Elevator operation fault detection method and system

Abstract

The present invention discloses a method and system for detecting elevator operation faults, the method comprising: collecting an operation data set during the operation of the elevator; dividing the operation data set into a plurality of operation data blocks according to data types; if there is abnormal operation data in the operation data set, using the operation data set to detect the corresponding overall fault mode; finding the operation data block corresponding to the abnormal operation data, using the operation data block to detect the corresponding local fault mode; and detecting the fault position of the elevator according to the overall fault mode and the local fault mode. By using the embodiments of the present invention, the fault position of the elevator can be accurately identified, thereby improving the timeliness and accuracy of fault detection and ensuring the safe operation of the elevator.

Description

Technical Field

[0001] The present invention belongs to the technical field of elevators, and in particular to an elevator operation fault detection method and system. Background Art

[0002] With the acceleration of urbanization, elevators, as an indispensable vertical transportation tool in modern buildings, are of vital importance to the safety of life and property of the public. Elevator failures may not only cause passengers to be trapped, but may also cause more serious safety accidents. Therefore, real-time monitoring and fault detection of elevators are particularly important. Traditional elevator fault detection methods mostly rely on regular manual inspections and maintenance, and use inspection personnel to identify faults through experience and judgment, which results in low detection efficiency. Summary of the invention

[0003] The purpose of the present invention is to provide an elevator operation fault detection method and system to solve the deficiencies in the prior art, and to accurately identify the fault location of the elevator, thereby improving the timeliness and accuracy of fault detection and ensuring the safe operation of the elevator.

[0004] An embodiment of the present application provides a method for detecting an elevator operation fault, the method comprising:

[0005] Collect operation data sets during elevator operation;

[0006] Dividing the operation data set into a plurality of operation data blocks according to data types;

[0007] If there is abnormal operation data in the operation data set, using the operation data set to detect the corresponding overall failure mode;

[0008] Finding an operation data block corresponding to the abnormal operation data, and using the operation data block to detect a corresponding local fault mode;

[0009] According to the overall failure mode and the local failure mode, the failure position of the elevator is detected.

[0010] Optionally, the method further includes:

[0011] If there is abnormal operation data in the operation data set, the fault degree index of the elevator is calculated using the operation data set, and the fault position and the fault degree index are sent to the elevator control terminal.

[0012] Optionally, the method further includes:

[0013] The potential local failure modes are predicted respectively by using the operation data blocks without abnormal operation data.

[0014] Optionally, if there is abnormal operation data in the operation data set, detecting a corresponding overall failure mode using the operation data set includes:

[0015] Obtain the historical operation data set when the overall failure occurred during the historical operation of the elevator and its corresponding historical overall failure mode;

[0016] Using the historical operation data set and its corresponding historical overall failure mode, training an overall failure detection model based on deep learning;

[0017] If there is abnormal operation data in the operation data set, the operation data set is input into the overall fault detection model, and the real-time overall fault mode of the elevator in the current period is output.

[0018] Optionally, searching for an operation data block corresponding to the abnormal operation data, and using the operation data block to detect a corresponding local fault mode includes:

[0019] Obtain historical operation data blocks and historical local failure modes when multiple local failures occurred during the historical operation of the elevator;

[0020] Using the historical operation data blocks and their historical local fault modes, respectively training corresponding multiple local fault detection models based on deep learning;

[0021] The operation data block corresponding to the abnormal operation data is searched, the operation data block is input into the corresponding local fault detection model, and the real-time local fault mode of the elevator in the current period is output.

[0022] Optionally, the using of each operation data block without abnormal operation data to predict potential local failure modes respectively includes:

[0023] Obtain the historical local failure mode when multiple local failures occurred during the historical operation of the elevator and the historical operation data block of the previous period;

[0024] Using the historical local fault mode and the historical operation data blocks of the previous period, respectively train corresponding multiple local fault prediction models based on deep learning;

[0025] Each operation data block without abnormal operation data is input into the corresponding local fault prediction model, and the potential local fault mode of the elevator in the next time period is output.

[0026] Optionally, the calculation formula of the fault degree index is:

[0027] , wherein the is the fault severity index, is the severity score of the kth abnormal event, is the weighting coefficient of the kth abnormal event, is the length of the entire operating cycle, is a relevance score for the mth relevant historical failure mode, where the relevant historical failure mode is a relevant historical failure mode mapped to the current abnormal event in the historical data, is the severity impact coefficient of the mth related historical failure mode, Score the repair effectiveness of the mth related historical failure mode, is the number of abnormal events detected in the running data set, is the number of relevant historical failure modes mapped to the current abnormal event in the historical data, is the adjustment coefficient.

