Method and system for predicting residual life of elevator traction system

Abstract

The invention relates to the technical field of elevator operation and maintenance, and provides a residual life prediction method and system for an elevator traction system. The method integrates links of data acquisition, storage, preprocessing, model training verification, result comparison and the like. Firstly, elevator operation data are comprehensively collected, a data storage mechanism is established, and a solid foundation is laid for data analysis and model training. The accuracy and generalization ability of model training are ensured through data preprocessing and set division. And an elevator number and data matching mechanism is introduced, and refined data management is achieved. The accuracy of the model is verified by comparing the first prediction result with the second prediction result, the prediction reliability is improved, and a basis is provided for model optimization. And finally, comparing the data features of the third prediction result with actual data to further ensure the prediction accuracy. According to the method and the system, powerful support is provided for safe operation and maintenance and preventive maintenance of the elevator, and accurate prediction of the residual life of the elevator traction system is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of elevator operation and maintenance, and in particular to a method and system for predicting the remaining life of an elevator traction system. Background Art

[0002] The remaining life of the elevator traction system depends on many factors, such as equipment quality, operating environment, maintenance, etc. Therefore, it is difficult to give a specific remaining life. Generally speaking, if the elevator traction system is well maintained and serviced, its service life can reach the design life. However, if the equipment quality is poor, the operating environment is harsh, or the maintenance is not in place, the service life of the elevator traction system may be greatly shortened.

[0003] In order to ensure the normal operation of the elevator traction system and extend its service life, it is recommended to take the following measures: Regularly check the elevator traction system, including the traction machine, wire rope, guide rails and other components to ensure that they are in good working condition. Perform maintenance according to the maintenance plan recommended by the manufacturer, replace worn parts in a timely manner, and keep the equipment in good condition. Avoid improper use such as overloading and overtime use to reduce wear and damage to the equipment. Handle faults in a timely manner: Once a fault is found in the elevator traction system, it should be stopped immediately and a professional should be asked to repair it to avoid the expansion of the fault and cause more serious consequences.

[0004] The existing remaining life prediction method of the elevator traction system is based on the prediction method of eddy current detection. It uses an oscillator and a detection device to monitor the running elevator traction wire rope in real time by eddy current detection, and converts the monitoring data into a detection signal through a signal conditioning module. The analysis and evaluation module compares the collected detection signal with the information on different wear degrees and life of the elevator traction wire rope pre-stored in the database, so as to predict the life of the detected elevator traction wire rope.

[0005] The accuracy of the existing prediction of the remaining life of the elevator traction system is highly dependent on the quality of the prediction model and the information stored in the database. If the model is not accurate enough or the database information is not complete, the prediction results may be biased. Although eddy current detection can be used for real-time monitoring, long-term continuous monitoring may generate a large amount of data, which places higher requirements on data processing and analysis capabilities. At the same time, real-time monitoring also means that the equipment needs to run continuously, which may increase the wear and energy consumption of the equipment, resulting in reduced monitoring accuracy of the real-time monitoring equipment. Summary of the invention

[0006] The purpose of the present invention is to provide a method and system for predicting the remaining life of an elevator traction system, so as to solve the problem that the accuracy of the existing prediction of the remaining life of an elevator traction system is highly dependent on the quality of the prediction model and the information stored in the database. If the model is not accurate enough or the database information is not perfect enough, the prediction result may be biased.

[0007] The present invention provides a method for predicting the remaining life of an elevator traction system, comprising:

[0008] Acquire the elevator traction system operation data and store it in a database for storing elevator operation and maintenance data, wherein the elevator traction system operation data includes the load data of the traction system, the elevator running speed, acceleration data, running time data, fault and abnormal situation records, use environment data, maintenance history data and elevator energy consumption data;

[0009] After data preprocessing, the elevator traction system operation data is divided into a training set, a validation set and a test set. The training set is used to train the model, the validation set is used to adjust the preset elevator traction system remaining life prediction model parameters and hyperparameters, and the test set is used to evaluate the final performance of the preset elevator traction system remaining life prediction model;

[0010] Establish elevator numbers for elevators in the operation and maintenance area, match the real-time collected elevator traction system operation data with the elevator number, and store them in the corresponding storage space in the database, retrieve the maintenance history data of the elevators in the operation and maintenance area, extract data features from the maintenance history data of the elevators in the operation and maintenance area, create a data inspection form based on the data features of the maintenance history data of the elevators, and collect the corresponding feature data in the data inspection form according to the preset time period based on the data inspection form;

[0011] Substitute the corresponding characteristic data in the inspection form of the preset time period data collection into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the first elevator traction system, retrieve the elevator traction system operation data corresponding to the time when the elevator failure occurred, and substitute the data into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the second elevator traction system, compare the remaining life prediction result of the first elevator traction system with the remaining life prediction result of the second elevator traction system, if the data comparison results are consistent, the remaining life prediction model of the elevator traction system and the elevator traction system data monitoring are normal, and output the remaining life prediction result of the third elevator traction system, if the data comparison results are inconsistent, optimize the remaining life prediction model of the elevator traction system, and detect the corresponding data collection interface in the elevator traction system operation data;

[0012] Extract the data features in the remaining life prediction result of the third elevator traction system, obtain the data feature information of the third prediction result, collect the corresponding data in the real-time third prediction result data, compare the remaining life prediction result of the third elevator traction system with the actual data, check the degree of consistency between the remaining life prediction result of the third elevator traction system and the corresponding data in the collected real-time third prediction result data, if the prediction result is significantly different from the actual data, it is necessary to adjust and optimize the remaining life prediction model of the elevator traction system.

