A crane operation and maintenance method based on machine learning, an electronic device, and a computer-readable storage medium

Abstract

The application provides a crane operation and maintenance method based on machine learning, which comprises the following steps: installing sensor groups on a metal structure, an electrical control element and a driving mechanism of a crane to monitor the intrinsic safety of the crane, and forming data groups of key positions; monitoring the real-time use safety of the state of the crane in the use process to form use process data groups; performing data processing on the key position data groups and the use process data groups to extract feature data, using a neural network to build a machine learning model and performing model training; collecting real-time data to input the model for judgment; when the prediction probability is higher than a certain threshold, issuing a fault alarm, and at the same time, using a convolutional neural network to classify the fault alarm, and performing maintenance work according to different types of fault alarms. The application solves the problem that the existing technology cannot effectively predict equipment failure in the operation and maintenance management process. An electronic device and a computer readable storage medium are also provided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of crane operation and maintenance management, and particularly relates to a crane operation and maintenance method based on machine learning, an electronic device and a computer readable storage medium. BACKGROUND

[0002] During the operation and use of the crane, the working environment thereof can be relatively harsh, and the traditional operation and maintenance mode can have certain human subjectivity and timeliness, which can easily cause resource waste. In addition, due to the need for high-altitude operation, frequent operation and maintenance plans can increase the risk of high-altitude operation of maintenance personnel, affect the point inspection efficiency and coverage, and manual inspection is difficult to achieve in some important security positions. Therefore, the interval of operation and maintenance also needs to be appropriately lengthened, which increases the equipment failure rate and safety hazards. During the operation and maintenance interval, if the equipment failure condition cannot be effectively predicted, the use of the equipment can be affected, the equipment can be used for a long time in a "sick operation" state, and thus the equipment failure shutdown can be caused, the production downtime can be long, and the loss can be large. Therefore, it is necessary to improve the efficiency and accuracy of crane operation and maintenance management. SUMMARY

[0003] In view of the deficiencies in the prior art, the present application provides a crane operation and maintenance method based on machine learning, which solves the problem that the equipment failure condition cannot be effectively predicted in the operation and maintenance management process in the prior art.

[0004] According to an embodiment of the present application, a crane operation and maintenance method based on machine learning comprises:

[0005] Sensors are installed on the metal structure, electrical control elements and driving mechanisms of the crane to monitor the intrinsic safety of the crane and form data groups of key positions;

[0006] The state during the use of the crane is monitored in real time to form a use process data group;

[0007] The data groups of key positions and the use process data group are processed to extract feature data and use a neural network to build a machine learning model and perform model training;

[0008] Real-time data is collected and input into the model for judgment, and when the prediction probability is higher than a certain threshold, a failure alarm is issued, and a convolutional neural network is used to classify the failure alarm, and maintenance work is performed according to different types of failure alarms.

[0009] In the embodiment, through the training of the machine learning model, the model can perform more accurate failure alarm and classification of the failure alarm, targeted operation and maintenance is performed according to different types of failure alarms, the operation and maintenance efficiency and accuracy are improved, and the "sick operation" of the equipment is avoided.

[0010] Further, the data processing includes data cleaning of the data sets, which includes one or more of filling in missing values, smoothing noisy data, identifying or deleting outliers.

[0011] Further, the data processing further includes adjusting the cleaned data into a format suitable for machine learning algorithms, combining and uniformly storing data in multiple data sets, establishing a data warehouse, and performing data integration.

[0012] Further, the data processing further includes:

[0013] extracting original data samples from the data warehouse, and performing feature extraction on the original data samples as original signal features;

[0014] extracting fault features from a public crane fault database, including time domain features, frequency domain features, and time-frequency domain features, for fault diagnosis.

