Elevator operation data real-time analysis and abnormity early warning system
Through real-time analysis and abnormal warning system for elevator operation data, the LSTM model is used to predict the change trend of elevator status and identify abnormal points, which solves the problem of inadvertently predicting the early warning of elevator operation status and faults in the existing technology, and improves the efficiency and safety of elevator maintenance .
Patent Information
- Application Number
- CN202411186229.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-28
- Publication Date
- 2024-11-12
AI Technical Summary
The existing elevator monitoring system cannot effectively conduct in-depth analysis and excavation of elevator operation data, resulting in the inability to predict the trend of elevator operation status and early warning of faults, affecting the elevator operation efficiency and passenger safety.
A real-time analysis and abnormal warning system for elevator operation data is designed, including real-time data acquisition layer, data preprocessing and storage layer, prediction model construction layer, monitoring and feedback layer and abnormal warning layer, using the long-term and short-term memory network (LSTM) model Predict the future trend of elevator cab speed and acceleration, and discover abnormal points in time through the abnormal identification module, and send early warning signals to maintenance personnel.
Accurate prediction of elevator operating status and early warning of faults, avoid potential safety hazards, improve elevator maintenance efficiency and operation efficiency, and transform maintenance from passive response to active prevention.
Smart Images

Figure CN118929364A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of elevator maintenance, and in particular to a real-time analysis and abnormality warning system for elevator operation data. Background Art
[0002] As an indispensable vertical transportation tool in modern buildings, the safety and efficiency of elevator operation are receiving more and more attention. Most of the existing elevator monitoring systems can only realize the real-time display and storage of basic data, and lack the ability to deeply analyze and mine elevator operation data. This makes it impossible to effectively predict the changing trend of the elevator's operating status, such as changes in car speed and acceleration, making it difficult to provide a scientific basis for elevator maintenance and optimization. Moreover, due to the lack of abnormality recognition technology, the existing elevator system can only respond passively after a fault occurs, and cannot provide early warning of the fault. This lagging processing method not only affects the operating efficiency of the elevator, but also poses a potential threat to the safety of passengers. Summary of the invention
[0003] The purpose of the present invention is to provide a real-time analysis and abnormality warning system for elevator operation data to solve the problems raised in the above background technology.
[0004] To achieve the above object, the present invention provides the following technical solution: a real-time analysis and abnormal warning system for elevator operation data, comprising:
[0005] The elevator operation data real-time analysis and abnormal warning system is characterized by including:
[0006] The real-time data collection layer includes sensor components for real-time collection of relevant data of elevator operation, including elevator car speed, acceleration, elevator load, floor span and running direction;
[0007] The data preprocessing and storage layer includes a data cleaning module, a data transmission module and a data storage module. The data cleaning module is used to clean and denoise the collected raw data, the data transmission module is used to transmit the preprocessed data in real time, and the data storage module is used to store the preprocessed data persistently.
[0008] The prediction model building layer includes a model building module, which is used to build an elevator operation status prediction model. The model is used to predict the future change trend of the elevator car speed and acceleration according to the elevator production parameters and the conditions of the elevator load, floor span and running direction;
[0009] The monitoring and feedback layer includes a real-time data collection module, a model application module, an anomaly identification module and a user interaction interface. The real-time data collection module is used to receive real-time elevator operation data. The model application module is used to call the elevator operation status prediction model to obtain prediction data. The anomaly identification module is used to compare the prediction data with the real-time data to detect anomalies. The user interaction interface is used to display the real-time monitoring data, prediction data, the difference data between the two and the abnormal point information to the user;
[0010] The abnormal warning layer includes a warning message sending module, which is used to receive the abnormal point information output by the abnormal identification module, generate a warning signal according to the preset warning rules, and further send the warning signal to the elevator maintenance personnel in time.