[0013] Further, the elevator traction system operation data is obtained, wherein the elevator traction system operation data includes the load data of the traction system, the elevator operation speed, acceleration data, operation time data, fault and abnormal condition records, use environment data, maintenance history data and elevator energy consumption data, including;

[0014] Determine the type of elevator traction system operating data and set the synchronization frequency;

[0015] The database architecture consists of a master database and a slave database. The master database is used to process write operations, and the slave database is used to process read operations and data backup.

[0016] The load data of the traction system, the elevator running speed, acceleration data, running time data, fault and abnormal condition records, use environment data, maintenance history data and elevator energy consumption data in the elevator traction system operation data are split into multiple elevator traction system data slices, and each elevator traction system data slice is stored on a different database node;

[0017] After each elevator traction system data shard is stored on different database nodes, the data is synchronously stored in the blockchain, and a corresponding hash value and timestamp are generated for each elevator traction system data shard stored on different database nodes;

[0018] If during the synchronization of the elevator traction system data, if a data conflict of the elevator traction system is encountered during the synchronization process, the conflicting data is distinguished according to the hash value and timestamp corresponding to the data. If the data is the same data, only the data label and the retrieval path of the data corresponding to the label are established.

[0019] Furthermore, after data preprocessing of the elevator traction system operation data, the data set is divided into a training set, a validation set and a test set, the training set is used to train the model, the validation set is used to adjust the preset parameters and hyperparameters of the elevator traction system remaining life prediction model, and the test set is used to evaluate the final performance of the preset elevator traction system remaining life prediction model, including;

[0020] Performing data cleaning on the elevator traction system operation data, removing abnormal values, noise data and duplicate data in the elevator traction system operation data, and obtaining the elevator traction system operation data after data cleaning;

[0021] Convert the elevator traction system operation data after data cleaning into a standard format to obtain a standardized elevator traction system operation data set;

[0022] The standardized elevator traction system operation dataset is divided into training set, validation set and test set.

[0023] Further, an elevator number is established for the elevators in the operation and maintenance area, the elevator traction system operation data collected in real time is matched with the elevator number, and stored in the corresponding storage space in the database, the maintenance history data of the elevators in the operation and maintenance area is retrieved, and data features are extracted from the maintenance history data of the elevators in the operation and maintenance area, and a data inspection form is created according to the data features of the maintenance history data of the elevators, and the corresponding feature data in the data inspection form is collected according to the preset time period according to the data inspection form, including;

[0024] Using sensors or data acquisition equipment installed on the elevator traction system, the elevator traction system operation data is collected in real time, and the collected elevator traction system operation data is matched with the corresponding elevator number;

[0025] Through database query statements, retrieve the maintenance history data of the specified elevator from the database;

[0026] Perform data analysis on historical maintenance data using a Weibull distribution model to extract key data features related to the remaining life prediction of the elevator traction system, including fault interval time, number of repairs, fault type, repair time, and abnormal data;

[0027] Based on the extracted key data features, a data inspection form is created to determine the data items and collection frequency to be collected regularly. The data inspection form includes the elevator number, inspection time, and key data feature value information.

[0028] Further, the corresponding characteristic data in the inspection form of the preset time period data collection is substituted into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the first elevator traction system, the elevator traction system operation data corresponding to the time when the elevator fails is retrieved, and the data is substituted into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the second elevator traction system, the remaining life prediction result of the first elevator traction system is compared with the remaining life prediction result of the second elevator traction system, if the data comparison results are consistent, the remaining life prediction model of the elevator traction system and the elevator traction system data monitoring are normal, and the remaining life prediction result of the third elevator traction system is output, if the data comparison results are inconsistent, the remaining life prediction model of the elevator traction system is optimized, and the corresponding data collection interface in the elevator traction system operation data is detected, including;

[0029] Clean, convert and standardize the corresponding characteristic data in the inspection form of the collected data in the preset time period. The data cleaning method includes processing missing values, abnormal values ​​and duplicate values, and standardizing the data;

[0030] Screen out features highly correlated with the remaining life of the elevator traction system from the preprocessed data. The features highly correlated with the remaining life of the elevator traction system include the failure interval time, the number of maintenance times, and the wear condition data of key components;

[0031] Substituting the selected characteristic data highly correlated with the remaining life of the elevator traction system into the remaining life prediction model of the elevator traction system, and using the established remaining life prediction model of the elevator traction system to process and analyze the data;

[0032] After the elevator traction system remaining life prediction model is run, the remaining life prediction result of the first elevator traction system will be output. The remaining life prediction result of the first elevator traction system includes a specific remaining life value of the traction system;

[0033] Retrieving the traction system operation data of the elevator when the failure occurs from the elevator operation and maintenance data database, including the speed, acceleration, temperature and pressure of the elevator when the failure occurs;

[0034] The traction system operation data when the elevator fails is substituted into the same elevator traction system remaining life prediction model, and the elevator traction system remaining life prediction model outputs the remaining life prediction result of the second elevator traction system.