[0028] Another embodiment of the present application provides an elevator operation fault detection system, the system comprising:

[0029] The acquisition module is used to collect the operation data set during the operation of the elevator;

[0030] A division module, used for dividing the operation data set into a plurality of operation data blocks according to data types;

[0031] A first detection module, configured to detect a corresponding overall fault mode using the operation data set if there is abnormal operation data in the operation data set;

[0032] A second detection module, used to find an operation data block corresponding to the abnormal operation data, and use the operation data block to detect a corresponding local fault mode;

[0033] The third detection module is used to detect the fault position of the elevator according to the overall fault mode and the local fault mode.

[0034] Yet another embodiment of the present application provides a storage medium, wherein the storage medium stores a computer program, wherein the computer program is configured to execute any of the above methods when running.

[0035] Yet another embodiment of the present application provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute any of the methods described above.

[0036] Compared with the prior art, the present invention provides an elevator operation fault detection method, which collects an operation data set during the operation of the elevator; divides the operation data set into multiple operation data blocks according to data types; if there is abnormal operation data in the operation data set, uses the operation data set to detect the corresponding overall fault mode; searches for the operation data block corresponding to the abnormal operation data, and uses the operation data block to detect the corresponding local fault mode; detects the fault position of the elevator according to the overall fault mode and the local fault mode, so that the fault position of the elevator can be accurately identified, thereby improving the timeliness and accuracy of fault detection and ensuring the safe operation of the elevator. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A hardware structure block diagram of a computer terminal of an elevator operation fault detection method provided by an embodiment of the present invention;

[0038] Figure 2 A schematic diagram of a flow chart of an elevator operation fault detection method provided by an embodiment of the present invention;

[0039] Figure 3 A schematic diagram of the structure of an elevator operation fault detection system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0040] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, but should not be construed as limiting the present invention.

[0041] The embodiment of the present invention first provides an elevator operation fault detection method, which can be applied to electronic equipment, such as a computer terminal, specifically a common computer, etc.

[0042] The following describes it in detail by taking running on a computer terminal as an example. Figure 1 The hardware structure block diagram of a computer terminal of an elevator operation fault detection method provided by an embodiment of the present invention. Figure 1 As shown, the computer terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data. Optionally, the computer terminal may also include a transmission device 106 for communication functions and an input and output device 108. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above-mentioned computer terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0043] The memory 104 can be used to store software programs and modules of application software, such as program instructions / modules corresponding to the elevator operation fault detection method in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implementing the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the computer terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0044] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of a computer terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0045] See also Figure 2 The embodiment of the present invention provides an elevator operation fault detection method, which may include the following steps:

[0046] S201, collecting operation data sets during the operation of the elevator;

[0047] Through various sensors and data acquisition devices, various types of operation data of the elevator during operation are collected in real time. These data not only include the basic operation information of the elevator (such as speed, position, load, etc.), but also should cover multi-dimensional data related to the safety and status of the elevator operation (such as current, temperature, fault alarm signal, etc.). Through high-frequency collection, the integrity and real-time nature of the data can be ensured, providing a reliable foundation for subsequent fault detection and analysis. Accurate and comprehensive operation data sets are the basis of fault detection methods. Through real-time monitoring of data, potential fault hazards can be discovered in time and predicted. This not only helps to improve the safety of the elevator, but also optimizes the operation efficiency and maintenance plan of the elevator through effective data analysis, and reduces operation and maintenance costs.

[0048] Specifically, you can choose a variety of sensors suitable for elevators, including acceleration sensors, position sensors, weight sensors, temperature sensors, and current sensors to ensure comprehensive coverage of the elevator's operating status information. Deploy sensors in the elevator car, shaft, and control room to form a multi-level monitoring network to obtain data of different dimensions in real time. Use high-frequency sampling technology to monitor sensor output signals in real time. Each sensor should have an independent acquisition unit and transmit data to the central processing unit at the same time. Add a timestamp to each data collection to ensure the timing and traceability of the data. At the same time, configure unique identifiers for different data blocks and sensors for subsequent accurate analysis. Store all operating data in a time series database for analysis and fault detection. Store the collected operating data in a cloud database for subsequent data analysis and training of machine learning models. At the same time, build a version management mechanism for the database to ensure that historical data is not overwritten so that long-term trend analysis can be performed. Implement a data backup and recovery system to deal with data loss due to equipment failure or network problems.

[0049] S202, dividing the operation data set into a plurality of operation data blocks according to data types;

[0050] In this step, the system will classify and organize the elevator operation data set collected in the previous step to facilitate subsequent data processing and fault detection. The operation data set may contain multiple types of data, such as speed, load, current, temperature, position, etc. Therefore, dividing the data into multiple operation data blocks according to their types can effectively simplify the data structure, making subsequent analysis more efficient and clear.

[0051] By dividing the data set into multiple operating data blocks, the relationship between various types of data and the elevator operating status can be more clearly identified and analyzed. This structured approach can help the fault detection system locate potential problems more quickly and improve the accuracy and efficiency of fault detection. In addition, accurately classified data blocks also facilitate subsequent deep learning model training, ensuring that each model only focuses on analyzing data related to it, thereby improving the overall system intelligence level.

