteristics during the operation of the inclined elevator, the mean of the transient deviation of various monitoring data at all collection times in each category is taken as the second characteristic value of the instantaneous response state of various monitoring data in each category at the current moment. The larger the second characteristic value, the more significant the deviation characteristics of various monitoring data during the operation of the inclined elevator are in the current instantaneous response state, that is, the greater the possibility that the deviation of various monitoring data is caused by the fault.

[0058] Furthermore, considering the instantaneous response states of deviations occurring at different acquisition times during the operation of the inclined elevator, the cumulative deviation characteristics of various monitoring data at the current moment are analyzed. Specifically, the cumulative deviation response characteristic value of various monitoring data at the current moment is obtained through the first characteristic value and the second characteristic value, expressed as:

[0059] In the formula, This represents the cumulative deviation response characteristic value of the x-th monitoring data at the current time; n represents the number of classes. The second feature value represents the instantaneous response state of the x-th monitoring data in the ith category at the current moment; B represents the first feature value of the instantaneous response state of class i at the current moment; B represents the sum of the first feature values ​​of the instantaneous response states of all classes at the current moment.

[0060] It should be noted that the larger the calculated cumulative deviation response characteristic value, the greater the likelihood that the various monitoring data up to the current moment will have a cumulative deviation impact on the inclined elevator's operating state, considering the instantaneous state characteristics observed at different acquisition times during the actual operation of the inclined elevator. This, in turn, will have a greater impact on the accuracy of fault detection in the actual operation of the inclined elevator. A schematic diagram of the process for obtaining the cumulative deviation response characteristic value is shown below. Figure 2 As shown.

[0061] Step 2.3: Using the accumulated deviation response characteristic value, adjust the time step when using the Long Short-Term Memory Network algorithm to perform time-series prediction on various monitoring data, and obtain the prediction results for various monitoring data.

[0062] Due to the inclined operation characteristics of inclined elevators, dynamic changes such as cable and traction rope sagging and mechanical component wear may occur under gravity, leading to instantaneous differences in various monitoring data. If the impact of such instantaneous changes is prolonged and the deviation response amplitude is significant, it indicates that the cumulative deviation characteristics of the monitoring data in response to changes in operating status are more prominent, requiring improved temporal resolution in predictive analysis to capture subtle anomalies. Therefore, this application employs the Long Short-Term Memory Networks (LSTM) algorithm for time-series prediction of various monitoring data and proposes an adaptive optimization mechanism based on the time step size of the cumulative deviation response characteristic value, specifically addressing the nonlinear dynamic characteristics of inclined elevators. The specific implementation process is as follows:

[0063] In real-time analysis, the time step for time-series prediction of various monitoring data is adjusted using the cumulative deviation response characteristic value of various monitoring data at the current moment. The expression is as follows:

[0064] In the formula, This represents the adjusted time step when using the Long Short-Term Memory network algorithm to perform time-series prediction on the xth type of monitoring data; This represents the preset initial value of the time step. The value is preset by a person, and the implementer can set it according to the actual situation. In this embodiment, The value of is 5; δ represents the preset upper limit of the time step. In order to avoid noise interference caused by an excessively large time step, the value of δ in this embodiment is 30. This represents the normalized value of the cumulative deviation response characteristic value of the x-th monitoring data at the current time.

[0065] In this embodiment, the Min-Max normalization method is used. The Min-Max normalization method is a well-known technique and will not be described in detail here.

[0066] Based on the adjusted time step calculated above, various monitoring data up to the current moment and their corresponding time steps are used as input. The LSTM algorithm is used to obtain the prediction results of various monitoring data, enabling the LSTM algorithm to automatically focus on high-risk periods based on real-time deviation characteristics. This achieves adaptive prediction analysis of "coarse-grained prediction during stable periods and fine-grained analysis during abnormal periods," thereby obtaining accurate analysis results of the inclined elevator's operating status.

[0067] Step 3: Detect the malfunction of the inclined elevator by comparing the predicted results with the preset safety range of various monitoring data.

[0068] Based on the prediction results of the various monitoring data obtained above, anomaly judgment is made on the monitoring data during elevator operation. If the speed, acceleration, angle, and vibration frequency are abnormal, it indicates that the inclined elevator may be faulty and needs to be repaired in time. Specifically, when the prediction result of any monitoring data exceeds its preset safety range, it is determined that the inclined elevator is faulty; otherwise, it is determined that the inclined elevator is not faulty.

[0069] In this embodiment, the inclined elevator is deemed to have malfunctioned if at least one of the following four conditions occurs: the speed exceeds 15% of its safety threshold, the acceleration deviates from the safety threshold by ±1.5m / s², the angle deviates from the safety threshold by more than ±5%, and the vibration frequency exceeds the safety threshold.

[0070] When a fault is detected in the inclined elevator, an alarm is generated that includes the fault type, real-time location, deviation range, and timestamp, notifying the elevator's maintenance personnel to carry out timely repairs.