[0015] Further, the data processing further includes:

[0016] filtering the extracted features, and calculating as follows:

[0017] S1: calculating the internal distance of the x features of the class y using the formula represented in the equation,

[0018]

[0019] ,

[0020] where N represents the number of samples, F represents the number of features, and the number of classes is represented by C, and are the x features of the C samples of the yth class and the x features of the N samples of the yth class, respectively;

[0021] S2: calculating the average value of the internal distance of the x features of the y class, represented by the equation:

[0022] ,

[0023] S3: calculating the average value of the x features of the N samples in the y class, represented by the equation:

[0024] ,

[0025] S4: calculating the average value of the inter-distance of the C classes of the x features using the equation:

[0026]

[0027]

[0028] S5: using the equation to calculate the evaluation factor of x features:

[0029] ,

[0030] Features are sorted according to the value of x, and then the number of features is increased.

[0031] Further, the parameters of the convolutional neural network for fault type classification include: accuracy, sensitivity, specificity, precision value and cross entropy.

[0032] Further, the model for judging includes:

[0033] The trained model is deployed to the actual crane, real-time monitoring of sensor data, and using the model for fault diagnosis;

[0034] According to the predicted situation, the fault probability is judged, if the fault probability is higher than the set threshold for many times, the fault alarm is triggered, and the operation and maintenance management system is accessed to dispatch operation and maintenance tasks.

[0035] Further, the operation and maintenance management system includes use safety management, intrinsic safety management, and operation and maintenance process management; the use safety management is used for supervising the use safety monitoring data of real-time monitoring, and also includes the monitoring of the safety of the outside of the running track, to realize display, alarm, and abnormal data output; the intrinsic safety management is used for supervising the monitoring situation of key components; the operation and maintenance process management is used for publishing the repair plan, repair task, and standard situation module information of the predicted equipment situation to relevant personnel.

[0036] According to the embodiment, an electronic device is also provided, which includes: one or more processors; a memory; and one or more application programs, wherein the one or more application programs are stored in the memory and configured to be executed by the one or more processors, and the one or more application programs are configured to: execute the above-mentioned deep learning-based high-altitude operation dangerous area detection method.

[0037] According to the embodiment, a computer readable storage medium is also provided, which includes: a computer program stored therein, which can be loaded and executed by a processor to execute the above-mentioned deep learning-based high-altitude operation dangerous area detection method.

[0038] Compared with the prior art, the present application has the following beneficial effects:

[0039] The efficiency and accuracy of operation and maintenance management can be improved. Through training of a machine learning model, the model can perform more accurate fault alarm and classification of fault alarm, and targeted operation and maintenance management is performed according to different types of fault alarm, thereby solving the problem that the existing technology cannot effectively predict equipment failure in the operation and maintenance management process, and avoiding the situation that the equipment is operated with disease. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 The figure is a schematic diagram of the operation and maintenance management system of the embodiment of the present application. DETAILED DESCRIPTION

[0041] The technical solutions in the present application will be further described below in combination with the drawings and embodiments.

[0042] In an exemplary embodiment, as shown in Figure 1 The present embodiment provides a crane operation and maintenance method based on machine learning, which comprises:

[0043] Sensors are installed on the metal structure, electrical control elements and driving mechanism of the crane to monitor the intrinsic safety of the crane, and form data groups of key parts;

[0044] The state during use of the crane is monitored in real time to form a use process data group;

[0045] The key part data group and the use process data group are processed to extract feature data, and a neural network is used to build a machine learning model and perform model training;

[0046] Real-time data is collected and input into the model for judgment. When the prediction probability is higher than a certain threshold, a fault alarm is issued, and a convolutional neural network is used to classify the fault alarm, and maintenance work is performed according to different types of fault alarm.

[0047] Specifically, the operation and maintenance management system comprises use safety management, intrinsic safety management and operation and maintenance process management. The use safety management is used to supervise the use safety monitoring data of real-time monitoring, and also comprises monitoring of the safety of the surrounding environment during operation, and realizes display, alarm and abnormal data output. The intrinsic safety management is used to supervise the monitoring of key components. The operation and maintenance process management is used to publish the repair plan, repair task and standard condition module information of the predicted equipment condition to relevant personnel.