[0052] Specifically, the preprocessing algorithm can be used to perform preliminary analysis on the collected operating data to extract the main features of each data sample, including data type (such as continuous value, discrete value), data source (such as sensor type) and data timestamp. Preprocessing can use clustering-based algorithms (such as K-means) to perform preliminary clustering of data features to identify the basic features of data categories.

[0053] According to the characteristics of elevator operation, a multi-dimensional classification standard is constructed. For example, data can be classified according to the physical meaning of the data (such as speed, load), the frequency of data collection (such as real-time data, historical data), and the time attribute of the data (such as current period, historical period). A dynamic data block algorithm is used. Since new data types may be generated due to emergencies during elevator operation, the system needs to flexibly respond to the division of data blocks. The implementation method may include sliding window technology to dynamically create data blocks according to the conditions of running time or event triggering. For example, continuously monitor the data stream, and when new data appears, sort and group it according to the type and its characteristics. According to the data type, a multi-level data block structure is constructed. Each data block should include a main data block (such as a current data block, a speed data block, etc.), which may contain multiple sub-blocks (such as current status in different time periods, speed records on different floors), so that the hierarchical relationship between data blocks is clearer.

[0054] Assign a unique identifier to each operation data block and generate relevant metadata (such as data collection time, data sample size, anomaly markers, etc.) to facilitate quick search and retrieval during subsequent analysis. In the data block, fields should be included to explain the source and analysis status of the data to improve the availability of the data. Once new data is collected, the system will automatically check whether the current data block needs to be updated or re-divided. If there is a major change in the operation status (such as an elevator failure), the system will automatically adjust the division of the data block and reclassify the data. At the same time, set up a data archiving mechanism to regularly save all operation data blocks in the current state for subsequent use and analysis.

[0055] S203, if there is abnormal operation data in the operation data set, using the operation data set to detect a corresponding overall fault mode;

[0056] In this step, when the system detects abnormal data in the collected operation data set, it needs to use this data to detect the overall fault mode. These abnormal data may show that the elevator has encountered an abnormal state during operation, such as overload, overheating, or system failure. The overall fault mode is detected by comparing these abnormal operation data with historical fault data, using machine learning or deep learning technology to identify the overall fault type that the elevator may have. The importance of detecting the overall fault mode is to promptly identify problems that may cause the elevator to be unsafe or shut down, thereby preventing accidents. This process can not only help managers quickly locate faults, but also provide guidance for subsequent maintenance and overhaul, reducing maintenance costs and downtime. At the same time, the accurate detection of the overall fault mode lays the foundation for the intelligent operation and management of the elevator, enabling the elevator to automatically alarm and provide corresponding solutions when a fault occurs, thereby improving the safety and reliability of the elevator.

[0057] Specifically, the historical operation data set when the overall failure occurred during the historical operation of the elevator and its corresponding historical overall failure mode can be obtained;

[0058] In this step, the system will extract the operating data sets recorded when the overall failure occurred before, as well as the failure modes corresponding to these data sets from the historical database. These historical data provide real cases for the training of subsequent models, allowing the system to learn the characteristics and patterns when the failure occurs. By collecting and analyzing historical failure data, the system can build a huge knowledge base from which to extract the performance and status of the elevator under various failure conditions. This not only helps to improve the accuracy and robustness of the fault detection model, but also enables managers to understand the frequency and conditions of failures, optimize the maintenance and management strategies of the elevator, and reduce the occurrence of failures.

[0059] Using the historical operation data set and its corresponding historical overall failure mode, training an overall failure detection model based on deep learning;

[0060] Using the extracted historical operation data set and the corresponding fault mode data, a holistic fault detection model based on deep learning is trained. Through training, the model will be able to learn the complex relationship between the normal state and the fault state during the operation of the elevator, so that when faced with new data, it can accurately determine whether the current operation state has a fault. The model trained by deep learning technology has powerful feature extraction and pattern recognition capabilities and can handle complex data relationships. This model not only improves the accuracy of fault detection, but also improves the system's adaptive ability, enabling it to continuously learn and adapt to new operating environments. In addition, the establishment of the model can also provide a scientific basis for subsequent elevator maintenance and detection, and warn of potential fault risks in advance.

[0061] You can choose a deep learning framework suitable for this task, such as TensorFlow or PyTorch, to ensure that the model can handle large-scale data and high-complexity calculations. Build a deep neural network structure that adapts to the characteristics of elevator operation data, which can include multiple fully connected layers, convolutional layers to extract spatial features, and recurrent neural network (RNN) layers to process time series data. Divide the collected historical data set into training set, validation set, and test set, and perform normalization to improve the learning efficiency of the model. At the same time, use data enhancement technology to improve the generalization ability of the model. Use the training set to train the model, update the model parameters through iterative optimization algorithms (such as Adam or SGD), and monitor the overfitting of the model through the validation set, and adjust the model hyperparameters (such as learning rate, batch size, etc.). After training, use the test set to evaluate the model performance, calculate indicators such as accuracy, precision, and recall, and ensure that the model has good fault detection capabilities. Save the trained model and deploy it to the elevator monitoring system so that real-time fault detection can be performed during actual operation.

