Construction method of mine digital workshop
By adopting multi-source data fusion, deep learning analysis and digital twin technology in the mining digital workshop, the problem of inaccurate equipment failure prediction is solved, and efficient and safe equipment maintenance and production management is achieved.
Patent Information
- Application Number
- CN202411601332.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-11
- Publication Date
- 2025-03-28
AI Technical Summary
The prior art is difficult to achieve a comprehensive understanding of the operating status of mining equipment, resulting in inaccurate equipment failure prediction, increasing maintenance costs and affecting production continuity.
Through the combination of multi-source data fusion, deep learning analysis and digital twin technology, a mining digital workshop is built to achieve accurate prediction of equipment failures and dynamic optimization of maintenance plans.
It improves the accuracy of equipment prediction and maintenance, reduces the probability of sudden accidents and failures, reduces maintenance costs, and improves the safety and efficiency of mine production.
Smart Images
Figure CN119720488A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of intelligent mine management, and in particular to a method for constructing a digital mine workshop. Background Art
[0002] With the transformation and development of the mining industry towards digitalization and intelligence, the automation and information management of mining production equipment has gradually become an important means to improve mining production efficiency and reduce equipment failure rates. In the existing technology, the maintenance of mining equipment mainly relies on regular maintenance and manual inspections. Although some systems have introduced equipment status monitoring methods, they are usually limited to the monitoring of a single parameter or a specific device, and it is difficult to achieve a comprehensive understanding of the equipment's operating status. At the same time, the existing equipment maintenance strategies are often based on fixed cycles or single fault warning indicators, which makes it difficult to accurately predict equipment failures. As a result, in the actual production process, sudden equipment failures still occur from time to time, which not only increases maintenance costs, but also has a serious impact on the continuity of mining production. Therefore, the construction of digital workshops in mines has become an important trend in the current development of the mining industry, which can realize real-time monitoring and intelligent maintenance of production equipment and improve the safety and efficiency of mining production. Summary of the invention
[0003] In order to solve the problems existing in the prior art, the present invention provides a method for constructing a digital mine workshop. The digital mine workshop is constructed based on the combination of multi-source data fusion, deep learning analysis and digital twin technology. It can realize accurate prediction of equipment failures and dynamic optimization of maintenance plans, thereby ensuring efficient and safe operation of the mine workshop.
[0004] The technical solution adopted by the present invention to solve the technical problem is: a method for constructing a mine digital workshop, comprising the following steps:
[0005] Step S1: Collect multi-source data of the mine workshop through various types of sensors, and transmit the multi-source data to the integrated management and control platform after processing through edge computing devices;
[0006] Step S2: the integrated management and control platform processes the multi-source data to form a multi-dimensional database;
[0007] Step S3: Use artificial intelligence algorithms to build equipment failure prediction models, environmental anomaly prediction models, personnel abnormal behavior prediction models, production scheduling models, and dispatch prediction models, and obtain prediction analysis results;
[0008] Step S4: Use digital twin technology to simulate the spatial layout, environmental parameters, workshop equipment, and production process of the mine workshop to build a digital twin model and obtain a maintenance plan;
[0009] Step S5: The equipment status, prediction analysis results and maintenance plan are displayed in a visual form through the visual data display and information interaction platform.
[0010] Furthermore, the multi-source data includes vibration parameters, temperature parameters, stress parameters, sound waves, and laser scanning data of the equipment.
[0011] Furthermore, the edge computing device processes multi-source data including: noise reduction processing and data cleaning processing.
[0012] Furthermore, the artificial intelligence algorithm includes: long short-term memory network, convolutional neural network and random forest model; and the model is trained by the following loss function: N is the number of samples, y i is the true value, is the predicted value, λ is the regularization parameter, w j is the model parameter.
[0013] Furthermore, the training process of the model includes: extracting features from the data, reducing the dimensionality of high-dimensional data using principal component analysis, and tuning the hyperparameters of the model through Bayesian optimization methods.
[0014] Furthermore, digital twin technology simulates the equipment operation status and production process through finite element analysis, and uses the following formula to calculate the stress-strain relationship of the equipment: σ=E·ε, where σ represents stress, E represents elastic modulus, and ε represents strain.
[0015] Furthermore, the digital twin model interacts bidirectionally with the real-time operating data of the equipment to achieve state synchronization between the virtual equipment and the actual equipment, and adjust the maintenance plan in real time in the digital twin model.
