Intelligent dust remover fault and health state diagnosis management method, device and equipment and storage medium

Abstract

The invention provides an intelligent dust remover fault and health state diagnosis management method, device and equipment and a storage medium, relates to the field of dust remover management, and aims to improve the timeliness and accuracy of bag-type dust remover maintenance. The method comprises the following steps: acquiring real-time operation data of key equipment of the dust remover, wherein the real-time operation data at least comprises dust removal fan vibration data, dust removal fan rotating speed, dust removal fan current, dust removal motor vibration data, dust removal motor cooling water flow and dust removal motor current; dividing the real-time operation data to obtain first operation data and second operation data; inputting the first operation data into a trained health state evaluation model to obtain a health index; inputting the second operation data into a trained equipment fault classification model to obtain a classification result; and when the health index is lower than a first threshold value, alarm information is sent out, and fault processing information is generated according to a classification result.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of dust collector management, and specifically, to an intelligent dust collector fault and health status diagnosis and management method, device, equipment and storage medium. Background Art

[0002] As an important environmental protection equipment in industrial production, the stable operation of bag dust collector is crucial to reducing air pollutant emissions, protecting the environment and maintaining the continuous operation of the production line. However, in the actual operation process, key sub-equipment of bag dust collector, such as dust removal fan, dust removal motor, etc., often have various faults due to long-term operation, wear and other factors, resulting in reduced dust removal efficiency, increased energy consumption, and even production interruption.

[0003] Traditional bag filter maintenance methods mostly rely on manual inspections and regular maintenance, which has many shortcomings. On the one hand, manual inspections are difficult to achieve real-time monitoring, and are often discovered only when a fault has already occurred or has had a significant impact on system performance, resulting in untimely fault handling, increased maintenance costs and production losses. On the other hand, although regular maintenance can prevent the occurrence of faults to a certain extent, it often lacks pertinence and accuracy, and is prone to over-maintenance or under-maintenance, which is also not conducive to the long-term stable operation of the system. Summary of the invention

[0004] The embodiments of the present application provide an intelligent dust collector fault and health status diagnosis and management method, device, equipment and storage medium, aiming to solve the problem of how to achieve real-time monitoring and health prediction of the operating status of key sub-equipment of bag dust collector, timely discover and identify equipment abnormalities, and improve the timeliness and accuracy of equipment maintenance.

[0005] The first aspect of the embodiment of the present application provides an intelligent dust collector fault and health status diagnosis and management method, including: Acquire real-time operation data of key equipment of the dust collector, the real-time operation data at least including: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, dust removal motor cooling water flow, dust removal motor current; The real-time operation data is divided to obtain first operation data and second operation data, wherein the first operation data includes: dust removal fan vibration data, dust removal motor vibration data, and dust removal motor current; and the second operation data includes: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, and dust removal motor cooling water flow; The first operating data is input into a trained health status assessment model to obtain a health index, wherein the health status assessment model is used to determine an abnormal state that is significantly different from a normal state according to the regularity of each indicator data under normal operation of the dust collector, and to determine the health status of the dust collector according to the difference in the abnormal state; Inputting the second operation data into a trained equipment fault classification model to obtain a classification result; When the health index is lower than a first threshold, an alarm message is issued, and fault handling information is generated according to the classification result.

[0006] In an optional implementation, the training step of the health status assessment model includes: Obtain the normal historical first operation data of the key equipment of the dust collector; The normal historical first operation data is input into the VAE-LSTM model for training, the potential features of the normal historical first operation data are learned, and the trained health status assessment model is obtained, and the output data of the health status assessment model is the health index.

[0007] In an optional implementation, the training step of the equipment fault classification model includes: Obtain the historical second operation data of key equipment of dust collector; Preprocessing the historical second operation data to obtain processed historical time series data; Performing cluster analysis on the processed historical time series data to obtain training data with labels, wherein the labels include: normal labels and abnormal labels; The training data is input into a plurality of models for training, and the model with the best performance is selected as the equipment fault classification model. The output data of the equipment fault classification model is the classification result.

[0008] In an optional implementation, preprocessing the historical second operation data to obtain processed historical time series data includes: Data interpolation and outlier removal are performed on the historical second operation data to obtain processed historical time series data.

[0009] In an optional implementation, cluster analysis is performed on the processed historical time series data to obtain labeled training data, including: The processed historical time series data is clustered using the K-means clustering algorithm, and the optimal number of clusters is determined based on the silhouette coefficient of the clustering result. Each cluster is used as a category or label. According to the number of clusters, the processed historical time series data is clustered again using the K-means clustering algorithm to obtain the training data with labels.

