Coal conveyer belt electrical fault early warning system and method based on Internet of Things

Abstract

The invention discloses a coal conveyor belt electrical fault early warning system and method based on the Internet of Things, and the method comprises the steps: obtaining operation state data and environment monitoring data, carrying out the preprocessing, analyzing an abnormal mode through a DBSCAN algorithm, generating a feature data set, building an early warning model based on an LSTM network, and carrying out the training, thereby obtaining an early warning result. And generating a maintenance and isolation instruction to regulate and control the equipment. The real-time performance and accuracy of electrical fault early warning can be effectively improved, the downtime of equipment is shortened, safe and stable operation of the coal conveying belt is guaranteed, and meanwhile the early warning comprehensiveness and reliability are improved in combination with multi-factor data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coal transportation and electrical equipment safety monitoring, and particularly relates to a coal transportation electrical fault early warning system and method based on the Internet of Things. BACKGROUND

[0002] As a key equipment in coal mine production, the running state of the coal transportation belt is directly related to the production efficiency and safety. However, the electrical system of the coal transportation belt is easily affected by factors such as environment, load change and equipment aging during long-term operation, resulting in frequent electrical faults. Such faults are sudden and hidden, and if not discovered and handled in time, they may cause equipment downtime or even safety accidents. Therefore, how to realize early warning of electrical faults of the coal transportation belt and ensure the safe and stable operation of the equipment has become a key problem in current research. At present, the early warning methods for electrical faults of the coal transportation belt mainly focus on two aspects: one is to monitor the changes of electrical parameters through traditional sensors to realize preliminary judgment of faults; the other is to use data analysis technology to mine historical operation data to improve the accuracy of fault prediction. Therefore, how to further improve the real-time performance and reliability of electrical fault early warning is a technical problem to be solved at the present stage. SUMMARY

[0003] The purpose of the present application is to provide a coal transportation belt electrical fault early warning system and method based on the Internet of Things, which solves the problems mentioned in the background art.

[0004] To achieve the above-mentioned purpose, the technical scheme of the present application is as follows: in the above-mentioned coal transportation belt electrical fault early warning method based on the Internet of Things, the early warning method comprises the following steps:

[0005] Obtain the running state data of the electrical system of the coal transportation belt, and perform data preprocessing on the running state data to obtain a standardized running state data set;

[0006] Collect environmental monitoring data of the environment where the coal transportation belt is located, and perform data preprocessing on the environmental monitoring data to obtain a standardized environmental monitoring data set;

[0007] Perform abnormal pattern analysis on the standardized running state data set and the standardized environmental monitoring data set using the DBSCAN density clustering algorithm to obtain an electrical fault feature data set;

[0008] Construct an electrical fault early warning model based on the LSTM long short-term memory network, input the electrical fault feature data set into the electrical fault early warning model for training, and obtain an electrical fault early warning result;

[0009] Based on the electrical fault early warning results, electrical equipment maintenance instructions and fault isolation instructions are generated, and the electrical equipment is controlled based on the electrical equipment maintenance instructions and fault isolation instructions.

[0010] Furthermore, in the above-mentioned IoT-based method for early warning of electrical faults in coal conveyor belts, the step of acquiring the operating status data of the electrical system of the coal conveyor belt and preprocessing the operating status data to obtain a standardized operating status dataset includes:

[0011] Obtain the operating status data of the electrical system of the coal conveyor belt. The operating status data includes at least current waveform data, voltage fluctuation data, power factor data, temperature change data, vibration frequency data, and equipment start-up and shutdown status data.

[0012] The outliers and duplicates in the running status data are removed to obtain the first running status data;

[0013] The missing values ​​in the first operating state data are supplemented using cubic spline interpolation to obtain the second operating state data;

[0014] The second running state data is subjected to min-max normalization to obtain a standardized running state dataset.

[0015] Furthermore, in the above-mentioned method for early warning of electrical faults in coal conveyor belts based on the Internet of Things, the step of collecting environmental monitoring data of the environment where the coal conveyor belt is located, and preprocessing the environmental monitoring data to obtain a standardized environmental monitoring dataset, includes:

[0016] Collect environmental monitoring data of the environment where the coal conveyor belt is located. The environmental monitoring data shall include at least dust concentration data, humidity data, temperature data, air pressure data, wind speed data, and light intensity data.

[0017] Outliers and duplicates in the environmental monitoring data are removed to obtain initial environmental monitoring data.