[0011] Preferably, the step of the model application module calling the elevator operation status prediction model includes:
[0012] i) Obtain model input parameters, including production parameters of the elevator, and conditions of elevator load, floor span and running direction;
[0013] ii) passing the input parameters into the elevator operation status prediction model;
[0014] iii) receiving the predicted change trend of the elevator car speed and acceleration output by the model under the current input parameters.
[0015] Preferably, the step of performing data comparison by the abnormality identification module includes:
[0016] i) respectively plotting the real-time elevator operation data received by the real-time data collection module and the prediction data obtained by calling the elevator operation status prediction model by the model application module into a curve changing over time;
[0017] ii) align the timestamps of the two curve data to make the data at each time point correspond to each other;
[0018] iii) Calculate the difference between the predicted value and the real-time value at the corresponding time point and generate a difference curve;
[0019] iv) judging whether each point on the difference curve exceeds the normal range according to a preset threshold value, and if so, marking it as an abnormal point;
[0020] v) Transmit the difference curve and abnormal point information to the user interaction interface.
[0021] Preferably, the step of obtaining the elevator operation state prediction model includes:
[0022] Step 1: Collect elevator operation data, where the operation data is time series data of changes in elevator car speed and acceleration under specific operating parameters, where the specific operating parameters include elevator production parameters, elevator load, floor span, and operating direction;
[0023] Step 2: Clean the collected historical data to handle missing values and outliers;
[0024] Step 3: Standardize the cleaned feature data to form a training data set;
[0025] Step 4: Divide the training data set into a training set and a validation set;
[0026] Step 5: Select the long short-term memory network LSTM model to build an elevator operation status prediction model and initialize the parameters of the model;
[0027] Step 6: Use the training set data to iteratively train the LSTM model. During the training process, regularly use the validation set to evaluate the model performance until the model performance meets the requirements.
[0028] Step 7: After training is completed, save the weights and biases of the LSTM model and deploy the model for subsequent elevator operation status prediction.
[0029] Preferably, the elevator operation status prediction model consists of an input layer, multiple LSTM hidden layers and an output layer; the input layer receives elevator operation data features, including elevator production parameters, elevator load, floor span and operation direction; the LSTM hidden layer is used to learn the time series dependencies in the data; the output layer outputs the future change trend of the elevator car speed and acceleration.
[0030] Preferably, the LSTM model is iteratively trained using the training set data. In each iteration, a batch of input data is sent to the LSTM model, and the memory units and hidden states are updated by calculating the values of the input layer and the output layer, and finally the predicted output sequence is obtained. During the training process, the validation set is regularly used to evaluate the model performance, and the learning rate is adjusted or the training is stopped early as needed to prevent overfitting.
[0031] Preferably, during the training process, the mean square error MSE algorithm is used to evaluate the performance of the elevator operation state prediction model. The specific algorithm is:
[0032]
[0033] Where n is the number of data points, Y pred,i is the model prediction value, Y actual,iis the actual measured value; the prediction accuracy of the model is evaluated according to the MSE value. The smaller the MSE value, the higher the prediction accuracy of the model.
[0034] Preferably, in the above step 2, for missing values, a linear interpolation algorithm is used to fill them; for outliers, a threshold judgment method is used to regard data points that exceed a preset threshold range as outliers and replace them with the mean of their neighboring points.
[0035] Preferably, in step 3, the Z-score standardization method is used to subtract the mean of each numerical data and divide it by its standard deviation so that the processed data conforms to the standard normal distribution. The formula for standardization is: Z=X-μ / σ; wherein X is the original data, μ is the mean of the original data, σ is the standard deviation of the original data, and Z is the standardized data.
[0036] Preferably, the user interaction interface includes a real-time monitoring data display area, a model prediction data display area and a difference data display area. The real-time monitoring data display area is used to display a curve of real-time elevator operation data changing over time, the model prediction data display area is used to display a curve of predicted elevator operation status changing over time, and the difference data display area is used to display a curve of the difference between real-time data and predicted data changing over time.