[0010] In an optional embodiment, the method further includes: The obtained health index and the classification results are visualized on an industrial Internet of Things platform, and the industrial Internet of Things platform is used to collect real-time operating data of key equipment of the dust collector and perform statistical analysis on the real-time operating data.

[0011] In an optional embodiment, the method further includes: Obtain compressed air property data of dust collector; According to the compressed air property data, the exhaust gas dust concentration data is predicted through a pre-trained energy consumption model; the energy consumption model training learns the relationship between the compressed air property and the exhaust gas dust concentration; When the exhaust gas dust concentration data does not meet the industrial emission standard, adjustment information for the exhaust gas dust emission data is generated according to the exhaust gas dust concentration data.

[0012] A second aspect of the embodiment of the present application provides an intelligent dust collector fault and health status diagnosis and management device, comprising: An acquisition module is used to acquire real-time operation data of key equipment of the dust collector, and the real-time operation data includes at least: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, dust removal motor cooling water flow, and dust removal motor current; A classification module, used for dividing the real-time operation data to obtain first operation data and second operation data, wherein the first operation data includes: dust removal fan vibration data, dust removal motor vibration data, and dust removal motor current; and the second operation data includes: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, and dust removal motor cooling water flow; A health status assessment module, used for inputting the first operation data into a trained health status assessment model to obtain a health index, wherein the health status assessment model is used for judging an abnormal state that is significantly different from a normal state according to the regularity presented by each indicator data under the normal operation state of the dust collector, and judging the health state of the dust collector according to the difference in the abnormal state; A fault classification module, used for inputting the second operation data into a trained equipment fault classification model to obtain a classification result; The processing module is used to issue an alarm message when the health index is lower than a first threshold value, and generate fault processing information according to the classification result.

[0013] The third aspect of an embodiment of the present application provides an electronic device, comprising: a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the intelligent dust collector fault and health status diagnosis and management method of the first aspect of an embodiment of the present application.

[0014] The fourth aspect of the embodiments of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the intelligent dust collector fault and health status diagnosis and management method of the first aspect of the embodiments of the present application is implemented.

[0015] The intelligent dust collector fault and health status diagnosis and management method provided by the present application obtains the real-time operation data of the key equipment of the dust collector, and the real-time operation data at least includes: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, dust removal motor cooling water flow, dust removal motor current; divide the real-time operation data to obtain the first operation data and the second operation data; input the first operation data into the trained health status assessment model to obtain the health index, and the health status assessment model is used to judge the abnormal state with a large difference from the normal state according to the law presented by each indicator data under the normal operation state of the dust collector, and judge the health state of the dust collector by the difference in abnormal state; input the second operation data into the trained equipment fault classification model to obtain the classification result; when the health index is lower than the first threshold, an alarm message is issued, and fault handling information is generated according to the classification result. Through the health status assessment model, the health status of the dust collector can be monitored in real time, and abnormal changes can be discovered in time; at the same time, the abnormal data or fault type of the dust collector can be accurately identified through the equipment fault classification model, which provides specific guidance for subsequent fault handling and maintenance, improves the timeliness and accuracy of equipment maintenance, and avoids further development of faults leading to greater losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments of the present application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0017] Figure 1 It is a flowchart of the steps of the intelligent dust collector fault and health status diagnosis and management method proposed in one embodiment of the present application; Figure 2 It is a schematic diagram of the interface of the industrial Internet of Things platform of the intelligent dust collector fault and health status diagnosis and management method proposed in one embodiment of the present application; Figure 3 It is a schematic diagram comparing the prediction results of the energy consumption model of the intelligent dust collector fault and health status diagnosis and management method proposed in one embodiment of the present application with the real-time data; Figure 4 It is a schematic diagram of the network structure of the VAE-LSTM model of the intelligent dust collector fault and health status diagnosis and management method proposed in one embodiment of the present application; Figure 5 It is a schematic diagram of the silhouette coefficient of the clustering result of the intelligent dust collector fault and health status diagnosis and management method proposed in one embodiment of the present application; Figure 6 It is a schematic diagram of an electronic device proposed in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0019] Reference Figure 1 , Figure 1 This is a flowchart of the steps of the intelligent dust collector fault and health status diagnosis and management method proposed in one embodiment of the present application. Figure 1 As shown, the method includes the following steps S11 to S15: Step S11: Acquire real-time operation data of key equipment of the dust collector, wherein the real-time operation data at least includes: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, dust removal motor cooling water flow, and dust removal motor current.