[0071] In summary, considering that inclined elevators are commonly used in complex installation environments such as outdoor scenic spots and mountains, the gravity caused by their tilt angle can easily lead to sagging of cables and traction ropes, resulting in changes in the elevator car's tilt angle and fluctuations in monitoring data. Furthermore, the differences in instantaneous characteristics at different data collection times lead to significant variations in cumulative deviation response characteristics, causing low accuracy in traditional fault detection methods. This application first calculates the cumulative variability, reflecting the state characteristics of various monitoring data during elevator operation up to each data collection time, helping to identify data fluctuation trends. Combining the state characteristics of various monitoring data, the transient deviation is obtained, which can more accurately capture the instantaneous characteristics of the monitoring data, providing key features for subsequent fault detection. By classifying and calculating the first and second eigenvalues, the frequency of anomalies and the degree of deviation of various monitoring data during the operation of the inclined elevator can be quantified. Using the first and second eigenvalues, a cumulative deviation response eigenvalue is calculated, reflecting the cumulative deviation impact of various monitoring data on the inclined elevator's operating status up to the current moment. The time step of the Long Short-Term Memory (LSTM) network algorithm for time-series prediction is dynamically adjusted based on the cumulative deviation response eigenvalue. This adaptive adjustment mechanism automatically focuses on high-risk periods based on real-time deviation characteristics, achieving adaptive predictive analysis of "coarse-grained prediction during stable periods and fine-grained analysis during abnormal periods." This allows for more accurate prediction of the inclined elevator's operating status, improving the accuracy and timeliness of fault detection. Its beneficial effect lies in fully considering the cumulative response characteristics of the deviation caused by the sagging of cables and traction ropes due to the inclined elevator's characteristics, thereby achieving accurate analysis and detection of inclined elevator operating faults and improving the safety of inclined elevator operation.

[0072] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0073] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from its essential characteristics. Therefore, the embodiments described above should be considered exemplary and non-limiting in all respects.

Claims

1. A fault detection method for inclined elevators based on real-time analysis of elevator operation data, characterized in that, The method includes the following steps: Real-time collection of various monitoring data during the operation of the inclined elevator; By analyzing the distribution of various monitoring data collected at each acquisition time and before, the cumulative change of various monitoring data at each acquisition time is obtained. By combining the cumulative change with the degree to which various monitoring data at each acquisition time deviates from its preset safety threshold, the transient deviation of various monitoring data at each acquisition time is obtained. By analyzing the similarity of the transient deviation of various monitoring data between different acquisition times before the current time, all acquisition times before the current time are classified. The first characteristic value of the instantaneous response state of each category at the current moment is obtained by taking the time interval of all adjacent acquisition times in each category; the second characteristic value of the instantaneous response state of each monitoring data in each category at the current moment is obtained by taking the average level of the transient deviation of each monitoring data at all acquisition times in each category. Calculate the sum of the first eigenvalues ​​of all classes at the current time, and calculate the ratio of the first eigenvalues ​​to the sum. Calculate the product of the second characteristic value and the ratio of various monitoring data at the current moment; The cumulative deviation response characteristic value is the mean of the product of various monitoring data in all classes at the current time. The time step when using the Long Short-Term Memory Network algorithm to make time-series predictions of various monitoring data is adjusted using the cumulative deviation response characteristic value to obtain the prediction results of various monitoring data. The malfunction of the inclined elevator is detected by comparing the predicted results with the preset safety range of various monitoring data. The method for adjusting the time step is as follows: Obtain the result of the difference between 1 and the normalized value of the cumulative deviation response characteristic value; The product of the preset upper limit of the time step and the difference calculation result is calculated. The sum of the product value and the preset initial value of the time step is used as the adjusted time step when using the Long Short-Term Memory network algorithm to make time-series predictions for various monitoring data.

2. The method for fault detection of inclined elevators based on real-time analysis of elevator operation data as described in claim 1, characterized in that, The monitoring data includes: velocity, acceleration, vibration frequency, and tilt angle data.

3. The method for fault detection of inclined elevators based on real-time analysis of elevator operation data as described in claim 1, characterized in that, The cumulative degree of change is the normalized value of the dispersion of various monitoring data collected at each acquisition time and before.

4. The method for fault detection of inclined elevators based on real-time analysis of elevator operation data as described in claim 1, characterized in that, The method for obtaining the transient deviation is as follows: Calculate the difference between various monitoring data at each acquisition time and their preset safety threshold; The transient deviation is the product of the difference value corresponding to various monitoring data at each acquisition time and the cumulative change.

5. The method for fault detection of inclined elevators based on real-time analysis of elevator operation data as described in claim 1, characterized in that, The process of classifying all data collection times prior to the current time is as follows: The transient deviations of all types of monitoring data at each acquisition time are combined to form a transient deviation sequence for each acquisition time; By using the similarity between the transient deviation sequences at different acquisition times, and combining it with a clustering algorithm, we can obtain the categories of all acquisition times before the current time.

6. The method for fault detection of inclined elevators based on real-time analysis of elevator operation data as described in claim 1, characterized in that, The first feature value is the average of the time intervals between all adjacent acquisition times in each category.

7. The method for fault detection of inclined elevators based on real-time analysis of elevator operation data as described in claim 1, characterized in that, The second characteristic value is the mean of the transient deviation of various monitoring data at all acquisition times in each category.

8. The method for fault detection of inclined elevators based on real-time analysis of elevator operation data as described in claim 1, characterized in that, The method for detecting malfunctions in inclined elevators is as follows: If the prediction result of any monitoring data exceeds its preset safety range, the inclined elevator is determined to be faulty; otherwise, the inclined elevator is determined not to be faulty.

Images