[0062] If there is abnormal operation data in the operation data set, the operation data set is input into the overall fault detection model, and the real-time overall fault mode of the elevator in the current period is output.

[0063] In this step, the system inputs the previously detected abnormal operation data set into the trained overall fault detection model to obtain the real-time overall fault mode of the elevator in the current period. The model analyzes the input data and generates corresponding outputs to help the system identify potential faults in a timely manner. By comparing and analyzing the abnormal data with the overall fault detection model in real time, the system can quickly identify the current fault state of the elevator and respond in time. This real-time monitoring mechanism greatly improves the safety and reliability of the elevator, ensures the safety of passengers, and can quickly take measures to reduce the occurrence of accidents through early warning.

[0064] Furthermore, if there is no abnormal operation data in the operation data set, the operation data set may also be used to predict potential overall failure modes.

[0065] In this step, when the system detects that there are no abnormal data in the operating data set, it will use these normal data to predict the possible overall failure mode. Through the analysis and modeling of the operating data, the system can identify the characteristics under normal operating conditions and infer the possible failure risks in a certain period of time in the future based on historical operating experience. This process usually involves the use of statistical modeling or machine learning algorithms to identify potential trends and patterns in the data.

[0066] The core significance of this step is to be able to identify potential failure risks in advance, thereby providing maintenance personnel with reliable early warning information. By predicting potential failure modes, the elevator management system can implement preventive maintenance, reduce the probability of failure, and improve the safety and reliability of the elevator. At the same time, accurate failure prediction can provide a decision-making basis for the operation and maintenance of the elevator, reduce operating costs, and improve customer satisfaction.

[0067] Specifically, if there is no abnormal operation data in the operation data set, the operation data set is used to predict the potential overall failure mode, and the historical overall failure mode when the overall failure occurred in the historical operation process of the elevator and the historical operation data set of the previous period can be obtained;

[0068] In this step, the system will extract relevant data from the historical database when the elevator had an overall failure in the past, including the corresponding failure mode and the operating data set before the failure. By analyzing this data, the system can better understand the root cause and characteristics of the failure. The significance of obtaining historical failure data is to provide high-quality training data for the fault prediction model. These historical data can help the model identify the operating characteristics before the failure and enhance its ability to predict future failures. Understanding past failure modes will also help maintenance personnel develop preventive strategies more efficiently and reduce downtime and maintenance costs.

[0069] Using the historical overall failure mode and its previous period historical operation data set, training an overall failure prediction model based on deep learning;

[0070] The key to this step is to use the extracted historical data set and failure mode information to train a deep learning model so that it can identify and predict future overall failure modes. Through deep learning technology, the model can learn complex nonlinear relationships and improve the accuracy of fault prediction. Training the overall fault prediction model of deep learning can improve the intelligence level of fault detection, allowing the elevator to actively warn of potential faults even when there is no abnormal data. This not only improves the safety of the elevator, but also provides data support for elevator operation and maintenance, optimizes maintenance plans, and reduces prevention costs.

[0071] You can choose a suitable deep learning framework (such as TensorFlow, PyTorch) based on your needs to ensure that it has the ability to process time series data. Design a neural network model structure suitable for elevator operation data, which may use a long short-term memory network (LSTM) or a convolutional neural network (CNN) to process time series data and extract time series features. Preprocess the collected historical operation data, including normalization, missing value processing, and time window segmentation to ensure that the data format meets the model input requirements. Use historical data sets to train the model, set the loss function and optimizer (such as Adam), optimize the model weights through back propagation, and use cross-validation methods to monitor model performance. Evaluate the prediction effect of the model on the validation set, including indicators such as accuracy, precision, and recall, to ensure that the model can effectively predict potential faults. After training, save the model and deploy it to the elevator monitoring system for real-time fault prediction.

[0072] If there is no abnormal operation data in the operation data set, the operation data set is input into the overall fault prediction model to output the potential overall fault mode of the elevator in the next period.

[0073] In this step, the system will input the current operation data set into the trained overall fault prediction model to obtain the prediction results of the overall fault mode that may occur in the next operation of the elevator. Through this process, the possible future failure risks of the elevator can be identified in time, thereby providing maintenance and management personnel with information about possible future failures. Doing so helps to deploy maintenance resources in advance and conduct necessary inspections and repairs to reduce downtime and user inconvenience caused by failures.