[0020] Install sensors such as vibration sensors, speed sensors, current sensors, etc. on the key equipment of the dust collector (such as dust removal fans and dust removal motors). Collect data from these sensors in real time through data acquisition systems (such as PLC and SCADA systems) to obtain real-time operation data of the key equipment of the dust collector. Real-time operation data at least includes: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, dust removal motor cooling water flow, and dust removal motor current. By collecting real-time operation data of key equipment of the dust collector, data support is provided for subsequent fault warning and diagnosis.

[0021] Step S12: Divide the real-time operation data to obtain first operation data and second operation data, wherein the first operation data includes: dust removal fan vibration data, dust removal motor vibration data, and dust removal motor current; the second operation data includes: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, and dust removal motor cooling water flow.

[0022] The real-time operation data is divided into the first operation data and the second operation data. The first operation data is used to evaluate the health status of the equipment, and may include dust removal fan vibration data, dust removal motor vibration data, dust removal motor current, etc. The second operation data contains more comprehensive information for classified diagnosis of equipment failures, and may include dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, dust removal motor cooling water flow, etc. Through data division, data can be used more specifically to improve the accuracy and efficiency of subsequent model analysis.

[0023] Step S13: Input the first operating data into a trained health status assessment model to obtain a health index. The health status assessment model is used to determine an abnormal state that is significantly different from the normal state based on the regularity presented by each indicator data under the normal operating state of the dust collector, and determine the health status of the dust collector based on the difference in the abnormal state.

[0024] The first operation data is input into the trained health status assessment model. The model calculates the health index of the dust collector based on the input data. The index reflects the degree of difference between the current state of the dust collector and the normal state. The health status assessment model is trained using the healthy historical first operation data through the VAE-LSTM (Variational Autoencoder-Long Short-Term Memory) model. The health status assessment model can judge the abnormal state that is significantly different from the normal state based on the law presented by each indicator data under the normal operation state of the dust collector, and judge the health status of the dust collector based on the difference in abnormal state. Through the health status assessment model, the health status of the dust collector can be monitored in real time, abnormal changes can be discovered in time, and a basis can be provided for subsequent fault warning and maintenance decisions.

[0025] Step S14: input the second operation data into the trained equipment fault classification model to obtain a classification result.

[0026] The second operation data is input into the trained equipment fault classification model, and the model outputs the classification result of the dust collector based on the input data. The classification result is determined by the different types of samples of the training model, and the classification result can be normal data or abnormal data, or a specific fault type. Through the equipment fault classification model, the abnormal data or fault type of the dust collector can be accurately identified, providing specific guidance for subsequent fault handling and maintenance.

[0027] Step S15: When the health index is lower than a first threshold, an alarm message is issued, and fault handling information is generated according to the classification result.

[0028] According to the actual situation of the dust collector and operation and maintenance experience, a reasonable health index threshold, namely the first threshold, is set. When the health index is lower than the first threshold, the alarm mechanism is triggered, an alarm message is sent to relevant personnel, and fault handling information is generated based on the classification results. Through the generation of alarm information and fault handling information, relevant personnel can be reminded in time to pay attention to abnormal conditions of the dust collector, and take effective measures to handle and repair the fault, so as to avoid further development of the fault and cause greater losses.

[0029] The intelligent dust collector fault and health status diagnosis and management method provided by the present application obtains the real-time operation data of the key equipment of the dust collector, and the real-time operation data at least includes: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, dust removal motor cooling water flow, dust removal motor current; divide the real-time operation data to obtain the first operation data and the second operation data; input the first operation data into the trained health status assessment model to obtain the health index, and the health status assessment model is used to judge the abnormal state with a large difference from the normal state according to the law presented by each indicator data under the normal operation state of the dust collector, and judge the health state of the dust collector by the difference in abnormal state; input the second operation data into the trained equipment fault classification model to obtain the classification result; when the health index is lower than the first threshold, an alarm message is issued, and fault handling information is generated according to the classification result. Through the health status assessment model, the health status of the dust collector can be monitored in real time, and abnormal changes can be discovered in time; at the same time, the abnormal data or fault type of the dust collector can be accurately identified through the equipment fault classification model, which provides specific guidance for subsequent fault handling and maintenance, improves the timeliness and accuracy of equipment maintenance, and avoids further development of faults leading to greater losses.