[0035] Furthermore, the present invention provides a method for predicting the remaining life of an elevator traction system, further comprising:

[0036] The data storage space used to store the operation data of the elevator traction system is tested for remaining data storage space according to a preset time period. If the remaining data storage space in the test result is less than 1T data storage space, an insufficient data storage space warning message will be generated, and a communication connection will be established with a third-party cloud storage space interface. If the data storage space is full, the data will be automatically stored in the third-party cloud storage space;

[0037] Before storing the data in a third-party cloud storage space, the data to be stored will be encrypted by privacy computing. After the data is encrypted by privacy computing, it will be stored in the third-party cloud storage space.

[0038] The local storage space is detected in real time. If the capacity of the local storage space increases, the data stored in the third-party cloud storage space is retrieved and stored in the local storage space, and the data stored in the third-party cloud storage space is deleted simultaneously.

[0039] In a second aspect, the present invention provides a system for predicting the remaining life of an elevator traction system, comprising: a server end, an elevator equipment end, a background control end, and an operator equipment end;

[0040] The server end establishes a communication connection with the elevator equipment end, the background control end and the operator equipment end, and the server end is used to process the data fed back by the elevator equipment end, the background control end and the operator equipment, and transmit the data processing results to the background control end and the operator equipment end;

[0041] The server side includes a data acquisition unit, a model building unit, a data inspection unit, an elevator traction system life prediction unit and a prediction model optimization unit;

[0042] Data acquisition unit: acquires the elevator traction system operation data and stores it in a database for storing elevator operation and maintenance data, wherein the elevator traction system operation data includes the load data of the traction system, the elevator operation speed, acceleration data, operation time data, fault and abnormal condition records, use environment data, maintenance history data and elevator energy consumption data;

[0043] Model building unit: After data preprocessing of the elevator traction system operation data, the data set is divided into a training set, a validation set and a test set. The training set is used to train the model, the validation set is used to adjust the preset parameters and hyperparameters of the elevator traction system remaining life prediction model, and the test set is used to evaluate the final performance of the preset elevator traction system remaining life prediction model;

[0044] Data inspection unit: establish elevator numbers for elevators in the operation and maintenance area, match the real-time collected elevator traction system operation data with the elevator number, and store them in the corresponding storage space in the database, retrieve the maintenance history data of the elevators in the operation and maintenance area, extract data features from the maintenance history data of the elevators in the operation and maintenance area, create a data inspection form based on the data features of the maintenance history data of the elevators, and collect the corresponding feature data in the data inspection form according to the preset time period;

[0045] Elevator traction system life prediction unit: Substitute the corresponding characteristic data in the preset time period data collection inspection form into the elevator traction system remaining life prediction model to obtain the first elevator traction system remaining life prediction result, retrieve the elevator traction system operation data corresponding to the elevator failure time, and substitute the data into the elevator traction system remaining life prediction model to obtain the second elevator traction system remaining life prediction result, compare the first elevator traction system remaining life prediction result with the second elevator traction system remaining life prediction result, if the data comparison results are consistent, the elevator traction system remaining life prediction model and the elevator traction system data monitoring are normal, and output the third elevator traction system remaining life prediction result, if the data comparison results are inconsistent, optimize the elevator traction system remaining life prediction model, and detect the corresponding data acquisition interface in the elevator traction system operation data;

[0046] Prediction model optimization unit: extract data features from the remaining life prediction result of the third elevator traction system, obtain data feature information of the third prediction result, collect corresponding data in the real-time third prediction result data, compare the remaining life prediction result of the third elevator traction system with the actual data, check the degree of consistency between the remaining life prediction result of the third elevator traction system and the corresponding data in the collected real-time third prediction result data. If the prediction result is significantly different from the actual data, the remaining life prediction model of the elevator traction system needs to be adjusted and optimized.

[0047] The beneficial effects of the present invention are as follows: The present invention provides a method and system for predicting the remaining life of an elevator traction system. The method for predicting the remaining life of an elevator traction system integrates multiple links such as data collection, storage, preprocessing, model training and verification, and result comparison to form a complete and systematic prediction process. First, by comprehensively acquiring various operating data of the elevator traction system and establishing an effective data storage mechanism, a solid foundation is provided for subsequent data analysis and model training. The accuracy and generalization ability of model training are ensured by preprocessing and set division of data. In addition, the present application also introduces an elevator number and data matching mechanism to achieve refined data management.

[0048] In the model application stage, the remaining life prediction results of the first and second elevator traction systems were obtained by substituting the characteristic data, and the accuracy of the model was verified by comparing the two sets of results. This double verification mechanism not only improves the reliability of the prediction, but also provides a basis for the optimization of the model. Finally, by extracting the data features of the third prediction result and comparing it with the actual data, the accuracy of the prediction result is further ensured.

[0049] This application achieves accurate prediction of the remaining life of the elevator traction system by comprehensively using technical means such as data analysis, machine learning and real-time monitoring, providing strong technical support for the safe operation and preventive maintenance of elevators. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0051] Figure 1 A flow chart of a method for predicting the remaining life of an elevator traction system is provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described in conjunction with the specific embodiments of the present invention and the corresponding drawings. 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. The technical solutions provided by the embodiments of the present invention are described in detail below in conjunction with the drawings.