[0018] The KNN nearest neighbor algorithm is used to fill in the missing values ​​in the initial environmental monitoring data to obtain complete environmental monitoring data;

[0019] The complete environmental monitoring data is Z-score standardized to obtain a standardized environmental monitoring dataset.

[0020] Furthermore, in the aforementioned IoT-based method for early warning of electrical faults in coal conveyor belts, the DBSCAN density clustering algorithm is used to perform anomaly pattern analysis on the standardized operating status dataset and the standardized environmental monitoring dataset to obtain an electrical fault feature dataset, including:

[0021] Set the neighborhood radius parameter and minimum number of sample points parameter for the DBSCAN density clustering algorithm;

[0022] Calculate the neighborhood density of each data point in the standardized operational status dataset and the standardized environmental monitoring dataset;

[0023] Data points are divided into core points, boundary points, and noise points based on neighborhood density.

[0024] Clusters are formed by connecting core points, and boundary points are assigned to the clusters to which the nearest core point belongs.

[0025] Unassigned data points are marked as outliers, forming an electrical fault feature dataset.

[0026] Furthermore, in the aforementioned IoT-based method for early warning of electrical faults in coal conveyor belts, the step of constructing an electrical fault early warning model based on an LSTM (Long Short-Term Memory) network, and inputting the electrical fault feature dataset into the electrical fault early warning model for training, to obtain electrical fault early warning results, includes:

[0027] An electrical fault early warning model is constructed based on an LSTM (Long Short-Term Memory) network. The electrical fault early warning model includes at least a forget gate, an input gate, and an output gate.

[0028] The Tanh activation function is used as the activation function for setting the electrical fault early warning model;

[0029] The learning rate of the electrical fault early warning model is set to 0.0015, and the number of hidden layer nodes is [20, 100].

[0030] The electrical fault feature dataset is divided into a training set and a validation set in a 7:3 ratio.

[0031] The training set is input into the electrical fault early warning model for iterative training until the loss function converges.

[0032] The validation set is input into the trained electrical fault early warning model to obtain the electrical fault early warning result.

[0033] Furthermore, in the above-mentioned method for early warning of electrical faults in coal conveyor belts based on the Internet of Things, the step of generating electrical equipment maintenance instructions and fault isolation instructions based on the early warning results, and regulating the electrical equipment based on the maintenance instructions and fault isolation instructions, includes:

[0034] Based on the electrical fault early warning results, electrical equipment maintenance instructions and fault isolation instructions are generated;

[0035] The electrical equipment maintenance instructions include at least cleaning and maintenance of the frequency converter within the next 7 days and replacement of the motor bearings within the next 14 days.

[0036] The fault isolation command includes at least automatically cutting off the relevant power supply line when a current fluctuation is detected to exceed a threshold.

[0037] The electrical equipment is controlled based on the electrical equipment maintenance instructions and fault isolation instructions. The electrical equipment includes at least a frequency converter, a motor controller, and a circuit breaker.

[0038] Furthermore, in an IoT-based early warning system for electrical faults in coal conveyor belts, the system includes the following modules:

[0039] The operation data acquisition module is used to acquire the operation status data of the electrical system of the coal conveyor belt, and to perform data preprocessing on the operation status data to obtain a standardized operation status dataset.

[0040] The environmental data acquisition module is used to collect environmental monitoring data of the environment where the coal conveyor belt is located, and to perform data preprocessing on the environmental monitoring data to obtain a standardized environmental monitoring dataset.

[0041] An anomaly pattern analysis module is used to perform anomaly pattern analysis on the standardized operating status dataset and the standardized environmental monitoring dataset using the DBSCAN density clustering algorithm to obtain an electrical fault feature dataset.

[0042] The early warning model construction module is used to build an electrical fault early warning model based on an LSTM long short-term memory network. The electrical fault feature dataset is input into the electrical fault early warning model for training to obtain electrical fault early warning results.

[0043] The equipment control module is used to generate electrical equipment maintenance instructions and fault isolation instructions based on the electrical fault early warning results, and to control the electrical equipment based on the electrical equipment maintenance instructions and fault isolation instructions.

[0044] Furthermore, in an IoT-based early warning system for electrical faults in coal conveyor belts, the operational data acquisition module includes the following sub-modules:

[0045] The operation data acquisition submodule is used to acquire the operating status data of the electrical system of the coal conveyor belt. The operating status data includes at least current waveform data, voltage fluctuation data, power factor data, temperature change data, vibration frequency data, and equipment start-up and shutdown status data.