[0030] In industrial production, dust collectors are important environmental protection equipment, and their operating status directly affects the efficiency and environmental compliance of the production line. Therefore, it is crucial to monitor and analyze the real-time operating data of key dust collector equipment and to promptly discover and handle potential faults in order to ensure the stable operation of the production line. This application divides the operating data of the dust removal fan and dust removal motor of the dust collector obtained in real time into: first operating data for health status assessment and second operating data for fault classification. The first operating data and the second operating data are respectively input into the health status assessment model and the equipment fault classification model to obtain the health index and classification results, so as to promptly discover and handle the fault risk. For example, in the production line of a certain enterprise, if the health status assessment model finds that the health index of a dust collector is lower than 0.2, which seriously exceeds the normal range, an alarm message is immediately issued; at the same time, the equipment fault classification model classifies the dust collector and obtains abnormal data, so that users can quickly find problems and take effective treatment measures.

[0031] In combination with the above embodiments, in one implementation, the present invention further provides an intelligent dust collector fault and health status diagnosis and management method. In this embodiment, the method further includes step S16 and step S17: Step S16: Visually display the obtained health index and the classification result on the industrial Internet of Things platform, where the industrial Internet of Things platform is used to collect real-time operating data of key equipment of the dust collector and perform statistical analysis on the real-time operating data.

[0032] like Figure 2 As shown, the health index and classification results are visualized on the industrial Internet of Things platform. The industrial Internet of Things platform has the functions of data collection, cleaning, processing and visualization. It can be used to collect real-time operation data of key equipment of the dust collector and perform statistical analysis on the real-time operation data so that managers can monitor and analyze the operation status of the dust collector in real time. Data visualization helps managers to intuitively understand the health status of the equipment and the results of fault classification, thereby improving management efficiency; the platform centrally collects and analyzes data, providing comprehensive data support for the operation and maintenance of the dust collector.

[0033] Step S17: Obtaining compressed air property data of the dust collector; According to the compressed air property data, the exhaust gas dust concentration data is predicted through a pre-trained energy consumption model; the energy consumption model training learns the relationship between the compressed air property and the exhaust gas dust concentration; When the exhaust gas dust concentration data does not meet the industrial emission standard, adjustment information for the exhaust gas dust emission data is generated according to the exhaust gas dust concentration data.

[0034] By installing sensors in the dust collector, the property data of compressed air (such as pressure, flow, temperature, and humidity) can be collected in real time. The exhaust gas dust concentration data can be predicted based on the property data of compressed air through the trained energy consumption model. The energy consumption model establishes the relationship between the equipment compressed air properties and the exhaust gas dust concentration based on the neural network. The predicted exhaust gas dust concentration data is compared with the industrial emission standards to determine whether it exceeds the standard. If it exceeds the standard, adjustment information is generated; if it does not exceed the standard, the data is recorded and monitoring continues. Environmental protection supervision data support can be provided to ensure that the dust removal effect meets environmental protection requirements, and it is convenient to adjust the property value of the equipment compressed air according to the feedback of the exhaust gas dust concentration data to achieve the effect of reducing energy consumption.

[0035] Among them, the training steps of the energy consumption model include: Obtain historical compressed air property data and historical exhaust gas dust concentration data for multiple dust collectors; The historical compressed air property data is used as input data, and the historical exhaust gas dust concentration data is used as output data. The historical compressed air property data is trained through an initial model to obtain the trained energy consumption model.

[0036] The historical compressed air property data and historical exhaust dust concentration data of multiple dust collectors can be directly obtained from the production center. The construction of the initial model involves LSTM recurrent neural network and fully connected layers.

[0037] The fully connected layer can be used in classification and regression tasks. It usually plays the role of a "classifier" in the entire convolutional neural network and is often located at the end of the neural network. The neural network maps the original data to the hidden feature space, and the fully connected layer maps the learned "feature representation" to the sample label space. x1, x2, x3 are the inputs of the fully connected layer, and a1, a2, a3 are the outputs. The forward calculation of the fully connected layer is:

[0038] Each neuron in the fully connected layer is connected to all neurons in the previous layer. Through this connection method, the feature representation is mapped to the sample label space, thereby achieving the effect of classification and regression.

[0039] The model network consists of four parts: input layer, intermediate LSTM layer, fully connected layer and output layer. The input layer is the compressed air attribute, and the output is the predicted exhaust dust concentration. The intermediate LSTM layer can effectively capture the temporal nature of the equipment attribute data, and the fully connected layer maps the learned features to the target sample space. The input compressed air attributes learn the corresponding features through the intermediate LSTM layer, and then the learned features are mapped to the target sample space through the fully connected layer, thereby establishing the relationship between the compressed air attributes and the exhaust dust concentration. Based on experimental experience and multiple attempts, the number of LSTM layers, the number of neurons, the time step (time-step), the batch size (batch_size) and other hyperparameters are determined, and the determined network parameters are shown in Table 1.