[0053] See also Figure 1 The present invention provides a method for predicting the remaining life of an elevator traction system, comprising:

[0054] In step S101, the elevator traction system operation data is obtained and stored in a database for storing elevator operation and maintenance data, wherein the elevator traction system operation data includes load data of the traction system, elevator operation speed, acceleration data, operation time data, fault and abnormal condition records, use environment data, maintenance history data and elevator energy consumption data;

[0055] In step S101, the system comprehensively collects the operation data of the elevator traction system, including but not limited to key information such as load, operating speed, acceleration, and operating time. In addition, faults and abnormal conditions, operating environment data (such as temperature, humidity, etc.), maintenance history data, and elevator energy consumption data will also be recorded. These data will be stored in a dedicated elevator operation and maintenance database for subsequent analysis and processing.

[0056] Comprehensive data collection provides a rich information foundation for subsequent data analysis and model training. Database storage makes data management and retrieval more efficient and facilitates long-term storage and analysis.

[0057] In step S102, after data preprocessing, the elevator traction system operation data is divided into a training set, a validation set and a test set. The training set is used to train the model, the validation set is used to adjust the preset elevator traction system remaining life prediction model parameters and hyperparameters, and the test set is used to evaluate the final performance of the preset elevator traction system remaining life prediction model;

[0058] The collected raw data will go through preprocessing steps such as cleaning, transformation and standardization to eliminate problems such as outliers, missing values ​​and duplicate values. The preprocessed data set will be divided into training set, validation set and test set. The training set is used to train the prediction model, the validation set is used to adjust the parameters and hyperparameters of the model, and the test set is used to evaluate the final performance of the model.

[0059] Data preprocessing improves the quality and consistency of data, making model training more accurate. Reasonable division of the data set helps to objectively evaluate the generalization ability of the model and prevent overfitting.

[0060] In step S103, an elevator number is established for the elevators in the operation and maintenance area, the elevator traction system operation data collected in real time is matched with the elevator number, and stored in the corresponding storage space in the database, the maintenance history data of the elevators in the operation and maintenance area is retrieved, and data features are extracted from the maintenance history data of the elevators in the operation and maintenance area. According to the data features of the maintenance history data of the elevators, a data inspection form is created, and the corresponding feature data in the data inspection form is collected according to the preset time period according to the data inspection form;

[0061] In step S103, a unique number is assigned to each elevator, and the real-time collected data is matched with this number to ensure the accuracy and traceability of the data. At the same time, the system will retrieve the historical maintenance data and extract key data features from it. Based on these features, a data inspection form is created, which will be used to guide subsequent data collection work.

[0062] Elevator numbering ensures accurate data matching and improves the efficiency and accuracy of data management. Data feature extraction helps identify key information in maintenance history and lays the foundation for subsequent prediction work.

[0063] In step S104, the corresponding characteristic data in the inspection form of the preset time period data collection is substituted into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the first elevator traction system, the elevator traction system operation data corresponding to the time when the elevator failure occurs is retrieved, and the data is substituted into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the second elevator traction system, and the remaining life prediction result of the first elevator traction system is compared with the remaining life prediction result of the second elevator traction system. If the data comparison results are consistent, the remaining life prediction model of the elevator traction system and the elevator traction system data monitoring are normal, and the remaining life prediction result of the third elevator traction system is output. If the data comparison results are inconsistent, the remaining life prediction model of the elevator traction system is optimized, and the corresponding data collection interface in the elevator traction system operation data is detected;

[0064] According to the preset time period, the system will collect the corresponding feature data in the data inspection form and substitute these data into the trained elevator traction system remaining life prediction model to obtain the first prediction result. At the same time, when the elevator fails, the system will retrieve the operating data at the time of the failure and substitute it into the model for a second prediction. The two prediction results will be compared to verify the accuracy of the model.

[0065] By comparing the two prediction results, possible problems in the model or data acquisition interface can be discovered in a timely manner. This double verification mechanism improves the reliability of the prediction and provides a strong guarantee for the safe operation and maintenance of the elevator.

[0066] In step S105, the data features in the remaining life prediction result of the third elevator traction system are extracted to obtain the data feature information of the third prediction result, the corresponding data in the real-time third prediction result data is collected, the remaining life prediction result of the third elevator traction system is compared with the actual data, and the degree of consistency between the remaining life prediction result of the third elevator traction system and the corresponding data in the collected real-time third prediction result data is checked. If the prediction result is significantly different from the actual data, the elevator traction system remaining life prediction model needs to be adjusted and optimized.

[0067] In step S105, the data features in the final prediction result are extracted and compared with the real-time collected data. If it is found that the prediction result is significantly different from the actual data, the prediction model needs to be adjusted and optimized.

[0068] By comparing with actual data, the prediction model can be continuously verified and optimized to improve its prediction accuracy. Continuous optimization of the model helps to more accurately predict the remaining life of the elevator traction system and provide decision support for preventive maintenance.

[0069] Specifically, the elevator traction system operation data is obtained, wherein the elevator traction system operation data includes the load data of the traction system, the elevator operation speed, acceleration data, operation time data, fault and abnormal situation records, use environment data, maintenance history data and elevator energy consumption data, including;

[0070] Determine the type of elevator traction system operating data and set the synchronization frequency;

[0071] The database architecture consists of a master database and a slave database. The master database is used to process write operations, and the slave database is used to process read operations and data backup.