[0046] The data cleaning submodule is used to remove outliers and duplicates from the running status data to obtain the first running status data;

[0047] The data filling submodule is used to fill in the missing values ​​in the first running state data using cubic spline interpolation to obtain the second running state data;

[0048] The data standardization submodule is used to perform min-max normalization on the second running state data to obtain a standardized running state dataset.

[0049] Furthermore, in an IoT-based early warning system for electrical faults in coal conveyor belts, the early warning model construction module includes the following sub-modules:

[0050] The model initialization submodule is used to construct an electrical fault early warning model based on an LSTM long short-term memory network. The electrical fault early warning model includes at least a forget gate, an input gate, and an output gate.

[0051] The activation function configuration submodule is used to use the Tanh activation function as the activation function for setting the electrical fault early warning model;

[0052] The parameter setting submodule is used to set the learning rate of the electrical fault early warning model to 0.0015 and the number of hidden layer nodes to [20, 100].

[0053] The model training submodule is used to divide the electrical fault feature dataset into a training set and a validation set in a 7:3 ratio, and input the training set into the electrical fault early warning model for iterative training until the loss function converges.

[0054] The result generation submodule is used to input the validation set into the trained electrical fault early warning model to obtain electrical fault early warning results.

[0055] Its beneficial effects lie in the following: By acquiring operational status data of the electrical system of the coal conveyor belt, preprocessing the operational status data to obtain a standardized operational status dataset; collecting environmental monitoring data of the environment where the coal conveyor belt is located, preprocessing the environmental monitoring data to obtain a standardized environmental monitoring dataset; using the DBSCAN density clustering algorithm to perform anomaly pattern analysis on the standardized operational status dataset and the standardized environmental monitoring dataset to obtain an electrical fault feature dataset; constructing an electrical fault early warning model based on an LSTM long short-term memory network, inputting the electrical fault feature dataset into the electrical fault early warning model for training, and obtaining electrical fault early warning results; generating electrical equipment maintenance instructions and fault isolation instructions based on the electrical fault early warning results, and controlling the electrical equipment based on the electrical equipment maintenance instructions and fault isolation instructions. This series of technical means can effectively improve the real-time performance and accuracy of electrical fault early warning, reduce equipment downtime caused by electrical faults, and ensure the safe and stable operation of the coal conveyor belt. Simultaneously, this method combines operational status data and environmental monitoring data, comprehensively considering multiple influencing factors, and improving the comprehensiveness and reliability of the early warning results. Attached Figure Description

[0056] Figure 1 This is a schematic diagram of the module structure of an IoT-based early warning system for electrical faults in a coal conveyor belt according to the present invention.

[0057] Figure 2 This is a schematic diagram of the internal structure of the operation data acquisition module of an IoT-based coal conveyor belt electrical fault early warning system according to the present invention.

[0058] Figure 3 This is an internal structural diagram of the early warning model construction module of an IoT-based early warning system for electrical faults in coal conveyors according to the present invention.

[0059] The attached diagram is labeled as follows: 1. Operational data acquisition module; 2. Environmental data acquisition module; 3. Anomaly pattern analysis module; 4. Early warning model construction module; 5. Equipment control module; 6. Data cleaning submodule; 7. Data filling submodule; 8. Data standardization submodule; 9. Model initialization submodule; 10. Model training submodule. Detailed Implementation

[0060] This invention provides an IoT-based electrical fault early warning system and method for coal conveyor belts, the detailed implementation process of which is described in conjunction with the appendix. Figure 1 Appendix Figure 2 and attached Figure 3 Explanation will be provided. (Attached) Figure 1The overall modular structure of the system is demonstrated, including module 1 for operational data acquisition, module 2 for environmental data acquisition, module 3 for anomaly pattern analysis, module 4 for early warning model construction, and module 5 for equipment control. (Appendix) Figure 2 It describes the specific process of data processing and analysis, and includes... Figure 3 The focus is on presenting the steps involved in training the early warning model and generating results.