[0040] Table 1. Network parameters

[0041] Divide the historical compressed air property data and historical exhaust dust concentration data into training sets, validation sets, and test sets for model training, validation, and testing. Use the training set to train the initial model, and adjust the model parameters through the back-propagation algorithm to minimize the loss value. Use the validation set data to evaluate the performance of the model, including prediction accuracy, generalization ability, etc., and adjust the network structure or hyperparameters based on the validation results to improve model performance. Use the test set data to evaluate the final performance of the model, and further adjust and optimize the model's network structure, parameters, and hyperparameters based on the evaluation results. Apply the trained energy consumption model to the production environment, and the exhaust dust concentration can be predicted based on the compressed air properties. The prediction results are shown in Figure 3 . Figure 3 The data of exhaust gas dust concentration is shown in the figure. Red represents the predicted results of the energy consumption model, and blue represents real-time data. Judging from the results in the figure and the experimental indicators, the established energy consumption model can meet the corresponding expectations and effectively establish the relationship between the compressed air properties of the equipment and the exhaust gas dust concentration.

[0042] In combination with the above embodiments, in one implementation, the present invention further provides an intelligent dust collector fault and health status diagnosis and management method. In this embodiment, the training step of the health status assessment model includes steps S21-S22: Step S21: Acquire normal historical first operation data of key equipment of the dust collector; The normal first historical operation data of the key equipment of the dust collector is derived from the data acquisition system. The first historical operation data includes: dust removal fan vibration data, dust removal motor vibration data, and dust removal motor current. The dust removal fan vibration data includes the fan front shaft x-item vibration data, the fan front shaft y-item vibration data, the fan rear shaft x-item vibration data, and the fan rear shaft y-item vibration data; the dust removal motor vibration data includes the motor front shaft vibration data, the motor rear shaft vibration data, the motor A-item vibration data, the motor B-item vibration data, and the motor C-item vibration data.

[0043] Step S22: Input the normal historical first operation data into the VAE-LSTM model for training, learn the potential features of the normal historical first operation data, and obtain the trained health status assessment model, and the output data of the health status assessment model is the health index.

[0044] like Figure 4 As shown in the figure, the model network of the VAE-LSTM model consists of an input layer, an output layer, an intermediate LSTM network layer, and a hidden layer. The intermediate LSTM network layer consists of 4 layers of LSTM units. The data is mapped to the hidden layer through the input layer and the LSTM network, and a new representation of the data is learned. This process is the encoding process. Similarly, the data is mapped to the original representation through the hidden layer and the LSTM network. This process is the decoding process. After the encoding and decoding process, the model can learn the characteristics of normal state data. When data with a particularly large difference appears, it means that the data is very different from the normal state data, so it can be judged that the data is abnormal data.

[0045] The normal historical first run data is divided into training set, validation set and test set. The VAE-LSTM model is trained using the training set data. During the training process, the model parameters are adjusted through the back propagation algorithm to minimize the loss function. The loss function usually consists of two parts: reconstruction loss and KL divergence loss. The reconstruction loss measures the difference between the reconstructed data and the original data, and the KL divergence loss measures the difference between the learned latent variable distribution and the prior distribution. Use the validation set data to evaluate the performance of the model and adjust the hyperparameters such as the learning rate and batch size as needed. Use the test set data to evaluate the performance of the trained model, calculate the accuracy and stability of the HI value, and the model's ability to detect abnormal data. According to the evaluation results, the network structure, parameters and hyperparameters of the model are further adjusted and optimized. Hyperparameter optimization techniques such as grid search, random search or Bayesian optimization can be used to find the best parameter combination.

[0046] The parameters of each layer of the network are adjusted accordingly. The determined network parameters are shown in Table 2. When training the VAE-LSTM model, the time step is set to 2 and the batch_size is set to 32.

[0047] Table 2. Network parameters

[0048] The VAE-LSTM model uses the healthy first historical operation data as training data, and the hidden variables of the VAE-LSTM model learn the probability distribution of the healthy data. The key equipment of the bag dust removal system (dust collector) is the dust removal fan and dust removal motor. The health index of the dust bag equipment is judged by predicting the operating status of the fan motor. The equipment health index (HI) value range is 0~1. According to the model training results, the total HI value is above 0.6, the HI curve is not monotonous, and the entire curve fluctuates above 0.6. It can be considered that the dust collector is in a normal state. If the HI value is lower than 0.2, it can be considered that the wear of the dust collector is very serious. The HI value of the dust collector can be used to perform timely maintenance and maintenance on the dust collector.