[0072] The load data of the traction system, the elevator running speed, acceleration data, running time data, fault and abnormal condition records, use environment data, maintenance history data and elevator energy consumption data in the elevator traction system operation data are split into multiple elevator traction system data slices, and each elevator traction system data slice is stored on a different database node;

[0073] After each elevator traction system data shard is stored on different database nodes, the data is synchronously stored in the blockchain, and a corresponding hash value and timestamp are generated for each elevator traction system data shard stored on different database nodes;

[0074] If during the synchronization of the elevator traction system data, if a data conflict of the elevator traction system is encountered during the synchronization process, the conflicting data is distinguished according to the hash value and timestamp corresponding to the data. If the data is the same data, only the data label and the retrieval path of the data corresponding to the label are established.

[0075] This step solves the problem that although eddy current detection can be used for real-time monitoring, long-term continuous monitoring may generate a large amount of data, which puts higher requirements on data processing and analysis capabilities.

[0076] Specifically, after data preprocessing of the elevator traction system operation data, the data set is divided into a training set, a validation set and a test set. The training set is used to train the model, the validation set is used to adjust the preset parameters and hyperparameters of the elevator traction system remaining life prediction model, and the test set is used to evaluate the final performance of the preset elevator traction system remaining life prediction model, including;

[0077] Performing data cleaning on the elevator traction system operation data, removing abnormal values, noise data and duplicate data in the elevator traction system operation data, and obtaining the elevator traction system operation data after data cleaning;

[0078] Convert the elevator traction system operation data after data cleaning into a standard format to obtain a standardized elevator traction system operation data set;

[0079] The standardized elevator traction system operation dataset is divided into training set, validation set and test set.

[0080] Specifically, an elevator number is established for the elevators in the operation and maintenance area, the elevator traction system operation data collected in real time is matched with the elevator number, and stored in the corresponding storage space in the database, the maintenance history data of the elevators in the operation and maintenance area is retrieved, and data features are extracted from the maintenance history data of the elevators in the operation and maintenance area, and a data inspection form is created according to the data features of the maintenance history data of the elevators, and the corresponding feature data in the data inspection form is collected according to the preset time period according to the data inspection form, including;

[0081] Using sensors or data acquisition equipment installed on the elevator traction system, the elevator traction system operation data is collected in real time, and the collected elevator traction system operation data is matched with the corresponding elevator number;

[0082] Through database query statements, retrieve the maintenance history data of the specified elevator from the database;

[0083] Perform data analysis on historical maintenance data using a Weibull distribution model to extract key data features related to the remaining life prediction of the elevator traction system, including fault interval time, number of repairs, fault type, repair time, and abnormal data;

[0084] Based on the extracted key data features, a data inspection form is created to determine the data items and collection frequency to be collected regularly. The data inspection form includes the elevator number, inspection time, and key data feature value information.

[0085] Specifically, the corresponding characteristic data in the inspection form of the preset time period data collection is substituted into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the first elevator traction system, the elevator traction system operation data corresponding to the time when the elevator fails is retrieved, and the data is substituted into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the second elevator traction system, and the remaining life prediction result of the first elevator traction system is compared with the remaining life prediction result of the second elevator traction system. If the data comparison results are consistent, the remaining life prediction model of the elevator traction system and the elevator traction system data monitoring are normal, and the remaining life prediction result of the third elevator traction system is output. If the data comparison results are inconsistent, the remaining life prediction model of the elevator traction system is optimized, and the corresponding data collection interface in the elevator traction system operation data is detected, including;

[0086] Clean, convert and standardize the corresponding characteristic data in the inspection form of the collected data in the preset time period. The data cleaning method includes processing missing values, abnormal values ​​and duplicate values, and standardizing the data;

[0087] Screen out features highly correlated with the remaining life of the elevator traction system from the preprocessed data. The features highly correlated with the remaining life of the elevator traction system include the failure interval time, the number of maintenance times, and the wear condition data of key components;

[0088] Substituting the selected characteristic data highly correlated with the remaining life of the elevator traction system into the remaining life prediction model of the elevator traction system, and using the established remaining life prediction model of the elevator traction system to process and analyze the data;

[0089] After the elevator traction system remaining life prediction model is run, the remaining life prediction result of the first elevator traction system will be output. The remaining life prediction result of the first elevator traction system includes a specific remaining life value of the traction system;

[0090] Retrieving the traction system operation data of the elevator when the failure occurs from the elevator operation and maintenance data database, including the speed, acceleration, temperature and pressure of the elevator when the failure occurs;

[0091] The traction system operation data when the elevator fails is substituted into the same elevator traction system remaining life prediction model, and the elevator traction system remaining life prediction model outputs the remaining life prediction result of the second elevator traction system.

[0092] Specifically, the present invention provides a method for predicting the remaining life of an elevator traction system, further comprising:

[0093] The data storage space used to store the operation data of the elevator traction system is tested for remaining data storage space according to a preset time period. If the remaining data storage space in the test result is less than 1T data storage space, an insufficient data storage space warning message will be generated, and a communication connection will be established with a third-party cloud storage space interface. If the data storage space is full, the data will be automatically stored in the third-party cloud storage space;

[0094] Before storing the data in a third-party cloud storage space, the data to be stored will be encrypted by privacy computing. After the data is encrypted by privacy computing, it will be stored in the third-party cloud storage space.