[0074] S204, searching for an operation data block corresponding to the abnormal operation data, and using the operation data block to detect a corresponding local fault mode;

[0075] In this step, the system will find the operation data block where the abnormal operation data is located based on the abnormal operation data previously identified. The operation data block is a subset of the collected elevator operation data divided by data type and time period, and each data block only contains relevant operation characteristics. When the corresponding operation data block is found, the system will use the data in the block to detect local fault modes and analyze the specific operation problems that may be caused by the abnormal data. The main function of this process is to refine fault detection, which can shift fault analysis from the overall to the specific, so as to more accurately locate the specific faulty components or systems of the elevator. Through the detection of local fault modes, maintenance personnel can take targeted maintenance measures in a timely manner to avoid more serious overall failures due to failure to deal with local problems in a timely manner, thereby improving the safety and reliability of elevator operation.

[0076] Specifically, historical operation data blocks and historical local failure modes of the elevator when multiple local failures occurred during its historical operation can be obtained;

[0077] In this step, the system will obtain the operation data blocks of various local faults that occurred in the past operation of the elevator from the historical data storage, as well as the fault modes corresponding to these data blocks. This step is intended to provide necessary training data for the subsequent local fault detection model. The purpose of obtaining these historical data blocks is to accumulate and analyze the operation characteristics of the elevator under local fault conditions, and help the system learn the laws when faults occur. This will help the training of subsequent models, enabling them to identify the characteristics of various local faults and improve the accuracy and response speed of detection.

[0078] Using the historical operation data blocks and their historical local fault modes, respectively training corresponding multiple local fault detection models based on deep learning;

[0079] The core of this step is to use the collected historical operation data blocks and their corresponding local fault modes to train multiple independent deep learning models. Each model focuses on identifying specific local fault characteristics, thereby enhancing the overall fault detection capability. By creating a separate model for each local fault, the system can improve the accuracy and response speed of fault detection. This segmented training method enables the model to be optimized specifically for each type of fault, thereby effectively reducing the false alarm rate and missed alarm rate, and improving the safety and operation and maintenance efficiency of the elevator.

[0080] You can select a suitable deep learning model architecture, such as a convolutional neural network (CNN) or a multi-layer perceptron (MLP), based on the local fault characteristics, and design the corresponding network layers and activation functions. Divide the historical data blocks into training sets, validation sets, and test sets to ensure the generalization ability of the model, including sufficient sample size and diversity. Train the model using the training set, optimize the model parameters using the back propagation algorithm, and monitor the overfitting phenomenon of the model through the validation set to make timely adjustments. Use the test set to evaluate the model performance, optimize the performance based on indicators such as precision and recall, and ensure that the model can perform well in practical applications. Save the trained model and integrate it into the elevator monitoring system to achieve real-time fault detection.

[0081] The operation data block corresponding to the abnormal operation data is searched, the operation data block is input into the corresponding local fault detection model, and the real-time local fault mode of the elevator in the current period is output.

[0082] In this step, the system will find the corresponding operating data block based on the previously detected abnormal data, and then input this data block into the specially trained local fault detection model to obtain the real-time local fault pattern of the current period. By inputting the operating data block corresponding to the abnormal data into the local fault detection model, the specific fault in the current elevator operation state can be quickly identified. This rapid response mechanism helps to solve potential problems in a timely manner and reduce the risk of faults developing into more serious problems, thereby improving the safety and reliability of elevators.

[0083] Furthermore, each operation data block without abnormal operation data may be used to predict potential local failure modes respectively.

[0084] In this step, the system processes each operating data block marked as normal and uses it to predict potential local failure modes. This process involves using a trained local failure prediction model to analyze the operating data under normal conditions to identify possible failure trends and hidden dangers. By predicting potential local failure modes for operating data blocks that do not have abnormalities, potential failure risks can be discovered early. This early warning mechanism helps maintenance personnel take preventive measures before a failure actually occurs, thereby reducing the probability of failure and improving the reliability and safety of the elevator. At the same time, this prediction can help optimize maintenance plans and achieve more efficient operation management.

[0085] Specifically, by using each operation data block without abnormal operation data, potential local failure modes are predicted respectively, so that historical local failure modes when multiple local failures occur in the historical operation process of the elevator and historical operation data blocks of the previous period can be obtained;

[0086] In this step, the system extracts relevant records of various local faults of the elevator from the historical database, including these fault modes and the running data blocks before the fault occurred. These data will serve as the basic data for training the local fault prediction model, which will help the model understand the characteristics of different local faults when they occur. Extracting relevant historical fault data is crucial for training the local fault prediction model. These data provide real cases, allowing the model to learn the characteristics of the fault and its occurrence conditions, enhance its prediction ability, and provide a reliable basis for subsequent elevator fault prediction.