[0049] In combination with the above embodiments, in one implementation, the present invention further provides an intelligent dust collector fault and health status diagnosis and management method. In this embodiment, the training step of the equipment fault classification model includes steps S31 to S34: Step S31: Obtain the historical second operation data of the dust collector key equipment.

[0050] The historical second operation data of the key equipment of the dust collector is derived from the data acquisition system. The sample distribution of the historical second operation data will have at least two categories, one is the positive sample obtained when the machine is operating normally, and the other is the negative sample obtained when the machine fails, as well as several possible transition intermediate states. The second operation data includes: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, and dust removal motor cooling water flow. The dust removal fan vibration data includes: fan front shaft Y-item vibration data, fan front shaft X-item vibration data, fan rear shaft Y-item vibration data, and fan rear shaft X-item vibration data. The dust removal motor vibration data includes: motor rear shaft Y-item vibration data, motor front shaft Y-item vibration data. The historical second operation data is the collected time series data, which contains 6 attributes, including name, type, value, timestamp, quality and version, as shown in Table 3.

[0051] Table 3. Time series data attribute table

[0052] Step S32: pre-processing the historical second operation data to obtain processed historical time series data.

[0053] The data collection frequency is once per second, and the collected data needs to be preprocessed. Preprocess the historical second operation data, such as data interpolation and outlier removal. For missing data points, use appropriate methods to interpolate, such as linear interpolation, polynomial interpolation, etc., to fill in the data gaps. Then, identify and remove outliers or noise in the data, which may be caused by sensor failure, data recording errors, etc.

[0054] Furthermore, the historical second operation data is preprocessed to obtain processed historical time series data, including: Data interpolation and outlier removal are performed on the historical second operation data to obtain processed historical time series data.

[0055] There is data with a missing timestamp in the historical second run data, and the data is supplemented by linear interpolation:

[0056] where x i is the time data of the interpolation point; y i is the attribute value at the interpolation point; x m and x n are the original time data at both ends of the interpolation point; m and n They are the attribute values ​​at both ends of the interpolation point. This ensures that the interpolation data only uses the original data and does not rely on the generated data. At this point, data supplementation and interpolation are completed, and then the outliers in the data are removed.

[0057] Step S33: Perform cluster analysis on the processed historical time series data to obtain training data with labels, wherein the labels include normal labels and abnormal labels.

[0058] The processed historical time series data is clustered by the K-means clustering algorithm to divide the data into different groups or clusters. Each cluster will be regarded as a category or label, that is, normal or abnormal. After cluster analysis, labeled training data is obtained, which will be used for subsequent model training. According to different needs and scenarios, the abnormal type can be further subdivided.

[0059] Furthermore, cluster analysis is performed on the processed historical time series data to obtain labeled training data, including: The processed historical time series data is clustered using the K-means clustering algorithm, and the optimal number of clusters is determined based on the silhouette coefficient of the clustering result. Each cluster is used as a category or label. According to the number of clusters, the processed historical time series data is clustered again using the K-means clustering algorithm to obtain the training data with labels.

[0060] The K-means clustering algorithm distributes the processed historical time series data into K clusters in an iterative manner, so that the data points in each cluster are as similar as possible, while the data points between different clusters are as different as possible. In order to determine how many clusters are appropriate, a grid search query method can be used, and the silhouette coefficient score of the clustering result is used as the standard to determine the optimal number of clusters. Figure 5 As shown in the figure, when the number of clusters is 2 and 3, the silhouette coefficient is at a high level, and when the number of clusters is 2, the silhouette coefficient is high, so the optimal number of clusters is determined to be 2. According to the determined optimal number of clusters, the K-means clustering algorithm is used again to cluster the processed historical time series data to obtain labeled training data. A total of 10-dimensional data is obtained, the first 9 dimensions are the attribute feature values ​​of the input time series data, and the last dimension is the label obtained by clustering, indicating the state judged by the corresponding input.

[0061] Step S34: input the training data into a plurality of models for training, and select the model with the best performance as the equipment fault classification model, the output data of the equipment fault classification model is the classification result.

[0062] Input the labeled training data into a variety of machine learning models for training. These models may include decision trees, random forests, GBDT (Gradient Boosting Decision Trees), SVM (Support Vector Machine), etc. By comparing the performance of different models on the validation set (such as accuracy, precision, F1 score, etc.), select the model with the best performance as the final equipment fault classification model.