[0095] The local storage space is detected in real time. If the capacity of the local storage space increases, the data stored in the third-party cloud storage space is retrieved and stored in the local storage space, and the data stored in the third-party cloud storage space is deleted simultaneously.

[0096] This step solves the problem that real-time monitoring also means that the device needs to run continuously, which may increase equipment wear and energy consumption, resulting in reduced monitoring accuracy of real-time monitoring equipment, by detecting local storage space and establishing a communication connection with cloud storage space.

[0097] In a second aspect, the present invention provides a system for predicting the remaining life of an elevator traction system, comprising: a server end, an elevator equipment end, a background control end, and an operator equipment end;

[0098] The server end establishes a communication connection with the elevator equipment end, the background control end and the operator equipment end, and the server end is used to process the data fed back by the elevator equipment end, the background control end and the operator equipment, and transmit the data processing results to the background control end and the operator equipment end;

[0099] The server side includes a data acquisition unit, a model building unit, a data inspection unit, an elevator traction system life prediction unit and a prediction model optimization unit;

[0100] Data acquisition unit: acquires the elevator traction system operation data and stores it in a database for storing elevator operation and maintenance data, wherein the elevator traction system operation data includes the load data of the traction system, the elevator operation speed, acceleration data, operation time data, fault and abnormal condition records, use environment data, maintenance history data and elevator energy consumption data;

[0101] Model building unit: After data preprocessing of the elevator traction system operation data, the data set is divided into a training set, a validation set and a test set. The training set is used to train the model, the validation set is used to adjust the preset parameters and hyperparameters of the elevator traction system remaining life prediction model, and the test set is used to evaluate the final performance of the preset elevator traction system remaining life prediction model;

[0102] Data inspection unit: establish elevator numbers for elevators in the operation and maintenance area, match the real-time collected elevator traction system operation data with the elevator number, and store them in the corresponding storage space in the database, retrieve the maintenance history data of the elevators in the operation and maintenance area, extract data features from the maintenance history data of the elevators in the operation and maintenance area, create a data inspection form based on the data features of the maintenance history data of the elevators, and collect the corresponding feature data in the data inspection form according to the preset time period;

[0103] Elevator traction system life prediction unit: Substitute the corresponding characteristic data in the preset time period data collection inspection form into the elevator traction system remaining life prediction model to obtain the first elevator traction system remaining life prediction result, retrieve the elevator traction system operation data corresponding to the elevator failure time, and substitute the data into the elevator traction system remaining life prediction model to obtain the second elevator traction system remaining life prediction result, compare the first elevator traction system remaining life prediction result with the second elevator traction system remaining life prediction result, if the data comparison results are consistent, the elevator traction system remaining life prediction model and the elevator traction system data monitoring are normal, and output the third elevator traction system remaining life prediction result, if the data comparison results are inconsistent, optimize the elevator traction system remaining life prediction model, and detect the corresponding data acquisition interface in the elevator traction system operation data;

[0104] Prediction model optimization unit: extract data features from the remaining life prediction result of the third elevator traction system, obtain data feature information of the third prediction result, collect corresponding data in the real-time third prediction result data, compare the remaining life prediction result of the third elevator traction system with the actual data, check the degree of consistency between the remaining life prediction result of the third elevator traction system and the corresponding data in the collected real-time third prediction result data. If the prediction result is significantly different from the actual data, the remaining life prediction model of the elevator traction system needs to be adjusted and optimized.

[0105] It can be seen from the above embodiments that the present invention provides a method and system for predicting the remaining life of an elevator traction system. The technical solution of this application first comprehensively collects various operating data of the elevator traction system, including load, speed, acceleration, running time, etc., and stores these data in a special elevator operation and maintenance database. The collected data is cleaned and preprocessed to eliminate outliers and noise, and then the data is divided into a training set, a validation set, and a test set, which are used for model training, parameter adjustment, and performance evaluation, respectively. A unique number is established for each elevator, and the real-time collected data is matched with the elevator number and stored in the corresponding database space. The key features in the maintenance history data are extracted, and a data inspection form is created based on these features, and these feature data are collected according to a preset time period. The collected feature data is substituted into the trained elevator traction system remaining life prediction model to obtain the first prediction result. When the elevator fails, the operating data at the time of the failure is retrieved and substituted into the model for a second prediction. Compare the first and second prediction results. If they are consistent, it means that the model and data monitoring are normal, and the final prediction results are output; if they are inconsistent, the model needs to be optimized and the data acquisition interface needs to be checked. Extract the data features in the final prediction results and compare them with the real-time collected data to verify the accuracy of the prediction. If it is found that the prediction results are significantly different from the actual data, the prediction model needs to be adjusted and optimized.

[0106] Users can effectively use this technical solution to predict the remaining life of the elevator traction system, timely discover potential safety hazards, and provide strong support for the operation, maintenance and management of the elevator. It solves the problem that the accuracy of the existing prediction of the remaining life of the elevator traction system is highly dependent on the quality of the prediction model and the information stored in the database. If the model is not accurate enough or the database information is not perfect, the prediction results may be biased.

[0107] Those skilled in the art can clearly understand that the technology in the embodiments of the present invention can be implemented by means of software plus a necessary general hardware platform. Based on this understanding, the technical solution in the embodiments of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., and includes a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention or some parts of the embodiments.

[0108] The above-described embodiments of the present invention do not limit the protection scope of the present invention.