[0061] In the specific implementation process, the operation data acquisition module 1 acquires the operation status data of the electrical system of the coal conveyor belt through its internal operation data acquisition submodule. This operation status data includes at least current waveform data, voltage fluctuation data, power factor data, temperature change data, vibration frequency data, and equipment start-up and shutdown status data. The operation data acquisition submodule is deployed at various key nodes of the coal conveyor belt through a sensor network and transmits the acquired data to the data cleaning submodule 6. The data cleaning submodule 6 removes outliers and duplicate values ​​from the operation status data to form the first operation status data. Subsequently, the data filling submodule 7 uses cubic spline interpolation to fill in the missing values ​​in the first operation status data to obtain the second operation status data. Finally, the data standardization submodule 8 performs min-max normalization on the second operation status data to form a standardized operation status dataset. The above submodules are connected through a data flow pipeline to ensure that data can be transmitted sequentially and preprocessed.

[0062] Meanwhile, the environmental data acquisition module 2 is responsible for collecting environmental monitoring data of the environment surrounding the coal conveyor belt. This environmental monitoring data includes at least dust concentration data, humidity data, temperature data, air pressure data, wind speed data, and light intensity data. The environmental data acquisition module 2 completes data acquisition through a sensor array deployed in the environment around the conveyor belt and transmits the raw data to the data cleaning submodule 6. The data cleaning submodule 6 removes outliers and duplicate values ​​from the environmental monitoring data to form initial environmental monitoring data. Subsequently, the data filling submodule 7 uses the KNN nearest neighbor algorithm to fill in missing values ​​in the initial environmental monitoring data to form complete environmental monitoring data. Finally, the data standardization submodule 8 performs Z-score standardization on the complete environmental monitoring data to form a standardized environmental monitoring dataset. The environmental data acquisition module 2 and the operational data acquisition module 1 achieve data synchronization and sharing through an internal data interface.

[0063] The anomaly pattern analysis module 3 receives the standardized operating status dataset from the operating data acquisition module 1 and the standardized environmental monitoring dataset from the environmental data acquisition module 2. The anomaly pattern analysis module 3 embeds the DBSCAN density clustering algorithm. First, it sets the neighborhood radius parameter and the minimum number of sample points parameter, then calculates the neighborhood density of each data point in the standardized operating status dataset and the standardized environmental monitoring dataset. Based on the neighborhood density, the data points are divided into core points, boundary points, and noise points. Clusters are formed by connecting core points, and boundary points are assigned to the clusters of the nearest core points. Unassigned data points are marked as anomalies, forming an electrical fault feature dataset. The anomaly pattern analysis module 3 uses its algorithm engine to classify data points and mark anomalies, and then transmits the final electrical fault feature dataset to the early warning model construction module 4.

[0064] The early warning model construction module 4 includes a model initialization submodule 9, an activation function configuration submodule, a parameter setting submodule, and a model training submodule 10. The model initialization submodule 9 constructs an electrical fault early warning model based on an LSTM (Long Short-Term Memory) network. This model includes at least a forget gate, an input gate, and an output gate. The activation function configuration submodule uses the Tanh activation function as the activation function for the electrical fault early warning model to enhance its nonlinear expressive power. The parameter setting submodule sets the learning rate of the electrical fault early warning model to 0.0015 and the number of hidden layer nodes to a range of 20 to 100. The model training submodule 10 divides the electrical fault feature dataset into a training set and a validation set in a 7:3 ratio and inputs the training set into the electrical fault early warning model for iterative training until the loss function converges. Subsequently, the model training submodule 10 inputs the validation set into the trained electrical fault early warning model to generate electrical fault early warning results and transmits the results to the equipment control module 5.

[0065] The equipment control module 5 generates electrical equipment maintenance instructions and fault isolation instructions based on the electrical fault early warning results. The electrical equipment maintenance instructions include at least cleaning and maintenance of the frequency converter within the next 7 days and replacing the motor bearings within the next 14 days. The fault isolation instructions include at least automatically disconnecting the relevant power supply line when a current fluctuation exceeds a threshold. The equipment control module 5 controls the electrical equipment, which includes at least the frequency converter, motor controller, and circuit breaker, through an execution unit. The equipment control module 5 interacts with the early warning model construction module 4 via a control signal interface and performs the actual control operations on the electrical equipment through the execution unit.

[0066] Throughout the system's operation, the various modules collaborate closely through data flow pipelines and control signal interfaces. The operational data acquisition module 1 and environmental data acquisition module 2 are responsible for collecting and preprocessing operational status data and environmental monitoring data, respectively, and transmitting the standardized datasets to the anomaly pattern analysis module 3. The anomaly pattern analysis module 3 extracts electrical fault feature datasets using the DBSCAN density clustering algorithm and transmits them to the early warning model construction module 4. The early warning model construction module 4 trains and generates the electrical fault early warning model based on an LSTM (Long Short-Term Memory) network and transmits the early warning results to the equipment control module 5. The equipment control module 5 generates maintenance and isolation commands based on the early warning results and controls the electrical equipment through execution units. This series of operations ensures the real-time performance and accuracy of electrical fault early warnings for the coal conveyor belt, thereby reducing equipment downtime caused by electrical faults and ensuring the safe and stable operation of the coal conveyor belt.