[0063] As shown in Table 4, the evaluation results of the GBDT model and the SVM model show that the accuracy, precision, and F1 score of the GBDT model are slightly higher than those of the SVM model, indicating that the GBDT model has better performance. Therefore, the trained GBDT model can be used as an equipment fault classification model. Among them, the parameters that the GBDT model focuses on include learning_rate (learning rate), n_estimators (number of weak learners), and max_depth (maximum depth of the decision tree). The learning rate is set to the default value of 0.1, and the GridSearchCV method is used to perform grid search using n_estimators and max_depth respectively. By comparison, the optimal parameter combination is obtained when n_estimators is 60 and max_depth is 3, which can obtain a GBDT model with better results. GridSearchCV is an exhaustive search method that tries all possible parameter combinations and evaluates the performance of each combination through cross-validation.

[0064] Table 4. Evaluation indicators of GBDT model and SVM model

[0065] Based on the same inventive concept, an embodiment of the present application provides an intelligent dust collector fault and health status diagnosis and management device, the device comprising: An acquisition module is used to acquire real-time operation data of key equipment of the dust collector, and the real-time operation data includes at least: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, dust removal motor cooling water flow, and dust removal motor current; A classification module, used for dividing the real-time operation data to obtain first operation data and second operation data, wherein the first operation data includes: dust removal fan vibration data, dust removal motor vibration data, and dust removal motor current; and the second operation data includes: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, and dust removal motor cooling water flow; A health status assessment module, used for inputting the first operation data into a trained health status assessment model to obtain a health index, wherein the health status assessment model is used for judging an abnormal state that is significantly different from a normal state according to the regularity presented by each indicator data under the normal operation state of the dust collector, and judging the health state of the dust collector according to the difference in the abnormal state; A fault classification module, used for inputting the second operation data into a trained equipment fault classification model to obtain a classification result; The processing module is used to issue an alarm message when the health index is lower than a first threshold value, and generate fault processing information according to the classification result.

[0066] In an optional implementation, the health status assessment module is specifically used to: Obtain the normal historical first operation data of the key equipment of the dust collector; The normal historical first operation data is input into the VAE-LSTM model for training, the potential features of the normal historical first operation data are learned, and the trained health status assessment model is obtained, and the output data of the health status assessment model is the health index.

[0067] In an optional implementation manner, the fault classification module is specifically used to: Obtain the historical second operation data of key equipment of dust collector; Preprocessing the historical second operation data to obtain processed historical time series data; Performing cluster analysis on the processed historical time series data to obtain training data with labels, wherein the labels include: normal labels and abnormal labels; The training data is input into a plurality of models for training, and the model with the best performance is selected as the equipment fault classification model. The output data of the equipment fault classification model is the classification result.

[0068] In an optional implementation manner, the fault classification module is specifically used to: Data interpolation and outlier removal are performed on the historical second operation data to obtain processed historical time series data.

[0069] In an optional implementation manner, the fault classification module is specifically used to: The processed historical time series data is clustered using the K-means clustering algorithm, and the optimal number of clusters is determined based on the silhouette coefficient of the clustering result. Each cluster is used as a category or label. According to the number of clusters, the processed historical time series data is clustered again using the K-means clustering algorithm to obtain the training data with labels.

[0070] In an optional embodiment, the device further comprises: A visualization module is used to visualize the obtained health index and the classification result on an industrial Internet of Things platform, and the industrial Internet of Things platform is used to collect real-time operating data of key equipment of the dust collector and perform statistical analysis on the real-time operating data.

[0071] Based on the same inventive concept, another embodiment of the present application provides an electronic device, referring to Figure 6 , Figure 6 is a schematic diagram of an electronic device shown in an embodiment of the present application. Figure 6As shown, the electronic device 100 includes: a memory 110 and a processor 120. The memory 110 and the processor 120 are connected via a bus communication. A computer program is stored in the memory 110. The computer program can be run on the processor 120 to implement the steps in the intelligent dust collector fault and health status diagnosis and management method disclosed in the embodiment of the present disclosure.

[0072] Based on the same inventive concept, the embodiment of the present disclosure also provides a computer-readable storage medium. When the instructions in the computer-readable storage medium are executed by the processor of a computer device, the computer device can execute the steps in the intelligent dust collector fault and health status diagnosis and management method described in any of the above embodiments of the present application.

[0073] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0074] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present application may adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the embodiments of the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0075] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0076] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1A function specified in one or more boxes.

[0077] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0078] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present application.

