Elevator health degree assessment method based on big data analysis

Abstract

The invention discloses an elevator health degree assessment method based on big data analysis, which comprises the following steps: acquiring elevator historical maintenance data and elevator operation data, then preprocessing, selecting elevator maintenance key features from the elevator historical maintenance data and the elevator operation data, and endowing the importance degree of elevator maintenance with a weight value according to the features; performing statistical feature extraction on key features in the elevator operation data in each maintenance period; according to historical maintenance data, weighted summation is conducted on maintenance item results in the same maintenance period, elevator health degree scores are obtained and classified, classification results serve as training labels of the maintenance period, and a training data set is obtained through gathering; constructing and training an elevator health degree evaluation model based on a decision tree algorithm; and the operation data collected in real time is preprocessed and then input into the trained elevator health degree evaluation model, and the health degree of the elevator is evaluated and predicted. The accuracy, the real-time performance and the intelligent level of elevator health degree evaluation are improved.

Description

Technical Field

[0001] The present invention relates to the field of elevator health assessment, and in particular to an elevator health assessment method based on big data analysis. Background Art

[0002] As an indispensable means of transportation in modern cities, the safety and reliability of elevators are vital to people's lives. In order to ensure the safe operation of elevators, maintenance is an indispensable link. At present, the elevator health assessment mainly adopts methods based on maintenance history and statistical analysis, which evaluates the health status of elevators by analyzing the maintenance records and fault data of elevators. However, this method has the problems of limited data volume and low evaluation accuracy. In order to improve the accuracy and precision of the evaluation, methods based on big data and machine learning have been widely used in the field of elevator health assessment. However, traditional big data and machine learning methods still face problems such as complex data processing and model construction, and lack of real-time performance. Summary of the invention

[0003] In view of the deficiencies of the prior art, the present invention proposes an elevator health assessment method based on big data analysis.

[0004] The specific technical solutions are as follows:

[0005] An elevator health assessment method based on big data analysis, comprising:

[0006] S1: Obtain the historical elevator maintenance data and elevator operation data, and pre-process them respectively;

[0007] S2: Select key features of elevator maintenance from the preprocessed historical elevator maintenance data and elevator operation data, and assign weight values ​​to the importance of elevator maintenance according to the features;

[0008] S3: Build a training dataset, which is implemented through the following sub-steps:

[0009] (3.1) Perform statistical feature extraction on key features of elevator operation data in each maintenance cycle;

[0010] (3.2) Based on the historical maintenance data, the results of each maintenance project in the same maintenance cycle are weighted and summed to obtain the elevator health score; the elevator health score is classified according to the elevator health score, and the classification result is used as the training label of the maintenance cycle; the statistical features of all maintenance cycles are arranged correspondingly with the training labels to obtain the training data set;

[0011] S4: Build and train an elevator health assessment model based on a decision tree algorithm;

[0012] S5: After the elevator historical maintenance data and the real-time collected elevator operation data are processed by S1 and S2, they are input into the trained elevator health assessment model to evaluate and predict the health of the elevator.

[0013] Furthermore, the elevator historical maintenance data includes: the elevator machine room environment, operating device status, drive host vibration, brake status, encoder status.

[0014] Furthermore, the elevator operation data includes: elevator door opening and closing status, elevator operation speed, elevator operation acceleration, elevator operation temperature, elevator operation load, and elevator operation amplitude.

[0015] Furthermore, the preprocessing includes: data cleaning, missing value processing and outlier processing.

[0016] Furthermore, in the step (3.1), for the elevator door opening and closing status belonging to binary data, the number of door opening and closing times is counted in each maintenance cycle; for continuous data such as elevator running speed, elevator running acceleration, and elevator running temperature, mathematical statistical features are calculated in each maintenance cycle, and the mathematical statistical features include: maximum value, minimum value, average value, and variance; all the obtained mathematical statistical features are spliced ​​to obtain the statistical feature data of the maintenance cycle.