[0016] Furthermore, the visual data display and information interaction platform is constructed using Grafana or Tableau, and the equipment status, predictive analysis results, and maintenance plans are visualized in the form of multi-dimensional charts, dashboards, and geographic information maps.
[0017] Furthermore, the maintenance plan is output based on an expert system and an artificial intelligence algorithm, and the maintenance plan is optimized and selected through a decision tree algorithm.
[0018] Furthermore, the visual data display and information interaction platform has a self-learning function and can continuously learn and optimize equipment operation data through a reinforcement learning algorithm.
[0019] The beneficial effects of the present invention are:
[0020] (1) The present invention uses intelligent algorithms to extract features and model the collected multi-source data, constructs a high-precision digital twin model, and combines historical data with real-time monitoring data to train and optimize the model. Through advanced algorithms such as long short-term memory networks, convolutional neural networks, and random forest models, it can accurately predict equipment failures, improve the accuracy of equipment predictive maintenance, reduce the probability of sudden accidents and failures, and ensure the continuity and orderliness of mine production.
[0021] (2) The present invention combines the prediction and analysis results with digital twin technology to simulate and verify and optimize the maintenance plan of the equipment. Digital twin technology simulates the actual operating status and maintenance process of the equipment, so that the maintenance plan can be dynamically adjusted according to the actual situation of the equipment, avoiding unnecessary maintenance operations, reducing maintenance costs, and improving the accuracy and efficiency of equipment maintenance.
[0022] (3) The present invention builds an integrated data visualization platform that can intuitively display equipment status, prediction results, maintenance plans and production processes in the form of multi-dimensional charts and dashboards. Managers can monitor the operation status of the workshop in real time through the visualization interface and make production and maintenance decisions quickly, thereby improving the management efficiency and decision-making level of the mine workshop. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] The present invention is further described below in conjunction with the accompanying drawings and embodiments.
[0024] Figure 1 It is a flow chart of the method for constructing a mine digital workshop of the present invention. DETAILED DESCRIPTION
[0025] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.
[0026] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.
[0027] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0028] like Figure 1 As shown, the method for constructing a mine digital workshop of the present invention comprises the following steps:
[0029] Step S1: Collect multi-source data from the mine workshop through various types of sensors, and transmit the multi-source data to the integrated management and control platform after processing through edge computing devices.
[0030] Step S2: The integrated management and control platform processes multi-source data to form a multi-dimensional database.
[0031] Step S3: Use artificial intelligence algorithms to build equipment failure prediction models, environmental anomaly prediction models, personnel abnormal behavior prediction models, production scheduling models, and dispatch prediction models, and obtain prediction analysis results.
[0032] Step S4: Use digital twin technology to simulate the spatial layout, environmental parameters, workshop equipment, and production process of the mine workshop to build a digital twin model and obtain a maintenance plan.
[0033] Step S5: The equipment status, prediction analysis results and maintenance plan are displayed in a visual form through the visual data display and information interaction platform.
[0034] It should be noted that there are many different types of sensors deployed in the mine workshop (such as temperature sensors, stress sensors, vibration sensors, acoustic wave sensors, laser scanning sensors, etc.) to collect equipment operating parameters and environmental parameters to form multi-source data. For example, vibration sensors are used to monitor the vibration characteristic signals of the equipment to identify the early characteristics of potential faults; temperature sensors monitor the temperature changes of equipment components in real time to detect potential overheating problems; stress sensors record the stress distribution of the equipment during operation to identify structural fatigue and stress concentration areas; acoustic wave sensors capture the sound characteristics emitted during the operation of the equipment to identify abnormal acoustic signals; laser scanning sensors are used to accurately measure the displacement and wear of the equipment. That is, multi-source data includes vibration parameters, temperature parameters, stress parameters, acoustic waves, laser scanning data, etc. of the equipment. The sensor transmits the data collected in real time to the edge computing device, which performs noise reduction processing (such as Yon Kalman filtering or wavelet transform) and data cleaning processing on the multi-source data to eliminate some interference data, and then transmits it to the comprehensive management and control platform by the edge computing device. The edge computing device can use NVIDIA Jetson or other high-performance edge computing devices to ensure that a large amount of data is processed quickly and in real time at the collection site, reducing the pressure of data transmission. During the data collection process, the edge computing device processes the collected raw data, mainly including data cleaning, noise reduction, compression and format conversion. The Kalman filter is used to smooth the continuous time series data to eliminate the high-frequency noise in the collection process. For non-stationary signals, the wavelet transform is used for multi-scale analysis to decompose the signal into different frequency bands to remove noise and extract effective information about the operation of the equipment. To ensure efficient data transmission, the industrial Internet of Things technology is used to connect each sensor and edge computing device to the integrated management and control platform through a wireless or wired network, and the message queue telemetry transmission protocol is used to transmit data. The message queue telemetry transmission protocol has the advantages of low bandwidth occupancy and high transmission efficiency, which is particularly suitable for the complex environment and large-scale data transmission requirements of the mine workshop. In terms of network architecture, the local area network of the mine workshop is built by combining industrial Ethernet with 5G network to realize the rapid transmission and aggregation of multi-source data. During the data transmission process, the network data encryption and verification mechanism is used to ensure the security and integrity of data transmission.