[0048] As shown in Figure 1As shown, in a further exemplary scheme, intrinsic safety management collects data from the crane's steel structure, electrical components, wire ropes, brakes, reducers, wheels, motors, etc., and transmits the collected data to the system. The system provides functions such as display, storage, and query for the data, and pushes abnormal items that have been judged as faults to the main system interface for display.

[0049] like Figure 1 As shown, in a further exemplary scheme, safety management is used to monitor the use of the crane by means of video surveillance, thermal imaging monitoring, usage status, load status, wind speed status, track status, wheel pressure status, etc. The monitoring data also supports functions such as display, storage, and query.

[0050] like Figure 1 As shown in the example, in a further exemplary solution, the operation and maintenance process management includes modules such as maintenance plans, maintenance tasks, maintenance management, and standards management; it formulates periodic inspection and maintenance plans to ensure that all systems and equipment are inspected and maintained as necessary within the predetermined time, while using historical data and predictive analysis to optimize maintenance plans and reduce unexpected downtime; it clarifies the specific content, responsible person, required tools and materials for each maintenance task, tracks the progress of maintenance tasks, and ensures timely completion and compliance with quality standards; it establishes a fault response mechanism to quickly respond to system or equipment faults, and records the maintenance process and results, including the cause of the fault, maintenance measures, and subsequent preventive measures.

[0051] In a further exemplary scheme, data processing includes data cleaning of the data set, which includes filling in missing values, smoothing noisy data, identifying or deleting outliers, or one or more of the following: data processing also includes adjusting the cleaned data into a format suitable for machine learning algorithms through smoothing aggregation, data generalization, normalization, etc., combining data from multiple data sets and storing them uniformly, establishing a data warehouse, and performing data integration.

[0052] In a further exemplary embodiment, data processing further includes:

[0053] Extract raw data samples from the data warehouse, perform feature extraction on the raw data samples, and use them as raw signal features;

[0054] Fault features, including time-domain features, frequency-domain features, and time-frequency-domain features, are extracted from publicly available crane fault databases for fault diagnosis.

[0055] In a further exemplary embodiment, data processing further includes:

[0056] The extracted features are filtered, and the calculations are as follows:

[0057] S1: Calculate the internal distance of the feature x of the class = y using the formula represented in the equation,

[0058]

[0059] ,

[0060] where N represents the number of samples, F represents the number of features, and the number of classes is represented by C, and are the x features of the C samples of the yth class and the x features of the N samples of the yth class, respectively;

[0061] S2: Calculate the average of the internal distance of the x features of the y class, represented by the equation:

[0062] ,

[0063] S3: Calculate the average of the x features of the N samples in the y class, represented by the equation:

[0064] ,

[0065] S4: Calculate the average of the inter-distance of the C classes of the x features using the equation:

[0066]

[0067]

[0068] S5: Calculate the evaluation factor of the x features using the equation:

[0069] ,

[0070] The features are sorted according to the value of x, and then the number of features is increased.

[0071] In a further exemplary scheme, when the convolutional neural network classifies, the initial network layer is responsible for perceiving information from the fault, the convolutional filter is further responsible for finding fault features that can identify and distinguish faults with different faults, and the output layer provides classification results obtained through feature information obtained at the fully connected layer; The parameters of the convolutional neural network in the fault type classification include: accuracy, sensitivity, specificity, precision value and cross entropy; wherein,

[0072] Accuracy (ACC): It defines the percentage of accurate correction results obtained in the total number of events;

[0073] Sensitivity (SEN.): The percentage of actual cases with error symptoms in the cases of error categories accurately indicated by the sensitivity;

[0074] Specificity (SEP.): The specificity provides a higher probability of truly indicating a faultless class without giving false positive results;

[0075] Precision value: This value indicates the accurate results correctly or precisely obtained from the total number of positive predictions;

[0076] Cross-entropy: The cross-entropy loss can provide the correct classification ability of the neural network in the range of 0 to 1.