[0017] Furthermore, the elevator health assessment model includes: an input unit, a decision unit, and an output unit;

[0018] The input unit is used to receive the training data set as an input of the elevator health assessment model;

[0019] The decision unit is used to make a decision classification on the health status of the elevator according to the input training data set; the decision unit is implemented based on a decision tree algorithm, and is used to capture the interactive relationship between data to achieve elevator health classification;

[0020] The output unit is used to output the elevator health classification result.

[0021] Furthermore, in S4, training the elevator health assessment model includes the following steps:

[0022] (4.1) Data preparation: Divide the training data set into a training set and a test set;

[0023] (4.2) Model training: The elevator health assessment model is trained using the training set. During the training process, the elevator health assessment model divides the data according to different branches of features and judgment conditions to minimize prediction errors or maximize information gain; the training is iterated repeatedly and the structure and parameters of the decision tree are adjusted;

[0024] (4.3) Model evaluation and tuning: Use the data from the test set to evaluate the trained elevator health assessment model and calculate the various indicators of the model on the test set, including accuracy, precision, recall, and F1 score. Based on the evaluation results, adjust the parameters and structure of the decision tree to tune and improve the model. Finally, a trained elevator health assessment model is obtained.

[0025] Furthermore, in the step (4.3), if the elevator health assessment model performs well on the training set but poorly on the test set, a regularization or pruning method is used to reduce the model complexity and avoid overfitting;

[0026] Adjust the parameters and structure of the decision tree, including: adjusting the maximum depth of the decision tree, the conditions for node splitting, and the minimum number of samples for leaf nodes;

[0027] The cross-validation method is used to evaluate the performance and stability of the model to avoid the accidental results of a single evaluation.

[0028] The beneficial effects of the present invention are:

[0029] The elevator health assessment method based on the integration of maintenance history, statistical analysis, big data and machine learning proposed in the present invention can make full use of maintenance history data, statistical analysis, big data processing and machine learning technology, improve the accuracy, real-time and intelligence level of elevator health assessment, and provide a scientific basis for elevator maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flow chart of an elevator health assessment method based on big data analysis in an embodiment of the present invention. DETAILED DESCRIPTION

[0031] The present invention will be described in detail below according to the accompanying drawings and preferred embodiments, and the purpose and effect of the present invention will become more clear. The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0032] like Figure 1 As shown, a method for evaluating elevator health based on big data analysis includes the following steps:

[0033] S1. Elevator data collection and preprocessing: obtain elevator data, including elevator historical maintenance data and elevator operation data; and preprocess the obtained elevator data.

[0034] The elevator's historical maintenance data includes: the elevator's machine room environment, operating device status, drive host vibration, brake status, encoder status, etc.

[0035] Elevator operation data refers to the data of elevator operation obtained through sensors and monitoring equipment installed in the elevator, including: elevator door opening and closing status, elevator operating speed, elevator operating acceleration, elevator operating temperature, elevator operating load, elevator operating amplitude, etc.

[0036] Preprocessing includes: data cleaning, missing value processing and outlier processing.

[0037] In this embodiment, the data in the quarterly maintenance list of an elevator company is taken as an example, and the maintenance items are used as feature data. The results of each item are divided into three situations: normal (meets the requirements, and the test and experimental functions are normal), abnormal (found to be non-compliant, and needs to be adjusted or replaced), and none (the elevator does not have this item or function). According to the needs of elevator maintenance, the feature value of the feature data with abnormal results is scored as 1, and the feature value of the feature data with normal or none results is scored as 0. That is, if it reaches 1 point, it means that the item needs maintenance.

[0038] S2. Feature selection and weight assignment: According to the needs, select the key features of elevator maintenance from the preprocessed elevator historical maintenance data and elevator operation data, and assign weight values ​​according to the importance of the features to elevator maintenance.