[0037] Compared with the prior art, the present invention has the following beneficial effects:
[0038] By collecting elevator operation data in real time and conducting in-depth analysis and mining, the system can accurately predict the changing trend of elevator operation status, such as changes in car speed and acceleration. The abnormality recognition module can promptly detect abnormal points in elevator operation and provide early warning of faults, thereby avoiding potential safety hazards and ensuring the safety of passengers.
[0039] The system can automatically identify abnormal conditions in elevator operation based on the comparison between the prediction model and real-time data, and promptly notify maintenance personnel to conduct maintenance. This transforms elevator maintenance from traditional passive response to active prevention, greatly reducing elevator downtime caused by faults and improving elevator operation and maintenance efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] Figure 1 It is the overall structure diagram of the present invention;
[0041] Figure 2 Execute step diagram for model application module;
[0042] Figure 3 Execute the steps diagram for the anomaly identification module. DETAILED DESCRIPTION
[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0044] See also Figure 1-3 The present invention provides a technical solution: a real-time analysis and abnormal warning system for elevator operation data, including a real-time data acquisition layer, a data preprocessing and storage layer, a prediction model building layer, a monitoring and feedback layer, and an abnormal warning layer.
[0045] In the prediction model building layer, the model building module first builds an elevator operation status prediction model based on the elevator's production parameters (such as elevator model, manufacturer, factory date, etc.) and various conditions encountered during elevator operation (such as elevator load, floor span and running direction). The model uses machine learning algorithms, especially time series prediction algorithms, to predict future changes in elevator car speed and acceleration. The model's training data comes from historical elevator operation data, and the prediction accuracy is improved by continuously optimizing model parameters.
[0046] The real-time data collection layer includes sensor components, which are installed at key locations inside and outside the elevator to collect relevant data on the elevator operation in real time. The sensor components include but are not limited to speed sensors, acceleration sensors, load sensors, and position sensors, which are used to measure the speed, acceleration, current load, floor span, and running direction of the elevator car. The collected data is transmitted to the data preprocessing and storage layer.
[0047] After receiving the raw data from the real-time data collection layer, the data preprocessing and storage layer first cleans and denoises the data through the data cleaning module to ensure the accuracy and reliability of the data. The cleaned data is transmitted to the monitoring and feedback layer in real time through the data transmission module, and is persistently stored through the data storage module for subsequent analysis and historical data backtracking.
[0048] The real-time data collection module of the monitoring and feedback layer continuously receives real-time elevator operation data from the data preprocessing and storage layer. The model application module calls the elevator operation status prediction model to generate prediction data for a period of time in the future based on real-time data. The anomaly identification module compares the prediction data with the real-time data and detects any significant deviations or abnormal patterns through a preset algorithm. The user interface is responsible for displaying real-time monitoring data, prediction data, the difference data between the two, and detected abnormal point information to the elevator management personnel or maintenance personnel.
[0049] The abnormal warning layer includes a warning message sending module, which receives the abnormal point information output by the abnormal identification module and generates a warning signal according to the preset warning rules (such as abnormal severity, duration, etc.). The warning signal can be sent to the elevator maintenance personnel in a timely manner through SMS, email, instant messaging or push of a specific application so that they can respond quickly and deal with potential elevator failure problems.
[0050] The present invention will be further described below in conjunction with Examples 1 to 3:
[0051] Embodiment 1:
[0052] The model application module is responsible for calling the elevator operation status prediction model to obtain the predicted change trend of the elevator car speed and acceleration. The execution steps of this module include:
[0053] i) Get model input parameters:
[0054] The model application module first obtains the real-time operation data of the elevator from the real-time data acquisition layer, and extracts the input parameters required for the model. The input parameters include the production parameters of the elevator (such as elevator model, manufacturer, date of manufacture, etc.), as well as the current load, floor span, and running direction of the elevator.