[0035] The integrated management and control platform stores, classifies and processes the received multi-source data to form a multi-dimensional database for subsequent query and use. After the data is aggregated to the integrated management and control platform, data fusion and preprocessing are performed to form a complete and accurate database. Data fusion includes time synchronization and spatial registration of different types of sensor data to ensure that all data are analyzed in the same time axis and spatial coordinate system. The data fusion process will combine the actual operating conditions of the equipment to achieve the association and matching between multi-source data. Data preprocessing includes data normalization, abnormal data removal and interpolation completion. For normalization, different types of sensor data are standardized to eliminate dimensional differences. Abnormal data can be removed by setting upper and lower data thresholds, and abnormal data points caused by sensor failure or environmental interference can be screened out. For missing data, interpolation algorithms (such as linear interpolation and spline interpolation) are used to complete the missing data to ensure data integrity. On the integrated management and control platform, a database based on distributed databases such as Hadoop and HBase is built to store and manage various types of data collected from sensors according to equipment, time and type to form a complete equipment operation data set. Use cloud storage technology to back up data in the cloud to ensure high availability and security of data. For example, use Extract, Transform, and Load tools to extract, transform, and load data, unify the data into the same standard format, and use feature engineering methods to process the data. For example, perform frequency domain analysis on equipment vibration signals to extract features such as characteristic frequency and energy spectrum; perform trend analysis on temperature data to extract features such as average temperature and temperature change rate. Use principal component analysis and other methods to reduce the dimensionality of high-dimensional data and extract key feature parameters to ensure the efficiency and accuracy of subsequent model training.
[0036] For example, during the data collection and fusion process, a real-time monitoring system is established to monitor the integrity and accuracy of data collection. When the system detects data transmission delays, data anomalies or sensor failures, alarm information is promptly fed back to ensure the continuity and effectiveness of data collection.
[0037] It should be noted that in order to improve the accuracy and reliability of data collection, calibration and data redundancy processing are performed on the data of different types of sensors to ensure the consistency and integrity of multi-source data.
[0038] Multi-sensor data calibration: For the vibration, temperature, stress, sound wave, laser scanning and other sensors installed on the mine workshop equipment, the calibration algorithm is used to calibrate the sensor data in real time, taking into account the possible measurement deviation or time delay between different sensors, and compare and synchronize the multi-source data of the same measurement object. For example, the least squares method is used to calibrate the sensor data to reduce the measurement error between different sensors, so that the data of each sensor remains consistent in the same time axis and spatial coordinate system.
[0039] Data redundancy processing: To improve the reliability of data collection, multiple sets of sensors are deployed for key equipment and important measurement parameters to form data redundancy. By comparing and fusing redundant data, rapid identification and correction of sensor failures or data anomalies can be achieved. In the data processing process, multiple redundant data fusion algorithms (such as weighted averaging) are used to perform weighted averaging on redundant data from different sensors to obtain more accurate and stable measurement values. The weight value can be dynamically adjusted according to the historical accuracy of the sensor and the stability of the current data to improve the reliability of the overall data.
[0040] For example, edge computing devices can be used to achieve real-time anomaly detection and data compression of collected data, further improving data transmission efficiency and system response speed.