[0067] In a practical application scenario, assuming a coal conveyor belt is located in an environment with high dust concentration, its electrical system frequently experiences current fluctuations and temperature anomalies. The operational data acquisition module 1 collects current waveform data, voltage fluctuation data, and temperature change data in real time through a sensor network and transmits this data to the data cleaning submodule 6 for preprocessing. Simultaneously, the environmental data acquisition module 2 collects dust concentration data, humidity data, and temperature data, and uses the data cleaning submodule 6 and data filling submodule 7 to form a standardized environmental monitoring dataset. The anomaly pattern analysis module 3 receives the two standardized datasets and uses the DBSCAN density clustering algorithm to discover a significant correlation between current fluctuations and dust concentration. The early warning model construction module 4 trains the model based on an LSTM long short-term memory network and generates early warning results to predict potential future electrical faults. The equipment control module 5 generates maintenance instructions based on the early warning results, requiring cleaning and maintenance of the frequency converter within the next 7 days and automatically cutting off relevant power supply lines when the dust concentration exceeds a certain threshold to prevent electrical faults. To better enable those skilled in the art to fully understand and implement this invention, the specific implementation principle of this invention is further explained below in conjunction with a specific application scenario.

[0068] First, the operational data acquisition module 1 collects operational status data of the electrical system of the coal conveyor belt through its internal operational data acquisition submodule. In a real-world scenario, assuming the conveyor belt is located in a coal mine environment with high dust concentrations, its electrical system frequently experiences current fluctuations and temperature anomalies. The operational data acquisition submodule acquires real-time current waveform data, voltage fluctuation data, and temperature change data through a sensor network deployed at key nodes of the conveyor belt. This data is then transmitted to the data cleaning submodule 6, which removes outliers and duplicates to form the first operational status data. Next, the data filling submodule 7 uses cubic spline interpolation to fill in missing values, obtaining the second operational status data. Finally, the data standardization submodule 8 performs min-max normalization on the data to form a standardized operational status dataset. This series of operations ensures the integrity and consistency of the operational status data, providing a reliable foundation for subsequent analysis.

[0069] Meanwhile, the environmental data acquisition module 2 is responsible for collecting monitoring data of the environment surrounding the conveyor belt. In the above scenario, the environmental data acquisition module 2 acquires dust concentration, humidity, and temperature data through a sensor array deployed around the conveyor belt. This raw data is also transmitted to the data cleaning submodule 6, where outliers and duplicates are removed to form initial environmental monitoring data. Subsequently, the data imputation submodule 7 uses the KNN nearest neighbor algorithm to fill in missing values, generating complete environmental monitoring data. Finally, the data standardization submodule 8 performs Z-score standardization on the complete environmental monitoring data to form a standardized environmental monitoring dataset. This process effectively eliminates noise interference in the environmental data, ensuring data quality and comparability.

[0070] Once the standardized operational status dataset and the standardized environmental monitoring dataset are ready, the data will be transferred to the anomaly pattern analysis module 3. The anomaly pattern analysis module 3 embeds the DBSCAN density clustering algorithm, which extracts potential electrical fault features through joint analysis of the two datasets. Specifically, module 3 first sets the neighborhood radius parameter and the minimum number of sample points parameter, and calculates the neighborhood density of each data point. Based on the neighborhood density, the data points are divided into core points, boundary points, and noise points. Core points are connected to form clusters, and boundary points are assigned to the cluster to which the nearest core point belongs. Unassigned data points are considered anomalies, ultimately forming the electrical fault feature dataset. This density-based clustering method can effectively identify anomaly patterns in complex data, and is particularly suitable for non-linearly distributed data.