[0055] ii) Input prediction model:
[0056] After obtaining the input parameters, the model application module passes these parameters to the elevator operation status prediction model. The model is built based on a machine learning algorithm, has been trained and optimized, and can predict the future trend of elevator car speed and acceleration based on the input parameters.
[0057] iii) Receive prediction output:
[0058] After receiving the input parameters, the elevator operation status prediction model outputs the predicted change trend of the elevator car speed and acceleration. The model application module receives these prediction outputs and passes them to the abnormality recognition module for subsequent analysis and processing.
[0059] The anomaly recognition module is responsible for comparing the real-time elevator operation data with the predicted data to detect possible anomalies. The execution steps of this module include:
[0060] i) Data curve drawing:
[0061] The abnormality identification module first receives the real-time elevator operation data from the real-time data collection module and the prediction data from the model application module. Then, these data are plotted into time-varying curves for intuitive comparison.
[0062] ii) Data timestamp alignment:
[0063] In order to ensure the temporal correspondence between real-time data and predicted data, the anomaly recognition module will align the timestamps of the two curve data to ensure that the data at each time point is consistent and comparable, providing an accurate basis for subsequent difference calculations.
[0064] iii) Difference calculation:
[0065] After aligning the data, the anomaly recognition module calculates the difference between the predicted value and the real-time value at the corresponding time point. These differences form a difference curve, which reflects the degree of deviation between the predicted data and the real-time data.
[0066] iv) Abnormal point judgment:
[0067] According to the preset threshold, the abnormality identification module will determine whether each point on the difference curve exceeds the normal range. If the difference of a point exceeds the threshold, it will be marked as an abnormal point, indicating that there may be a problem with the operation status of the elevator at that time.
[0068] v) Information transmission and display:
[0069] Finally, the abnormality identification module transmits the difference curve and abnormal point information to the user interface. The user interface is responsible for displaying this information to the elevator management or maintenance personnel in an intuitive way so that they can promptly discover and deal with potential elevator failures.
[0070] The following are specific application examples:
[0071] Obtain model input parameters: The real-time data acquisition layer provides real-time operation data of elevator A. From it, the input parameters required for the model are extracted: elevator model (Model X), manufacturer (Company Y), production date (January 2020), current load (800 kg), floor span (10 floors) and running direction (ascending).
[0072] Pass these parameters into the prediction model: Pass these parameters into the elevator operation status prediction model. The model is based on a machine learning algorithm, has been trained and optimized, and can predict the future trend of elevator car speed and acceleration.
[0073] Receive prediction output: The model outputs the predicted change trend of the speed and acceleration of elevator A. The model application module receives these prediction outputs and passes them to the anomaly recognition module.
[0074] Data curve drawing: The real-time data collection module provides the real-time operation data of elevator A. The abnormal identification module receives the prediction data from the model application module and draws these data into curves that change over time.
[0075] Data timestamp alignment: Align the timestamps of real-time data and forecast data to ensure that the data at each time point is consistent and comparable.
[0076] Difference calculation: Calculate the difference between the predicted value and the real-time value at the corresponding time point to form a difference curve.
[0077] Abnormal point judgment: The preset threshold is that the deviation of speed and acceleration shall not exceed 5%. At a certain point in time, the point on the difference curve exceeds the normal range, indicating that there may be a problem with the operation status of elevator A.
[0078] Information transmission and display: The abnormality identification module transmits the difference curve and abnormal point information to the user interface. The user interface is displayed to the elevator management personnel in an intuitive way, and the management personnel promptly discover and notify the maintenance personnel to deal with potential elevator failures.