[0041] Real-time anomaly detection: Anomaly detection algorithms are embedded in edge computing devices to monitor and analyze sensor data in real time, identify and filter abnormal data, avoid abnormal data transmission to the central data processing platform, and reduce data processing pressure. Anomaly detection algorithms use sliding window averaging and Z-score standard score methods to detect anomalies in sensor data. Specifically, the mean and standard deviation of the data in the sliding window are calculated. For data that exceeds the mean ±3 times the standard deviation, it is determined to be an outlier and is removed or marked. At the same time, edge computing devices are used to establish abnormal data logs to record detailed information for each anomaly for subsequent analysis and maintenance.
[0042] Data compression: To reduce the bandwidth occupied by data transmission, a data compression algorithm is used to compress sensor data in real time. For vibration signal data, discrete cosine transform is used for compression, and the time domain data is converted into frequency domain data. Only the main frequency components are retained, and high-frequency noise and redundant information are discarded to achieve the purpose of data compression. For other sensor data, such as temperature, stress, and displacement data, differential encoding and run-length encoding compression algorithms are used to encode the difference between adjacent data points, reduce data redundancy, and achieve efficient compression.
[0043] Use artificial intelligence algorithms to extract features from data in a multidimensional database, build multiple prediction models such as equipment failure prediction models, environmental anomaly prediction models, personnel abnormal behavior prediction models, production scheduling models, and scheduling prediction models, and train, test, and verify the prediction models so that the prediction models can output accurate prediction results. Extract features from data in a multidimensional database to facilitate intelligent analysis and model training in subsequent steps. For vibration data, perform frequency domain analysis to extract information such as characteristic frequency, amplitude, and energy spectrum; for temperature data, calculate features such as average value, maximum value, and rate of change; for stress data, analyze stress distribution and stress concentration. In the feature extraction process, Fourier transform can be used to convert time domain signals into frequency domain signals to analyze the vibration frequency characteristics of the equipment. In addition, the time-frequency characteristics of the vibration signal are obtained through time-frequency joint analysis (such as short-time Fourier transform and continuous wavelet transform) to identify abnormal equipment conditions.
[0044] For example, artificial intelligence algorithms include: long short-term memory networks, convolutional neural networks, and random forest models. Based on historical data and real-time data, machine learning and deep learning algorithms are used to build equipment failure prediction models. For time series data, long short-term memory networks are selected to predict the operating status of the equipment and capture dynamic changes during the operation of the equipment. For image data or two-dimensional signal data, convolutional neural networks are selected for feature extraction and classification. For a variety of feature data, a random forest model is used for comprehensive analysis. When training the model, a large amount of historical equipment operation data is used in combination with equipment maintenance records for supervised learning training. The model is verified by cross-validation method to ensure the prediction accuracy and robustness of the model. The loss function of the model adopts a combination of mean square error (MSE) and regularization term. The loss function formula is: N is the number of samples, y i is the true value, is the predicted value, λ is the regularization parameter, w j For example, the Bayesian optimization method can be used to tune the model's hyperparameters to improve the model's prediction performance. After the model training is completed, the model is self-learned and optimized through the reinforcement learning algorithm to adjust the model's prediction strategy under different working conditions and environmental conditions.
[0045] For example, digital twin technology simulates the equipment operation status and production process through finite element analysis, and uses the following formula to calculate the stress-strain relationship of the equipment: σ = E·ε, where σ represents stress, E represents elastic modulus, and ε represents strain. Digital twins make full use of physical models, sensors, operation history and other data, integrate multi-disciplinary, multi-physical quantity, multi-scale, and multi-probability simulation processes, complete mapping in virtual space, and thus reflect the entire life cycle of the corresponding physical equipment. The digital twin model interacts bidirectionally with the real-time operation data of the equipment to synchronize the status of the virtual equipment with the actual equipment, and adjust the maintenance plan in real time in the digital twin model.
[0046] For example, the visual data display and information interaction platform is built using Grafana or Tableau, and visualizes the equipment status, predictive analysis results, and maintenance plans in the form of multi-dimensional charts, dashboards, and geographic information maps. The maintenance plan is based on the output of the expert system and artificial intelligence algorithm, and the maintenance plan is optimized and selected through the decision tree algorithm. The visual data display and information interaction platform has a self-learning function and can continuously learn and optimize the equipment operation data through the reinforcement learning algorithm.