[0039] In this embodiment, 45 key features of elevator maintenance are extracted, as shown in Table 1. In Table 1, the requirements for judging the "machine room and pulley room environment" numbered 1 as normal are cleanliness, intact doors and windows, and normal lighting; the requirements for the "manual emergency operation device status" numbered 2 are that the device is complete and located at the specified position; the requirements for the "drive host status" numbered 3 are that there is no abnormal vibration and abnormal sound during operation; the requirements for the "brake clearance status" numbered 5 are that the brake lining and the brake wheel should not rub when opened, and the clearance value meets the requirements of the manufacturing unit; the "brake self-monitoring status" numbered 6 specifically refers to the self-monitoring status of the brake when it is used as the stopping subsystem of the car accidental movement protection device; The requirement for "encoder status" No. 7 is to be clean and firmly installed; the requirement for "oil cup status on guide shoe" No. 13 specifically refers to whether the oil amount in the oil cup is appropriate and whether there is any leakage; the requirement for "engagement length of floor door locking element" No. 29 is to be no less than 7mm; the requirement for "pit environment status" No. 30 is to be clean, free of water seepage and accumulation, and normal lighting; the requirement for "reducer lubricating oil status" No. 33 is to have an appropriate amount of lubricating oil and no leakage; the requirement for "traction wheel groove and suspension device cleanliness, wire rope status" No. 37 is to be free of severe grease, have uniform tension, and meet the requirements of the manufacturing unit.

[0040] Table 1 Elevator maintenance feature number table

[0041]

[0042] According to the importance of these key features to elevator maintenance, these key features are divided into three levels: A, B, and C. The weights are assigned to Class A features as 0.2, Class B features as 0.5, and Class C features as 1. The classification is shown in Table 2.

[0043] Table 2 Elevator maintenance feature classification and weight assignment table

[0044]

[0045] S3: Build a training dataset, which is implemented through the following sub-steps:

[0046] (3.1) The elevator operation data is grouped based on the maintenance cycle and the mathematical statistical features of the elevator operation data in each maintenance cycle are extracted. Specifically, for the elevator door opening and closing status, which is binary data, the number of door opening and closing times is counted in each maintenance cycle; for continuous data such as elevator operation speed, elevator operation acceleration, and elevator operation temperature, the maximum value, minimum value, average value, variance and other mathematical statistical features are calculated in each maintenance cycle. All the obtained mathematical statistical features are spliced ​​to obtain the statistical feature data (i.e., training data) of the maintenance cycle.

[0047] (3.2) Score the health of the elevator based on historical maintenance data and construct training labels for the maintenance cycle, as follows:

[0048] (3.2.1) Elevator health score: The results of each maintenance item in the same maintenance cycle are weighted and summed to obtain the elevator health score, which is expressed as follows:

[0049]

[0050] In the formula, i represents the i-th item in the maintenance order, i.e., the number, m is the total number of items in the maintenance order, and w is i is the weight value of maintenance item i, x i is the result of maintenance item i, that is, abnormal score is 1, and normal score or no such item is 0.

[0051] The higher the final elevator health score is, the greater the risks and problems that exist in the elevator.

[0052] (3.2.2) Elevator health classification is performed based on the elevator health score, and the obtained health classification results will be used as training labels for this maintenance cycle.

[0053] In this embodiment, the elevator health classification is specifically performed according to the method shown in Table 3 below.

[0054] Table 3 Elevator health scores and health classification correspondence

[0055] Elevator Health Score Elevator health classification 0-10 points normal 11-25 points Need attention 26-35 points warn 36-45 points Fault

[0056] (3.3) Arrange the training data (i.e., statistical features) of all maintenance cycles corresponding to the training labels, and finally obtain the training data set.

[0057] S4. Build and train an elevator health assessment model: Build an elevator health assessment model based on a decision tree algorithm, and through learning from a large amount of elevator operation data, establish the relationship between elevator health and the historical maintenance data and elevator operation data of the maintenance elevator, and finally obtain a trained elevator health assessment model.

[0058] The elevator health assessment model includes: input unit, decision unit and output unit.

[0059] The input unit is used to receive the training dataset constructed by S3 as model input.