[0079] Embodiment 2:
[0080] Elevator operation status data usually contains complex nonlinear relationships and patterns. LSTM (Long Short-Term Memory Network) has the ability to learn nonlinear functions, which enables the model to better capture these complex features and make accurate predictions based on them. LSTM can effectively capture and retain long-term dependencies through its unique memory units and gating mechanisms (including input gates, forget gates, and output gates), which is particularly important for elevator operation status prediction because the operation status of elevators is often affected by multiple previous time steps. The memory units of LSTM enable the model to remember key information from the past and apply it to the prediction of the current time step, thereby improving the accuracy of the prediction.
[0081] The method for obtaining the elevator operation state prediction model comprises the following steps:
[0082] Step 1: Collect elevator operation data. This data is the time series data of the elevator's car body speed and acceleration changes under specific operating parameters. The specific operating parameters cover the elevator production parameters (such as elevator model, manufacturer, factory date, etc.), elevator load, floor span and running direction, ensuring the comprehensiveness and accuracy of the data.
[0083] Step 2: Clean the collected historical data to deal with missing values and outliers in the data to improve the quality and reliability of the data. For missing values, linear interpolation algorithm is used to fill them; for outliers, threshold judgment method is used to treat data points that exceed the preset threshold range as outliers and replace them with the mean of their neighboring points. Through the cleaning process, the effectiveness and accuracy of subsequent model training can be ensured.
[0084] Step 3: Standardize the cleaned feature data to eliminate the dimensional differences between different features, making the data more suitable for model training. Using the Z-score standardization method, subtract the mean of each numerical data and divide it by its standard deviation so that the processed data conforms to the standard normal distribution. The formula for standardization is: Z = X-μ / σ; where X is the original data, μ is the mean of the original data, σ is the standard deviation of the original data, and Z is the standardized data. After standardization, the data will form a training data set, providing a basis for subsequent model training.
[0085] Step 4: Divide the training data set into a training set and a validation set for effective cross-validation during model training to evaluate the performance and generalization ability of the model. By dividing the training set and the validation set, the stability and reliability of the model during the training process can be ensured.
[0086] Step 5: Select the long short-term memory network (LSTM) model to build the elevator operation status prediction model and initialize the parameters of the LSTM model.
[0087] Step 6: Use the training set data to iteratively train the LSTM model. During the training process, the validation set will be used regularly to evaluate the performance of the model to ensure the effectiveness and accuracy of the model during the training process. Through continuous iterative training, until the performance of the model meets the requirements, that is, it can accurately predict the changing trend of the elevator operation status.
[0088] Step 7: After training is complete, save the weights and biases of the LSTM model so that the trained model can be deployed.
[0089] The elevator operation status prediction model consists of an input layer, multiple LSTM hidden layers and an output layer; the input layer receives the elevator operation data features, including elevator production parameters, elevator load, floor span and operation direction; the LSTM hidden layer is used to learn the time series dependencies in the data; the output layer outputs the future change trend of the elevator car speed and acceleration.
[0090] Model training process:
[0091] Iterative training: Use the training set data to iteratively train the LSTM model. In each iteration, a batch of input data is fed into the LSTM model, and the memory cells and hidden states are updated by calculating the values of the input layer and the output layer. This process enables the model to gradually learn the time series dependencies in the elevator operation data and optimize its prediction performance. During the training process, the mean square error MSE algorithm is used to evaluate the performance of the elevator operation status prediction model. The specific algorithm is:
[0092]
[0093] Where n is the number of data points, Y pred,i is the model prediction value, Y actual,i is the actual measured value; the prediction accuracy of the model is evaluated according to the MSE value. The smaller the MSE value, the higher the prediction accuracy of the model.
[0094] Performance evaluation and adjustment: During the training process, the performance of the model is evaluated regularly using the validation set. The accuracy and generalization ability of the model can be evaluated by calculating the difference between the predicted output and the actual running data. Based on the results of the performance evaluation, the learning rate can be adjusted or training can be stopped early to prevent the model from overfitting and improve its generalization performance.