[0077] In a further exemplary scheme, the model making a judgment comprises:

[0078] The trained model is deployed to an actual crane, real-time monitoring of sensor data is performed, and the model is used for fault diagnosis;

[0079] According to the predicted situation, the fault probability is judged, if the fault probability is higher than the set threshold for multiple times, a fault alarm is triggered, and an operation and maintenance management system is accessed to dispatch operation and maintenance tasks.

[0080] Finally, it should be pointed out that the above embodiments are only used to illustrate the technical solutions of the present application and not to limit it. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the purpose and scope of the present application, and they should be covered in the scope of the claims of the present application.

Claims

1. A machine learning-based crane operation and maintenance method, characterized by, Comprise: Install sensor groups on the metal structure, electrical control elements, and driving mechanisms of the crane to monitor the intrinsic safety of the crane, forming data groups of key positions; Real-time monitoring of the state during the use of the crane to monitor the use safety, forming use process data groups; Data processing of the key position data groups and the use process data groups to extract feature data and use neural networks to build a machine learning model and perform model training; Collecting real-time data to input the model for judgment, issuing a fault alarm when the prediction probability is higher than a certain threshold, and classifying the fault alarm using a convolutional neural network to perform maintenance work according to different types of fault alarms; the data processing includes data cleaning of the data groups, which includes one or more of filling in missing values, smoothing noisy data, and identifying or deleting outliers; the data processing also includes adjusting the cleaned data into a format suitable for machine learning algorithms, combining and uniformly storing the data in multiple data groups, establishing a data warehouse, and performing data integration; The data processing also includes: Extracting original data samples from the data warehouse and extracting feature data from the original data samples as original signal features; Extracting fault features from a public crane fault database, including time domain features, frequency domain features, and time-frequency domain features, for fault diagnosis; the data processing also includes: Filtering the extracted features and calculating as follows: S1: Calculate the internal distance of the x feature of class y using the formula in the equation, , where N represents the number of samples, F represents the number of features, and the number of classes is represented by C, and are the x features of the C samples of the yth class and the x features of the N samples of the yth class, respectively. S2: Calculate the average value of the internal distance of the x feature of class y using the equation: ， S3: Calculate the average value of the x feature of N samples in class y using the equation: ， S4: Calculate the average value of the inter-distance of class C of the x feature using the equation: S5: Calculate the evaluation factor of the x feature using the equation: ， The features are sorted according to the value of x, and then the number of features is increased.

2. The machine learning based crane operation method of claim 1, wherein, The parameters of the convolutional neural network for fault type classification include accuracy, sensitivity, specificity, precision value, and cross-entropy.

3. The machine learning based crane operation method of claim 2, wherein, The model judgment includes: Deploying the trained model to the actual crane, monitoring the sensor data in real time, and using the model for fault diagnosis; According to the prediction, judge the fault probability, if the fault probability is higher than the set threshold for many times, trigger the fault alarm, access the operation and maintenance management system to assign operation and maintenance tasks.

4. The machine learning based crane operation method of claim 3, wherein, The operation and maintenance management system includes use safety management, intrinsic safety management, and operation and maintenance process management; the use safety management is used to supervise the use safety monitoring data of real-time monitoring, and also includes monitoring of the safety of the crane operation periphery, realizing display, alarm, and abnormal data output; the intrinsic safety management is used to supervise the monitoring of key components; the operation and maintenance process management is used to publish repair plans, repair tasks, and standard condition module information to relevant personnel according to the predicted equipment condition.

5. An electronic device, comprising: Comprise: One or more processors; Memory; one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured for performing the machine learning based crane operation method of any one of claims 1-4.

6. A computer-readable storage medium, characterized in that, comprising: a computer program stored in the memory and loadable and executable by the processor, the computer program comprising instructions for performing the machine learning based crane operation method of any one of claims 1-4.

Images