Claims

1. A method for predicting the remaining life of an elevator traction system, characterized in that: include; Acquire the elevator traction system operation data and store it in a database for storing elevator operation and maintenance data, wherein the elevator traction system operation data includes the load data of the traction system, the elevator running speed, acceleration data, running time data, fault and abnormal situation records, use environment data, maintenance history data and elevator energy consumption data; After data preprocessing, the elevator traction system operation data is divided into a training set, a validation set and a test set. The training set is used to train the model, the validation set is used to adjust the preset elevator traction system remaining life prediction model parameters and hyperparameters, and the test set is used to evaluate the final performance of the preset elevator traction system remaining life prediction model; Establish elevator numbers for elevators in the operation and maintenance area, match the real-time collected elevator traction system operation data with the elevator number, and store them in the corresponding storage space in the database, retrieve the maintenance history data of the elevators in the operation and maintenance area, extract data features from the maintenance history data of the elevators in the operation and maintenance area, create a data inspection form based on the data features of the maintenance history data of the elevators, and collect the corresponding feature data in the data inspection form according to the preset time period based on the data inspection form; Substitute the corresponding characteristic data in the inspection form of the preset time period data collection into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the first elevator traction system, retrieve the elevator traction system operation data corresponding to the time when the elevator failure occurred, and substitute the data into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the second elevator traction system, compare the remaining life prediction result of the first elevator traction system with the remaining life prediction result of the second elevator traction system, if the data comparison results are consistent, the remaining life prediction model of the elevator traction system and the elevator traction system data monitoring are normal, and output the remaining life prediction result of the third elevator traction system, if the data comparison results are inconsistent, optimize the remaining life prediction model of the elevator traction system, and detect the corresponding data collection interface in the elevator traction system operation data; Extract the data features in the remaining life prediction result of the third elevator traction system, obtain the data feature information of the third prediction result, collect the corresponding data in the real-time third prediction result data, compare the remaining life prediction result of the third elevator traction system with the actual data, check the degree of consistency between the remaining life prediction result of the third elevator traction system and the corresponding data in the collected real-time third prediction result data, if the prediction result is significantly different from the actual data, it is necessary to adjust and optimize the remaining life prediction model of the elevator traction system.

2. The method according to claim 1, characterized in that Acquire the elevator traction system operation data, wherein the elevator traction system operation data includes the load data of the traction system, the elevator operation speed, acceleration data, operation time data, fault and abnormal condition records, use environment data, maintenance history data and elevator energy consumption data, including; Determine the type of elevator traction system operating data and set the synchronization frequency; The database architecture consists of a master database and a slave database. The master database is used to process write operations, and the slave database is used to process read operations and data backup. The load data of the traction system, the elevator running speed, acceleration data, running time data, fault and abnormal condition records, use environment data, maintenance history data and elevator energy consumption data in the elevator traction system operation data are split into multiple elevator traction system data slices, and each elevator traction system data slice is stored on a different database node; After each elevator traction system data shard is stored on different database nodes, the data is synchronously stored in the blockchain, and a corresponding hash value and timestamp are generated for each elevator traction system data shard stored on different database nodes; If during the synchronization of the elevator traction system data, if a data conflict of the elevator traction system is encountered during the synchronization process, the conflicting data is distinguished according to the hash value and timestamp corresponding to the data. If the data is the same data, only the data label and the retrieval path of the data corresponding to the label are established.

3. The method according to claim 1, characterized in that After data preprocessing of the elevator traction system operation data, the data set is divided into a training set, a validation set and a test set. The training set is used to train the model, the validation set is used to adjust the preset parameters and hyperparameters of the elevator traction system remaining life prediction model, and the test set is used to evaluate the final performance of the preset elevator traction system remaining life prediction model, including; Performing data cleaning on the elevator traction system operation data, removing abnormal values, noise data and duplicate data in the elevator traction system operation data, and obtaining the elevator traction system operation data after data cleaning; Convert the elevator traction system operation data after data cleaning into a standard format to obtain a standardized elevator traction system operation data set; The standardized elevator traction system operation dataset is divided into training set, validation set and test set.

4. The method according to claim 1, characterized in that Establishing elevator numbers for elevators in the operation and maintenance area, matching the real-time collected elevator traction system operation data with the elevator numbers, and storing them in the corresponding storage space in the database, retrieving the maintenance history data of the elevators in the operation and maintenance area, extracting data features from the maintenance history data of the elevators in the operation and maintenance area, creating a data inspection form based on the data features of the maintenance history data of the elevators, and collecting corresponding feature data in the data inspection form according to a preset time period, including; Using sensors or data acquisition equipment installed on the elevator traction system, the elevator traction system operation data is collected in real time, and the collected elevator traction system operation data is matched with the corresponding elevator number; Through database query statements, retrieve the maintenance history data of the specified elevator from the database; Perform data analysis on the maintenance history data through the Weibull distribution model to extract key data features related to the remaining life prediction of the elevator traction system, including the fault interval time, maintenance times, fault type, maintenance time, and abnormal data; Based on the extracted key data features, a data inspection form is created to determine the data items and collection frequency to be collected regularly. The data inspection form includes the elevator number, inspection time, and key data feature value information.