[0095] The following is a Python code example for building and training an elevator operation status prediction model:
[0096] import numpy as np
[0097] import pandas as pd
[0098] from sklearn.model_selection import train_test_split
[0099] from sklearn.preprocessing import StandardScaler
[0100] from keras.models import Sequential
[0101] from keras.layers import LSTM.Dense
[0102] #DataFrame containing elevator operation data, named df
[0103] # df has columns including 'timestamp', 'speed', 'acceleration', 'load', 'floorspan', 'direction', etc.
[0104] #Step 1: Data cleaning
[0105] #Step 2: Feature Standardization
[0106] scaler = StandardScaler()
[0107] features=df[['load','floor_sρan','speed','acceleration']]
[0108] features_scaled=scaler.fit_Transform(features)
[0109] #Step 3: Divide the training set and validation set
[0110] X = features_scaled[:_1] # features (except the last data point)
[0111] y = df['speed'][1:].values # target variable (speed, except the first data point)
[0112] X_train, X_val, y_train, y_val=train_test_split(X, y, test_size=0.2, random_state=42)
[0113] #For LSTM input, we need to reshape the data into [samples, time_steps, features]
[0114] time_steps=1
[0115] X_train=X_train.reshape(X_train.shape[0], time_steps, X_train.shape[1])
[0116] X_val=X_val.reshape(X_val.shape[0], time_steps, X_val.shape[1])
[0117] #Step 4: Build LSTM model
[0118] model=Sequential0
[0119] model.add(LSTM(50, activation='relu′, input_shape=(time_steps, X_train.shape[2])))
[0120] model.add(Dense(1))
[0121] model.compile(optimizer='adam', loss='mse')
[0122] #Step 5: Train the model
[0123] model.fit(X_train, y_train, epochs=100, batch_size=32, validation_data=(X_val, y_val),
[0124] verbose=1)
[0125] #Step 6: Save the model
[0126] model.save(′elevator_prediction_model.h5′)
[0127] The above code implements an elevator operation status prediction model based on the long short-term memory network (LSTM). The following is a detailed implementation description:
[0128] Data preparation: A DataFrame containing elevator operation data, named df. The columns of df include timestamp, speed, acceleration, load, floor span, and direction.
[0129] Feature standardization: StandardScaler is used to standardize the feature data to eliminate the dimensional differences between different features. Load, floor span, speed, and acceleration are selected as features.
[0130] Divide the training set and validation set: Divide the feature data X and the target variable y (speed) into a training set and a validation set. Use the train_test_split function to divide, and the test set size is 20%.
[0131] Data reshaping: The LSTM model requires the input data to be in the shape of [samples, time_steps, features].
[0132] In this example, time_steps is set to 1, indicating that each sample contains only one time step of data. Reshape the data of the training set and validation set into the input shape required by the LSTM model.
[0133] Build LSTM model: Use Sequential model to build LSTM network. Add an LSTM layer with 50 neurons and use ReLU activation function. Add a fully connected layer (Dense layer) with output dimension of 1, corresponding to the predicted speed value. Use Adam optimizer and mean square error (MSE) loss function to compile the model.
[0134] Training model: Use the training set data to iteratively train the LSTM model. Set the number of iterations (epochs) to 100 and the batch size (batch_size) to 32. During the training process, use the validation set data to evaluate the performance of the model.
[0135] Save model: Save the trained LSTM model as an HDF5 file (.h5) for subsequent elevator operation status prediction.
[0136] Embodiment 3:
[0137] The abnormal warning layer includes a warning message sending module, the main function of which is to receive the abnormal point information output by the abnormal identification module. This information contains abnormal data points in the operation of the elevator, such as excessive speed, abnormal acceleration, etc. After receiving this information, the warning message sending module will analyze and process these abnormal point information according to the preset warning rules.