[0079] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0080] The above is a detailed introduction to the intelligent dust collector fault and health status diagnosis and management method, device, equipment and storage medium provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the ideas of the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. An intelligent dust collector fault and health status diagnosis and management method, characterized in that: include: Acquire real-time operation data of key equipment of the dust collector, the real-time operation data at least including: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, dust removal motor cooling water flow, dust removal motor current; The real-time operation data is divided to obtain first operation data and second operation data, wherein the first operation data includes: dust removal fan vibration data, dust removal motor vibration data, and dust removal motor current; and the second operation data includes: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, and dust removal motor cooling water flow; The first operating data is input into a trained health status assessment model to obtain a health index, wherein the health status assessment model is used to determine an abnormal state that is significantly different from a normal state according to the regularity of each indicator data under normal operation of the dust collector, and to determine the health status of the dust collector according to the difference in the abnormal state; Inputting the second operation data into a trained equipment fault classification model to obtain a classification result; When the health index is lower than a first threshold, an alarm message is issued, and fault handling information is generated according to the classification result.

2. The intelligent dust collector fault and health status diagnosis and management method according to claim 1 is characterized in that: The training steps of the health status assessment model include: Obtain the normal historical first operation data of the key equipment of the dust collector; The normal historical first operation data is input into the VAE-LSTM model for training, the potential features of the normal historical first operation data are learned, and the trained health status assessment model is obtained, and the output data of the health status assessment model is the health index.

3. The intelligent dust collector fault and health status diagnosis and management method according to claim 1 is characterized in that: The training steps of the equipment fault classification model include: Obtain the historical second operation data of key equipment of dust collector; Preprocessing the historical second operation data to obtain processed historical time series data; Performing cluster analysis on the processed historical time series data to obtain training data with labels, wherein the labels include: normal labels and abnormal labels; The training data is input into a plurality of models for training, and the model with the best performance is selected as the equipment fault classification model. The output data of the equipment fault classification model is the classification result.

4. The intelligent dust collector fault and health status diagnosis and management method according to claim 3 is characterized in that: Preprocessing the historical second operation data to obtain processed historical time series data includes: Data interpolation and outlier removal are performed on the historical second operation data to obtain processed historical time series data.

5. The intelligent dust collector fault and health status diagnosis and management method according to claim 3 is characterized in that: Perform cluster analysis on the processed historical time series data to obtain labeled training data, including: The processed historical time series data is clustered using the K-means clustering algorithm, and the optimal number of clusters is determined based on the silhouette coefficient of the clustering result. Each cluster is used as a category or label. According to the number of clusters, the processed historical time series data is clustered again using the K-means clustering algorithm to obtain the training data with labels.

6. The intelligent dust collector fault and health status diagnosis and management method according to claim 1 is characterized in that: The method further comprises: The obtained health index and the classification results are visualized on an industrial Internet of Things platform, and the industrial Internet of Things platform is used to collect real-time operating data of key equipment of the dust collector and perform statistical analysis on the real-time operating data.

7. The intelligent dust collector fault and health status diagnosis and management method according to claim 1 is characterized in that: The method further comprises: Obtain compressed air property data of dust collector; According to the compressed air property data, the exhaust gas dust concentration data is predicted through a pre-trained energy consumption model; the energy consumption model training learns the relationship between the compressed air property and the exhaust gas dust concentration; When the exhaust gas dust concentration data does not meet the industrial emission standard, adjustment information for the exhaust gas dust emission data is generated according to the exhaust gas dust concentration data.

8. An intelligent dust collector fault and health status diagnosis and management device, characterized in that: include: An acquisition module is used to acquire real-time operation data of key equipment of the dust collector, and the real-time operation data includes at least: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, dust removal motor cooling water flow, and dust removal motor current; A classification module, used for dividing the real-time operation data to obtain first operation data and second operation data, wherein the first operation data includes: dust removal fan vibration data, dust removal motor vibration data, and dust removal motor current; and the second operation data includes: dust removal fan vibration data, dust removal fan speed, dust removal fan current, dust removal motor vibration data, and dust removal motor cooling water flow; A health status assessment module, used for inputting the first operation data into a trained health status assessment model to obtain a health index, wherein the health status assessment model is used for judging an abnormal state that is significantly different from a normal state according to the regularity presented by each indicator data under the normal operation state of the dust collector, and judging the health state of the dust collector according to the difference in the abnormal state; A fault classification module, used for inputting the second operation data into a trained equipment fault classification model to obtain a classification result; The processing module is used to issue an alarm message when the health index is lower than a first threshold value, and generate fault processing information according to the classification result.

9. An electronic device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor implements the intelligent dust collector fault and health status diagnosis and management method described in any one of claims 1 to 6 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the intelligent dust collector fault and health status diagnosis and management method according to any one of claims 1 to 6.

Images