[0060] The decision unit is used to make decisions and classify the health status of the elevator according to the input training data set; specifically, the decision unit is implemented based on a decision tree algorithm to capture the interactive relationship between data, thereby achieving the elevator health classification effect.

[0061] The output unit is used to output the elevator health classification result.

[0062] After the elevator health assessment model is built, data can be input to train the model. The specific training steps are as follows:

[0063] (4.1) Data preparation: Divide the training data set into a training set and a test set. Usually 70% of the data is used as the training set and 30% of the data is used as the test set.

[0064] (4.2) Model training: The elevator health assessment model is trained using the statistical feature data of the training set and the corresponding training labels. During the training process, the elevator health assessment model will divide the data according to the different branches of the features and the judgment conditions to minimize the prediction error or maximize the information gain. The accuracy and generalization ability of the model are improved by repeatedly iterating the training and adjusting the structure and parameters of the decision tree.

[0065] (4.3) Model evaluation and tuning: Use the test set data to evaluate the performance and prediction accuracy of the trained elevator health assessment model. Based on the evaluation results, tune and improve the model, such as adjusting the depth of the decision tree, pruning operations, etc., to improve the performance, accuracy and robustness of the model. The following are the specific steps:

[0066] Model evaluation: Use the data from the test set to evaluate the model, calculate the accuracy, precision, recall, F1 score and other indicators of the model on the test set, evaluate the performance and prediction accuracy of the model, and check whether the model has overfitting or underfitting problems. If the model performs well on the training set but poorly on the test set, there may be an overfitting problem, and you need to consider regularization or pruning to reduce the complexity of the model. Analyze the types and causes of model errors, and understand the prediction effects of the model on samples of different categories in order to improve and tune the model.

[0067] Model tuning: Based on the results and analysis of model evaluation, the model is tuned for existing problems. Tuning includes adjusting the parameters and structure of the decision tree, such as adjusting the maximum depth of the decision tree, the judgment conditions for node splitting, the minimum number of samples of leaf nodes, etc., to optimize the performance and generalization ability of the model. Feature selection and feature engineering optimization can also be performed. Based on the S2 selection, features with higher importance and contribution to health evaluation can be screened out to further improve the prediction ability of the model.

[0068] Cross-validation: Use the cross-validation method to further evaluate the performance and stability of the model. Divide the training set into multiple subsets, use each subset as the validation set in turn, and the remaining subsets as the training set, perform multiple rounds of training and validation, and obtain the average evaluation results. Cross-validation can help detect the generalization ability and stability of the model and avoid the randomness of a single evaluation result.

[0069] Iterative optimization: Based on the evaluation and tuning results, iterative optimization is continuously performed until the error is within the threshold range, that is, satisfactory model performance is achieved. You can also try to use different decision tree algorithms and parameter combinations, compare their performance and effects, and select the optimal model for further optimization and deployment.

[0070] Through the above steps, we can comprehensively evaluate and tune the elevator health assessment model to improve the performance and accuracy of the model. The continuous iterative optimization process will make the model more adaptable to the health assessment needs of the elevator, provide more reliable and accurate health status judgments, and provide a scientific decision-making basis for elevator maintenance.

[0071] S5. Health assessment and prediction: The historical elevator maintenance data and real-time elevator operation data are used as the elevator data to be evaluated. After processing, they are input into the trained elevator health assessment model for calculation and analysis to obtain the elevator health assessment results, and the future health status of the elevator is predicted based on the assessment results.

[0072] In S4, a decision tree-based elevator health assessment model was established and evaluated and tuned. In S5, how to use the model for online health assessment and prediction is introduced in detail. The following are the detailed steps:

[0073] (5.1) Data preparation: Obtain the elevator data to be evaluated, including the parameter values ​​of each feature. Process the data as in S1 and S2 to ensure the accuracy and completeness of the data.

[0074] (5.2) Data processing: The elevator operation data to be evaluated is processed as in step (3.1) to obtain the statistical characteristics of the elevator operation data within the period to be evaluated.