[0087] Using the historical local fault mode and the historical operation data blocks of the previous period, respectively train corresponding multiple local fault prediction models based on deep learning;

[0088] This step involves training deep learning models using the collected historical data, especially the local fault pattern and its previous period of operation data blocks. Each model is optimized for a specific type of local fault to improve the accuracy of fault prediction. By training specialized deep learning models for different local faults, the system can deeply understand the characteristics and patterns of each fault. In this way, when the model is faced with new data, it can more accurately predict the occurrence of local faults, thereby effectively reducing maintenance costs and downtime risks, and ensuring the efficient and safe operation of the elevator.

[0089] You can choose a suitable deep learning framework (such as TensorFlow or PyTorch) according to the project requirements to support the implementation of the model. Design a reasonable neural network structure according to the fault characteristics, such as convolutional neural network (CNN) or long short-term memory network (LSTM), to ensure that the model can process time series data and extract effective features. Preprocess the historical data, including denoising, standardization, data segmentation (training set, validation set, test set), etc. At the same time, you can use data enhancement methods to increase the robustness of the model. Use the training set to train the model, use appropriate optimization algorithms (such as Adam), and monitor the model performance through the validation set, and adjust the hyperparameters in time to prevent overfitting. Use the test set to evaluate the trained model, and select the best performing model for subsequent application based on indicators such as precision and recall. Save the final selected model and deploy it to the elevator monitoring system to achieve real-time local fault prediction.

[0090] Each operation data block without abnormal operation data is input into the corresponding local fault prediction model, and the potential local fault mode of the elevator in the next time period is output.

[0091] In this step, the system inputs the current normal operation data block into each previously trained local fault prediction model to predict the local fault mode that may occur in the future. Each data block corresponds to a specific model, which can achieve a higher prediction accuracy. By inputting the normal operation data block into the local fault prediction model, maintenance personnel can promptly learn about the impending local fault risk of the elevator. This technology can effectively support the adjustment of maintenance strategies, avoid the occurrence of full-scale faults, and improve the operational efficiency and user safety of the elevator.

[0092] S205: Detect the fault location of the elevator according to the overall fault mode and the local fault mode.

[0093] Specifically, the fault location can be initially located in a large range according to the overall fault mode, and then further refined to a small range of precise locations according to the local fault mode. If there is a location conflict or inconsistency, the combined fault locations of the two corresponding locations will be used as the final fault location, and maintenance personnel will perform separate inspections.

[0094] In this step, the system first performs a preliminary large-scale fault location based on the detected overall fault pattern. This process uses the system-level abnormal state reflected by the overall fault pattern to determine the main areas or components where the fault may exist. Then, the system further refines the location through local fault patterns, narrowing the fault scope to specific components or functional areas. By combining these two stages, the system can more efficiently and accurately confirm the specific location of the fault, thereby guiding subsequent maintenance and repair work.

[0095] This hierarchical fault detection method has significant practical value. First, it can effectively reduce the time and resource waste of maintenance personnel in the troubleshooting process, because the initial broad positioning can quickly narrow the fault search range. Secondly, the refined positioning can ensure the accuracy of maintenance measures and avoid misjudgment and unnecessary repairs caused by incorrect fault positioning. In addition, accurate fault positioning can also provide important data support for subsequent fault analysis and preventive measures, thereby optimizing the elevator maintenance management system.

[0096] Furthermore, if there is abnormal operation data in the operation data set, the operation data set can also be used to calculate the fault degree index of the elevator, and the fault location and the fault degree index are sent to the elevator control terminal.

[0097] Specifically, using abnormal operation data to calculate the Failure Severity Index (FSI) can provide an important reference for the safe operation of elevators. Through in-depth analysis of abnormal data, the specific location of the fault and the severity of the fault are determined, and this information is sent to the elevator control end, so that the control end can take corresponding measures, such as issuing an alarm, adjusting the operation strategy, or performing maintenance inspections. This timely data transmission and feedback mechanism is the key to elevator fault prevention and safety management, which helps reduce the risk of accidents and ensure passenger safety and the normal operation of elevators.

[0098] Specifically, a calculation formula for the fault degree index may be:

[0099] The formula is designed to comprehensively consider the severity of abnormal events during elevator operation and their impact on overall safety, so as to quantify the degree of elevator failure. It comprehensively evaluates the operating status of the elevator through variables in multiple dimensions and provides a reliable basis for fault detection.

[0100] Among them, the FSI is the fault severity index, which is used to quantify the severity of the elevator fault. The higher the FSI, the greater the risk and impact of the fault, and vice versa. is the severity score of the kth abnormal event. Each abnormal event is assigned a score according to its impact on the elevator operation. The higher the score, the greater the impact of the abnormal event on the safe operation. is the weighting coefficient of the kth abnormal event, which reflects the importance of each abnormal event in the overall fault assessment. By weighting, more critical abnormal events can be emphasized. is the length of the entire operating cycle, is a relevance score for the mth relevant historical failure mode, where the relevant historical failure mode is a relevant historical failure mode mapped to the current abnormal event in the historical data, is the severity impact coefficient of the mth related historical failure mode, Score the repair effectiveness of the mth related historical failure mode, which is used to measure the success rate and effectiveness of historical repair measures. The higher the effectiveness, the better the repair measures can solve the failure in similar situations. is the number of abnormal events detected in the running data set, The number of related historical failure modes mapped to the current abnormal event in the historical data indicates the number of failure modes of historical abnormal events that are the same as the current abnormal event in the historical records. The more related modes there are, the more complex the potential failure types of the abnormal event may be. This is the adjustment coefficient, which is a flexible parameter that can be adjusted according to actual conditions to enhance the adaptability of the formula.