[0047] In summary, the present invention uses intelligent algorithms to extract features and model the collected multi-source data, builds a high-precision digital twin model, and combines historical data with real-time monitoring data to train and optimize the model. Through advanced algorithms such as long short-term memory networks, convolutional neural networks, and random forest models, it can accurately predict equipment failures, improve the accuracy of equipment prediction and maintenance, reduce the probability of sudden accidents and failures, and ensure the continuous orderliness of mine production. The present invention combines the prediction and analysis results with digital twin technology to simulate and verify and optimize the maintenance plan of the equipment. Digital twin technology simulates the actual operating status and maintenance process of the equipment, so that the maintenance plan can be dynamically adjusted according to the actual situation of the equipment, avoiding unnecessary maintenance operations, reducing maintenance costs, and improving the accuracy and efficiency of equipment maintenance. The present invention constructs an integrated data visualization platform that can intuitively display equipment status, prediction results, maintenance plans, and production processes in the form of multi-dimensional charts and dashboards. Managers can monitor the operation status of the workshop in real time through the visualization interface, make production and maintenance decisions quickly, and improve the management efficiency and decision-making level of the mine workshop.
[0048] Based on the above ideal embodiments of the present invention, the relevant staff can make various changes and modifications without departing from the technical concept of the present invention through the above description. The technical scope of the present invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.
Claims
1. A method for constructing a digital mine workshop, characterized in that: The following steps are involved: Step S1: Collect multi-source data of the mine workshop through various types of sensors, and transmit the multi-source data to the integrated management and control platform after processing through edge computing devices; Step S2: the integrated management and control platform processes the multi-source data to form a multi-dimensional database; Step S3: Use artificial intelligence algorithms to build equipment failure prediction models, environmental anomaly prediction models, personnel abnormal behavior prediction models, production scheduling prediction models, and dispatch prediction models, and obtain prediction analysis results; Step S4: Use digital twin technology to simulate the spatial layout, environmental parameters, workshop equipment, and production process of the mine workshop to build a digital twin model and obtain a maintenance plan; Step S5: The equipment status, prediction analysis results and maintenance plan are displayed in a visual form through the visual data display and information interaction platform.
2. The method for constructing a digital mine workshop according to claim 1, characterized in that: The multi-source data include vibration parameters, temperature parameters, stress parameters, sound waves, and laser scanning data of the equipment.
3. The method for constructing a digital mine workshop according to claim 1, characterized in that: The edge computing device processes multi-source data including: noise reduction processing and data cleaning processing.
4. The method for constructing a digital mine workshop according to claim 1, characterized in that: The artificial intelligence algorithm includes: long short-term memory network, convolutional neural network and random forest model; and the model is trained through the following loss function: N is the number of samples, y i is the true value, is the predicted value, λ is the regularization parameter, w j is the model parameter.
5. The method for constructing a digital mine workshop according to claim 4, characterized in that: The model training process includes: extracting features from the data, reducing the dimensionality of high-dimensional data using principal component analysis, and tuning the model's hyperparameters through Bayesian optimization methods.
6. The method for constructing a digital mine workshop according to claim 1, characterized in that: Digital twin technology simulates the equipment operation status and production process through finite element analysis, and uses the following formula to calculate the stress-strain relationship of the equipment: σ=E·ε, where σ represents stress, E represents elastic modulus, and ε represents strain.
7. The method for constructing a mine digital workshop according to claim 1, characterized in that: The digital twin model interacts bidirectionally with the real-time operation data of the equipment to synchronize the status of the virtual equipment with the actual equipment, and adjusts the maintenance plan in real time in the digital twin model.
8. The method for constructing a digital mine workshop according to claim 1, characterized in that: The visual data display and information interaction platform is built using Grafana or Tableau, and visualizes the equipment status, predictive analysis results, and maintenance plans in the form of multi-dimensional charts, dashboards, and geographic information maps.
9. The method for constructing a digital mine workshop according to claim 1, characterized in that: The maintenance plan is based on the output of the expert system and artificial intelligence algorithm, and the maintenance plan is optimized and selected through the decision tree algorithm.
10. The method for constructing a digital mine workshop according to claim 1, characterized in that: The visual data display and information interaction platform has a self-learning function and can continuously learn and optimize equipment operation data through a reinforcement learning algorithm.