5. The method according to claim 1, characterized in that Substituting the corresponding characteristic data in the inspection form of the data collected in the preset time period into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the first elevator traction system, retrieving the elevator traction system operation data corresponding to the time when the elevator failure occurred, and substituting the data into the remaining life prediction model of the elevator traction system to obtain the remaining life prediction result of the second elevator traction system, comparing the remaining life prediction result of the first elevator traction system with the remaining life prediction result of the second elevator traction system, if the data comparison results are consistent, the remaining life prediction model of the elevator traction system and the elevator traction system data monitoring are normal, and outputting the remaining life prediction result of the third elevator traction system, if the data comparison results are inconsistent, optimizing the remaining life prediction model of the elevator traction system, and detecting the corresponding data acquisition interface in the elevator traction system operation data, including; Clean, convert and standardize the corresponding characteristic data in the inspection form of the collected data in the preset time period. The data cleaning method includes processing missing values, abnormal values ​​and duplicate values, and standardizing the data; Features highly correlated with the remaining life of the elevator traction system are screened out from the preprocessed data. The features highly correlated with the remaining life of the elevator traction system include fault interval time, maintenance times, and wear condition data of key components; the screened feature data highly correlated with the remaining life of the elevator traction system are substituted into the remaining life prediction model of the elevator traction system, and the data are processed and analyzed using the established remaining life prediction model of the elevator traction system; After the elevator traction system remaining life prediction model is run, the remaining life prediction result of the first elevator traction system will be output. The remaining life prediction result of the first elevator traction system includes a specific remaining life value of the traction system; Retrieving the traction system operation data of the elevator when the failure occurs from the elevator operation and maintenance data database, including the speed, acceleration, temperature and pressure of the elevator when the failure occurs; The traction system operation data when the elevator fails is substituted into the same elevator traction system remaining life prediction model, and the elevator traction system remaining life prediction model outputs the remaining life prediction result of the second elevator traction system.

6. The method according to claim 2, characterized in that Also includes; The data storage space used to store the operation data of the elevator traction system is tested for remaining data storage space according to a preset time period. If the remaining data storage space in the test result is less than 1T data storage space, an insufficient data storage space warning message will be generated, and a communication connection will be established with a third-party cloud storage space interface. If the data storage space is full, the data will be automatically stored in the third-party cloud storage space; Before storing the data in a third-party cloud storage space, the data to be stored will be encrypted by privacy computing. After the data is encrypted by privacy computing, it will be stored in the third-party cloud storage space. The local storage space is detected in real time. If the capacity of the local storage space increases, the data stored in the third-party cloud storage space is retrieved and stored in the local storage space, and the data stored in the third-party cloud storage space is deleted simultaneously.

7. An elevator traction system remaining life prediction system, characterized in that: Including; server side, elevator equipment side, background control side and operator equipment side; The server end establishes a communication connection with the elevator equipment end, the background control end and the operator equipment end, and the server end is used to process the data fed back by the elevator equipment end, the background control end and the operator equipment, and transmit the data processing results to the background control end and the operator equipment end; The server side includes a data acquisition unit, a model building unit, a data inspection unit, an elevator traction system life prediction unit and a prediction model optimization unit; Data acquisition unit: acquires the elevator traction system operation data and stores it in a database for storing elevator operation and maintenance data, wherein the elevator traction system operation data includes the load data of the traction system, the elevator operation speed, acceleration data, operation time data, fault and abnormal condition records, use environment data, maintenance history data and elevator energy consumption data; Model building unit: After data preprocessing of the elevator traction system operation data, the data set is divided into a training set, a validation set and a test set. The training set is used to train the model, the validation set is used to adjust the preset parameters and hyperparameters of the elevator traction system remaining life prediction model, and the test set is used to evaluate the final performance of the preset elevator traction system remaining life prediction model; Data inspection unit: establish elevator numbers for elevators in the operation and maintenance area, match the real-time collected elevator traction system operation data with the elevator number, and store them in the corresponding storage space in the database, retrieve the maintenance history data of the elevators in the operation and maintenance area, extract data features from the maintenance history data of the elevators in the operation and maintenance area, create a data inspection form based on the data features of the maintenance history data of the elevators, and collect the corresponding feature data in the data inspection form according to the preset time period; Elevator traction system life prediction unit: Substitute the corresponding characteristic data in the preset time period data collection inspection form into the elevator traction system remaining life prediction model to obtain the first elevator traction system remaining life prediction result, retrieve the elevator traction system operation data corresponding to the elevator failure time, and substitute the data into the elevator traction system remaining life prediction model to obtain the second elevator traction system remaining life prediction result, compare the first elevator traction system remaining life prediction result with the second elevator traction system remaining life prediction result, if the data comparison results are consistent, the elevator traction system remaining life prediction model and the elevator traction system data monitoring are normal, and output the third elevator traction system remaining life prediction result, if the data comparison results are inconsistent, optimize the elevator traction system remaining life prediction model, and detect the corresponding data acquisition interface in the elevator traction system operation data; Prediction model optimization unit: extract data features from the remaining life prediction result of the third elevator traction system, obtain data feature information of the third prediction result, collect corresponding data in the real-time third prediction result data, compare the remaining life prediction result of the third elevator traction system with the actual data, check the degree of consistency between the remaining life prediction result of the third elevator traction system and the corresponding data in the collected real-time third prediction result data. If the prediction result is significantly different from the actual data, the remaining life prediction model of the elevator traction system needs to be adjusted and optimized.

Images