[0101] Through such a design, the fault severity index provides a comprehensive and dynamic assessment of the elevator's operating status, ensuring that the location and severity of problems can be identified as early as possible and appropriate measures can be taken when problems occur, thereby ensuring the safety and reliability of the elevator.

[0102] It can be seen that an operation data set is collected during the operation of the elevator; the operation data set is divided into multiple operation data blocks according to the data type; if there is abnormal operation data in the operation data set, the corresponding overall fault mode is detected using the operation data set; the operation data block corresponding to the abnormal operation data is found, and the corresponding local fault mode is detected using the operation data block; according to the overall fault mode and the local fault mode, the fault position of the elevator is detected, so that the fault position of the elevator can be accurately identified, thereby improving the timeliness and accuracy of fault detection and ensuring the safe operation of the elevator.

[0103] Another embodiment of the present invention provides an elevator operation fault detection system, see Figure 3 , the system may include:

[0104] The collection module 301 is used to collect the operation data set during the operation of the elevator;

[0105] A division module 302 is used to divide the operation data set into a plurality of operation data blocks according to data types;

[0106] A first detection module 303 is used to detect a corresponding overall fault mode using the operation data set if there is abnormal operation data in the operation data set;

[0107] A second detection module 304 is used to find an operation data block corresponding to the abnormal operation data, and use the operation data block to detect a corresponding local fault mode;

[0108] The third detection module 305 is used to detect the fault position of the elevator according to the overall fault mode and the local fault mode.

[0109] It can be seen that an operation data set is collected during the operation of the elevator; the operation data set is divided into multiple operation data blocks according to the data type; if there is abnormal operation data in the operation data set, the corresponding overall fault mode is detected using the operation data set; the operation data block corresponding to the abnormal operation data is found, and the corresponding local fault mode is detected using the operation data block; according to the overall fault mode and the local fault mode, the fault position of the elevator is detected, so that the fault position of the elevator can be accurately identified, thereby improving the timeliness and accuracy of fault detection and ensuring the safe operation of the elevator.

[0110] An embodiment of the present invention further provides a storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0111] Specifically, in this embodiment, the above storage medium may be configured to store a computer program for performing the following steps:

[0112] S201, collecting operation data sets during the operation of the elevator;

[0113] S202, dividing the operation data set into a plurality of operation data blocks according to data types;

[0114] S203, if there is abnormal operation data in the operation data set, using the operation data set to detect a corresponding overall fault mode;

[0115] S204, searching for an operation data block corresponding to the abnormal operation data, and using the operation data block to detect a corresponding local fault mode;

[0116] S205: Detect the fault location of the elevator according to the overall fault mode and the local fault mode.

[0117] It can be seen that an operation data set is collected during the operation of the elevator; the operation data set is divided into multiple operation data blocks according to the data type; if there is abnormal operation data in the operation data set, the corresponding overall fault mode is detected using the operation data set; the operation data block corresponding to the abnormal operation data is found, and the corresponding local fault mode is detected using the operation data block; according to the overall fault mode and the local fault mode, the fault position of the elevator is detected, so that the fault position of the elevator can be accurately identified, thereby improving the timeliness and accuracy of fault detection and ensuring the safe operation of the elevator.

[0118] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0119] Specifically, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0120] Specifically, in this embodiment, the processor may be configured to perform the following steps through a computer program:

[0121] S201, collecting operation data sets during the operation of the elevator;

[0122] S202, dividing the operation data set into a plurality of operation data blocks according to data types;

[0123] S203, if there is abnormal operation data in the operation data set, using the operation data set to detect a corresponding overall fault mode;

[0124] S204, searching for an operation data block corresponding to the abnormal operation data, and using the operation data block to detect a corresponding local fault mode;

[0125] S205: Detect the fault location of the elevator according to the overall fault mode and the local fault mode.

[0126] It can be seen that an operation data set is collected during the operation of the elevator; the operation data set is divided into multiple operation data blocks according to the data type; if there is abnormal operation data in the operation data set, the corresponding overall fault mode is detected using the operation data set; the operation data block corresponding to the abnormal operation data is found, and the corresponding local fault mode is detected using the operation data block; according to the overall fault mode and the local fault mode, the fault position of the elevator is detected, so that the fault position of the elevator can be accurately identified, thereby improving the timeliness and accuracy of fault detection and ensuring the safe operation of the elevator.