[0138] The warning rule is a set of preset conditions and logic used to determine whether the received abnormal point information meets the criteria for sending a warning signal. If the abnormal point information meets the conditions in the warning rule, the warning message sending module will generate a corresponding warning signal. The generated warning signal contains detailed information about the abnormal point, such as the abnormality type, occurrence time, specific location, etc., to help elevator maintenance personnel quickly and accurately locate and handle the abnormality.
[0139] Finally, the warning message sending module will send the generated warning signal to the elevator maintenance personnel in a timely manner. This is usually achieved through email, SMS, mobile application push, etc., to ensure that the maintenance personnel can receive the warning signal in time and take appropriate measures to deal with the abnormality of the elevator.
[0140] The user interaction interface includes a real-time monitoring data display area, a model prediction data display area, and a difference data display area.
[0141] Real-time monitoring data display area: The main function of this area is to display real-time elevator operation data. It clearly depicts how the elevator operation data changes over time in the form of a curve chart. Through this area, users can intuitively observe the changes in various parameters of the elevator in actual operation, such as speed, acceleration, etc.
[0142] Model prediction data display area: This area is specifically used to display the predicted data of the elevator operation status. It also uses a curve chart to intuitively present the predicted trend of the elevator operation status over time. Through this area, users can clearly see the model's prediction results for the future operation status of the elevator.
[0143] Difference data display area: The main task of this area is to display the difference between real-time data and predicted data. It accurately depicts how the difference between real-time data and predicted data changes over time in the form of a curve graph. Through this area, users can intuitively understand the accuracy of the model prediction and the deviation between real-time data and predicted data.
[0144] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.
[0145] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. Real-time analysis and abnormal warning system of elevator operation data, characterized by: include: The real-time data collection layer includes sensor components for real-time collection of relevant data of elevator operation, including elevator car speed, acceleration, elevator load, floor span and running direction; The data preprocessing and storage layer includes a data cleaning module, a data transmission module and a data storage module. The data cleaning module is used to clean and denoise the collected raw data, the data transmission module is used to transmit the preprocessed data in real time, and the data storage module is used to store the preprocessed data persistently. The prediction model building layer includes a model building module, which is used to build an elevator operation status prediction model. The model is used to predict the future change trend of the elevator car speed and acceleration according to the elevator production parameters and the conditions of the elevator load, floor span and running direction; The monitoring and feedback layer includes a real-time data collection module, a model application module, an anomaly identification module and a user interaction interface. The real-time data collection module is used to receive real-time elevator operation data. The model application module is used to call the elevator operation status prediction model to obtain prediction data. The anomaly identification module is used to compare the prediction data with the real-time data to detect anomalies. The user interaction interface is used to display the real-time monitoring data, prediction data, the difference data between the two and the abnormal point information to the user; The abnormal warning layer includes a warning message sending module, which is used to receive the abnormal point information output by the abnormal identification module, generate a warning signal according to the preset warning rules, and further send the warning signal to the elevator maintenance personnel in time.
2. The elevator operation data real-time analysis and abnormal warning system according to claim 1 is characterized in that: The step of the model application module calling the elevator operation status prediction model comprises: i) Obtain model input parameters, including production parameters of the elevator, and conditions of elevator load, floor span and running direction; ii) passing the input parameters into the elevator operation status prediction model; iii) receiving the predicted change trend of the elevator car speed and acceleration output by the model under the current input parameters.
3. The elevator operation data real-time analysis and abnormal warning system according to claim 2 is characterized by: The step of performing data comparison by the abnormality identification module includes: i) respectively plotting the real-time elevator operation data received by the real-time data collection module and the prediction data obtained by calling the elevator operation status prediction model by the model application module into a curve changing over time; ii) align the timestamps of the two curve data to make the data at each time point correspond to each other; iii) Calculate the difference between the predicted value and the real-time value at the corresponding time point and generate a difference curve; iv) judging whether each point on the difference curve exceeds the normal range according to a preset threshold value, and if so, marking it as an abnormal point; v) Transmit the difference curve and abnormal point information to the user interaction interface.