[0075] (5.3) Health assessment and prediction: Input the statistical features obtained in step (5.2) into the trained elevator health assessment model, and predict the health status of the elevator according to the branching rules and decision paths of the model. The model will classify the elevator into different health levels based on the conditional judgment and branching rules of the features.

[0076] (5.4) Result interpretation and report generation: Generate a corresponding explanatory report for each elevator health assessment result. The report should clearly explain the elevator health status, possible problems and risks, and provide targeted maintenance recommendations and measures.

[0077] Furthermore, continuous monitoring and updating are carried out: regular health assessments are carried out on elevators to monitor changes in elevator status and health. Based on new data and feedback information, the elevator health assessment model is continuously updated and optimized to maintain the accuracy and adaptability of the model.

[0078] Through the above steps, the trained elevator health assessment model can be used to evaluate and predict the health of the elevator data to be evaluated. Through result interpretation and report generation, the health status of the elevator can be clearly understood, and corresponding maintenance suggestions and measures can be provided. At the same time, the model is continuously monitored and updated to ensure the accuracy of real-time evaluation and prediction of the elevator health status.

[0079] In practical applications, after completing the above health assessment, targeted maintenance strategies can also be generated based on the results of the health assessment. According to the health status and prediction results of the elevator, a corresponding maintenance plan is formulated, including maintenance frequency, maintenance measures, etc. The maintenance plan is optimized and adjusted through optimization algorithms, such as genetic algorithms and simulated annealing algorithms, to achieve a reasonable allocation of maintenance resources and optimization of maintenance effects. Specifically, it includes:

[0080] (1) Maintenance needs analysis: Based on the health assessment results, analyze the maintenance needs of each elevator, including identifying elevators that require repair, maintenance, or other maintenance operations, and determining their priority and urgency.

[0081] (2) Maintenance plan generation: Generate a maintenance plan based on the elevator health assessment results and maintenance needs analysis. Determine the maintenance frequency, maintenance content and maintenance measures for each elevator. Maintenance measures include routine maintenance, fault repair, parts replacement, etc.

[0082] (3) Maintenance resource optimization: Consider the limitations of maintenance resources and optimize the maintenance plan. Adjust the arrangement and priority of the maintenance plan based on factors such as the availability and cost of maintenance personnel, tools and equipment to maximize maintenance efficiency and resource utilization.

[0083] (4) Maintenance execution and monitoring: Perform maintenance operations according to the maintenance plan generated in step (3). Monitor the execution of the maintenance process and record the maintenance results and maintenance data.

[0084] (5) Maintenance feedback and optimization: Collect feedback information during the maintenance process, including maintenance records, fault repair status, maintenance costs, etc. Analyze feedback information, evaluate maintenance results, and identify potential problems and room for improvement. Optimize and adjust maintenance strategies based on feedback information to improve maintenance quality and efficiency.

[0085] Furthermore, we continuously improve the maintenance plan: regularly evaluate the effectiveness and feasibility of the maintenance strategy, and make continuous improvements based on actual conditions. Based on data such as elevator health assessment, maintenance execution, and maintenance feedback, we continuously optimize the maintenance strategy and process to adapt to changing needs and environments.

[0086] Through the above steps, the maintenance strategy of the elevator can be generated and optimized to improve maintenance efficiency and reduce costs. Maintenance demand analysis and maintenance plan generation ensure the pertinence and standardization of maintenance work, maintenance resource optimization improves resource utilization efficiency, maintenance execution and monitoring ensure the implementation and supervision of maintenance operations, and maintenance feedback, optimization and continuous improvement ensure the continuous optimization and adaptability of maintenance strategies. By continuously improving and optimizing maintenance strategies, the reliability, safety and availability of elevators can be improved, providing users with a better elevator experience.

[0087] Those skilled in the art can understand that the above are only preferred examples of the invention and are not intended to limit the invention. Although the invention is described in detail with reference to the above examples, those skilled in the art can still modify the technical solutions recorded in the above examples or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, etc. made within the spirit and principle of the invention shall be included in the protection scope of the invention.