[0127] The above describes in detail the structure, features and effects of the present invention based on the embodiments shown in the drawings. The above is only a preferred embodiment of the present invention, but the present invention is not limited to the scope of implementation shown in the drawings. Any changes made according to the concept of the present invention, or modifications to equivalent embodiments with equivalent changes, which still do not exceed the spirit covered by the description and drawings, should be within the protection scope of the present invention.

Claims

1. A method for detecting elevator operation failure, characterized in that: The method comprises: Collecting an operation data set during the operation of the elevator; dividing the operation data set into a plurality of operation data blocks according to data types; if there is abnormal operation data in the operation data set, using the operation data set to detect a corresponding overall fault mode; finding an operation data block corresponding to the abnormal operation data, and using the operation data block to detect a corresponding local fault mode; detecting a fault location of the elevator according to the overall fault mode and the local fault mode; If there is abnormal operation data in the operation data set, the fault degree index of the elevator is calculated using the operation data set, and the fault position and the fault degree index are sent to the elevator control terminal; the calculation formula of the fault degree index is: ； Among them, the is the fault severity index, is the severity score of the kth abnormal event, is the weighting coefficient of the kth abnormal event, is the length of the entire operating cycle, is a relevance score for the mth relevant historical failure mode, where the relevant historical failure mode is a relevant historical failure mode mapped to the current abnormal event in the historical data, is the severity impact coefficient of the mth related historical failure mode, Score the repair effectiveness of the mth related historical failure mode, is the number of abnormal events detected in the running data set, is the number of relevant historical failure modes mapped to the current abnormal event in the historical data, is the adjustment coefficient.

2. The method according to claim 1, characterized in that The method further comprises: The potential local failure modes are predicted respectively by using the operation data blocks without abnormal operation data.

3. The method according to claim 2, characterized in that If there is abnormal operation data in the operation data set, detecting the corresponding overall failure mode using the operation data set includes: Obtain the historical operation data set when the overall failure occurred during the historical operation of the elevator and its corresponding historical overall failure mode; Using the historical operation data set and its corresponding historical overall failure mode, training an overall failure detection model based on deep learning; If there is abnormal operation data in the operation data set, the operation data set is input into the overall fault detection model, and the real-time overall fault mode of the elevator in the current period is output.

4. The method according to claim 3, characterized in that The step of searching for an operation data block corresponding to the abnormal operation data and using the operation data block to detect a corresponding local fault mode includes: Obtain historical operation data blocks and historical local failure modes when multiple local failures occurred during the historical operation of the elevator; Using the historical operation data blocks and their historical local fault modes, respectively training corresponding multiple local fault detection models based on deep learning; The operation data block corresponding to the abnormal operation data is searched, the operation data block is input into the corresponding local fault detection model, and the real-time local fault mode of the elevator in the current period is output.

5. The method according to claim 4, characterized in that The method of using each operation data block without abnormal operation data to predict potential local failure modes includes: Obtain the historical local failure mode when multiple local failures occurred during the historical operation of the elevator and the historical operation data block of the previous period; Using the historical local fault mode and the historical operation data blocks of the previous period, respectively train corresponding multiple local fault prediction models based on deep learning; Each operation data block without abnormal operation data is input into the corresponding local fault prediction model, and the potential local fault mode of the elevator in the next time period is output.

6. An elevator operation fault detection system, characterized in that: The system comprises: The acquisition module is used to collect the operation data set during the operation of the elevator; A division module, used for dividing the operation data set into a plurality of operation data blocks according to data types; A first detection module, configured to detect a corresponding overall fault mode using the operation data set if there is abnormal operation data in the operation data set; A second detection module, used to find an operation data block corresponding to the abnormal operation data, and use the operation data block to detect a corresponding local fault mode; A third detection module, used to detect a fault position of the elevator according to the overall fault mode and the local fault mode; If there is abnormal operation data in the operation data set, the fault degree index of the elevator is calculated using the operation data set, and the fault location and the fault degree index are sent to the elevator control terminal; the calculation formula of the fault degree index is: ； Among them, the is the fault severity index, is the severity score of the kth abnormal event, is the weighting coefficient of the kth abnormal event, is the length of the entire operating cycle, is a relevance score for the mth relevant historical failure mode, where the relevant historical failure mode is a relevant historical failure mode mapped to the current abnormal event in the historical data, is the severity impact coefficient of the mth related historical failure mode, Score the repair effectiveness of the mth related historical failure mode, is the number of abnormal events detected in the running data set, is the number of relevant historical failure modes mapped to the current abnormal event in the historical data, is the adjustment coefficient.

7. A storage medium, characterized in that: The storage medium stores a computer program, wherein the computer program is configured to execute the method according to any one of claims 1 to 5 when executed.

8. An electronic device, comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to perform the method according to any one of claims 1 to 5.

Images