4. The elevator operation data real-time analysis and abnormal warning system according to claim 1 is characterized in that: The steps of obtaining the elevator operation status prediction model include: Step 1: Collect elevator operation data, where the operation data is time series data of changes in elevator car speed and acceleration under specific operating parameters, where the specific operating parameters include elevator production parameters, elevator load, floor span, and operating direction; Step 2: Clean the collected historical data to handle missing values and outliers; Step 3: Standardize the cleaned feature data to form a training data set; Step 4: Divide the training data set into a training set and a validation set; Step 5: Select the long short-term memory network LSTM model to build an elevator operation status prediction model and initialize the parameters of the model; Step 6: Use the training set data to iteratively train the LSTM model. During the training process, regularly use the validation set to evaluate the model performance until the model performance meets the requirements. Step 7: After training is completed, save the weights and biases of the LSTM model and deploy the model for subsequent elevator operation status prediction.
5. The elevator operation data real-time analysis and abnormal warning system according to claim 4 is characterized by: The elevator operation status prediction model consists of an input layer, multiple LSTM hidden layers and an output layer; the input layer receives the elevator operation data features, including elevator production parameters, elevator load, floor span and operation direction; the LSTM hidden layer is used to learn the time series dependencies in the data; the output layer outputs the future change trend of the elevator car speed and acceleration.
6. The elevator operation data real-time analysis and abnormal warning system according to claim 5 is characterized by: The LSTM model is iteratively trained using the training set data. In each iteration, a batch of input data is fed into the LSTM model, and the values of the input layer and the output layer are calculated, the memory cells and hidden states are updated, and finally the predicted output sequence is obtained; During training, periodically evaluate model performance using the validation set and adjust the learning rate or stop training early as needed to prevent overfitting.
7. The elevator operation data real-time analysis and abnormal warning system according to claim 6 is characterized in that: During the training process, the mean square error (MSE) algorithm is used to evaluate the performance of the elevator operation status prediction model. The specific algorithm is: Where n is the number of data points, Y pred,i is the model prediction value, Y actual,i is the actual measured value; the prediction accuracy of the model is evaluated according to the MSE value. The smaller the MSE value, the higher the prediction accuracy of the model.
8. The elevator operation data real-time analysis and abnormal warning system according to claim 4 is characterized by: In the above step 2, for missing values, linear interpolation algorithm is used to fill them; for outliers, threshold judgment method is used to regard data points beyond the preset threshold range as outliers and replace them with the mean of their neighboring points.
9. The elevator operation data real-time analysis and abnormal warning system according to claim 4 is characterized by: In step 3, the Z-score standardization method is used to subtract the mean of each numerical data and divide it by its standard deviation so that the processed data conforms to the standard normal distribution. The standardization formula is: Z = X-μ / σ; where X is the original data, μ is the mean of the original data, σ is the standard deviation of the original data, and Z is the standardized data.
10. The elevator operation data real-time analysis and abnormal warning system according to claim 1 is characterized by: The user interaction interface includes a real-time monitoring data display area, a model prediction data display area and a difference data display area. The real-time monitoring data display area is used to display a curve showing the real-time elevator operation data changing over time. The model prediction data display area is used to display a curve showing the predicted elevator operation status changing over time. The difference data display area is used to display a curve showing the difference between the real-time data and the predicted data changing over time.
Citation Information
Patent Citations
Elevator system
CN101678994A
Fire early warning and state monitoring device and method based on 5G edge computing
CN113763667A
Dynamic monitoring method for force application load of multi-car intelligent parallel elevator
CN115535759A
Early warning method and system of dynamic environment monitoring system based on time sequence
CN117872839A
Group management control device of elevators
CN1802303A
Cited By
Comprehensive experimental device for refrigeration and heat exchange of air conditioner
CN118913738A
Microorganism data visualization chart generation system and method based on model parameters
CN120388622A