Claims

1. An elevator health assessment method based on big data analysis, characterized in that: include: S1: Obtain the historical elevator maintenance data and elevator operation data, and pre-process them respectively; S2: Select key features of elevator maintenance from the preprocessed historical elevator maintenance data and elevator operation data, and assign weight values ​​to the importance of elevator maintenance according to the features; S3: Build a training dataset, which is implemented through the following sub-steps: (3.1) Perform statistical feature extraction on key features of elevator operation data in each maintenance cycle; (3.2) Based on the historical maintenance data, the results of each maintenance project in the same maintenance cycle are weighted and summed to obtain the elevator health score; Elevator health is classified according to the elevator health score, and the classification result is used as the training label of the maintenance cycle; Arrange the statistical features of all maintenance cycles and the training labels accordingly to obtain the training data set; S4: Build and train an elevator health assessment model based on a decision tree algorithm; S5: After the elevator historical maintenance data and the real-time collected elevator operation data are processed by S1 and S2, they are input into the trained elevator health assessment model to evaluate and predict the health of the elevator.

2. The elevator health assessment method based on big data analysis according to claim 1 is characterized in that: The elevator historical maintenance data includes: elevator machine room environment, operating device status, drive host vibration, brake status, encoder status.

3. The elevator health assessment method based on big data analysis according to claim 1 is characterized in that: The elevator operation data includes: elevator door opening and closing status, elevator operation speed, elevator operation acceleration, elevator operation temperature, elevator operation load, and elevator operation amplitude.

4. The elevator health assessment method based on big data analysis according to claim 1 is characterized in that: The preprocessing includes: data cleaning, missing value processing and outlier processing.

5. The elevator health assessment method based on big data analysis according to claim 3 is characterized in that: In the step (3.1), for the elevator door opening and closing status belonging to binary data, the number of door opening and closing times is counted in each maintenance cycle; For continuous data such as elevator running speed, elevator running acceleration, and elevator running temperature, mathematical statistical characteristics are calculated in each maintenance cycle. The mathematical statistical characteristics include: maximum value, minimum value, average value, and variance; all the obtained mathematical statistical characteristics are spliced ​​to obtain the statistical characteristic data of the maintenance cycle.

6. The elevator health assessment method based on big data analysis according to claim 1 is characterized in that: The elevator health assessment model includes: an input unit, a decision unit, and an output unit; The input unit is used to receive the training data set as an input of the elevator health assessment model; The decision unit is used to make a decision classification on the health status of the elevator according to the input training data set; the decision unit is implemented based on a decision tree algorithm, and is used to capture the interactive relationship between data to achieve elevator health classification; The output unit is used to output the elevator health classification result.

7. The elevator health assessment method based on big data analysis according to claim 1 is characterized in that: In S4, training the elevator health assessment model includes the following steps: (4.1) Data preparation: Divide the training data set into a training set and a test set; (4.2) Model training: The elevator health assessment model is trained using the training set. During the training process, the elevator health assessment model divides the data according to different branches of features and judgment conditions to minimize prediction errors or maximize information gain; the training is iterated repeatedly and the structure and parameters of the decision tree are adjusted; (4.3) Model evaluation and tuning: Use the data from the test set to evaluate the trained elevator health assessment model and calculate the various indicators of the model on the test set, including accuracy, precision, recall, and F1 score. Based on the evaluation results, adjust the parameters and structure of the decision tree to tune and improve the model. Finally, a trained elevator health assessment model is obtained.

8. The elevator health assessment method based on big data analysis according to claim 7 is characterized in that: In the step (4.3), if the elevator health assessment model performs well on the training set but poorly on the test set, regularization or pruning methods are used to reduce the model complexity and avoid overfitting; Adjust the parameters and structure of the decision tree, including: adjusting the maximum depth of the decision tree, the conditions for node splitting, and the minimum number of samples for leaf nodes; The cross-validation method is used to evaluate the performance and stability of the model to avoid the accidental results of a single evaluation.

Images