[0071] Subsequently, the early warning model construction module 4 receives the electrical fault feature dataset and constructs an electrical fault early warning model based on an LSTM (Long Short-Term Memory) network. The model initialization submodule 9 first constructs an LSTM network structure including forget gates, input gates, and output gates to capture long-term dependencies in the time-series data. The activation function configuration submodule uses the Tanh activation function to enhance the model's non-linear expressive power, while the parameter setting submodule sets the learning rate to 0.0015 and the number of hidden layer nodes to a range of 20 to 100. The model training submodule 10 divides the electrical fault feature dataset into training and validation sets in a 7:3 ratio, and inputs the training set into the model for iterative training until the loss function converges. After training, the validation set is used to evaluate the model performance and generate electrical fault early warning results. This process fully utilizes the advantages of LSTM networks in processing time-series data, significantly improving the accuracy and reliability of early warnings.

[0072] Based on the early warning results, the equipment control module 5 generates specific maintenance and isolation instructions. For example, in the scenario described above, when the model predicts a potential electrical fault due to excessive dust concentration, the equipment control module 5 generates a maintenance instruction requiring cleaning and maintenance of the frequency converter within the next 7 days and replacement of the motor bearings within the next 14 days. Furthermore, when a current fluctuation exceeds a set threshold, module 5 generates a fault isolation instruction, automatically disconnecting the relevant power supply lines to prevent the fault from escalating. These instructions are transmitted to electrical equipment such as the frequency converter, motor controller, and circuit breaker through the execution unit, triggering corresponding control operations. This mechanism enables early intervention in faults, effectively reducing equipment downtime and potential safety hazards.

[0073] Throughout the system's operation, the various modules collaborate closely through data flow pipelines and control signal interfaces. The operational data acquisition module 1 and environmental data acquisition module 2 are responsible for collecting and preprocessing operational status data and environmental monitoring data, respectively, and then transmit the standardized datasets to the anomaly pattern analysis module 3. The anomaly pattern analysis module 3 extracts electrical fault feature datasets using the DBSCAN density clustering algorithm and transmits them to the early warning model construction module 4. The early warning model construction module 4 completes model training and result generation based on an LSTM long short-term memory network and transmits the early warning results to the equipment control module 5. The equipment control module 5 generates maintenance and isolation commands based on the early warning results and executes the actual control operations on the electrical equipment through its execution unit. This series of steps ensures the real-time performance and accuracy of electrical fault early warnings for the coal conveyor belt, thereby guaranteeing the safe and stable operation of the equipment.

[0074] As can be seen from the implementation steps of the specific application scenarios described above, this invention achieves comprehensive monitoring and accurate early warning of electrical faults in coal conveyor belts through the collaborative work of multiple modules. From the collection and preprocessing of operational status data and environmental monitoring data, to the analysis of abnormal patterns and the construction of early warning models, and finally to equipment control, each step is closely linked and logically clear. This systematic solution not only improves the real-time performance and reliability of fault early warning but also significantly reduces equipment downtime caused by electrical faults, providing a strong guarantee for the safe and efficient operation of coal mine production.

[0075] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for early warning of electrical faults in coal conveyor belts based on the Internet of Things, characterized in that, The early warning method includes the following steps: The operating status data of the electrical system of the coal conveyor belt is acquired, and the operating status data is preprocessed to obtain a standardized operating status dataset. Environmental monitoring data of the environment where the coal transport belt is located is collected, and the environmental monitoring data is preprocessed to obtain a standardized environmental monitoring dataset. The DBSCAN density clustering algorithm was used to perform anomaly pattern analysis on the standardized operating status dataset and the standardized environmental monitoring dataset to obtain an electrical fault feature dataset. An electrical fault early warning model is constructed based on an LSTM (Long Short-Term Memory) network. The electrical fault feature dataset is input into the electrical fault early warning model for training, and the electrical fault early warning result is obtained. Based on the electrical fault early warning results, electrical equipment maintenance instructions and fault isolation instructions are generated, and the electrical equipment is controlled based on the electrical equipment maintenance instructions and fault isolation instructions.

2. The method for early warning of electrical faults in coal conveyor belts based on the Internet of Things as described in claim 1, characterized in that, The step of acquiring the operating status data of the electrical system of the coal conveyor belt and preprocessing the operating status data to obtain a standardized operating status dataset includes: acquiring the operating status data of the electrical system of the coal conveyor belt, wherein the operating status data includes at least current waveform data, voltage fluctuation data, power factor data, temperature change data, vibration frequency data, and equipment start-up and shutdown status data; The outliers and duplicates in the running status data are removed to obtain the first running status data; The missing values ​​in the first operating state data are supplemented using cubic spline interpolation to obtain the second operating state data; The second running state data is subjected to min-max normalization to obtain a standardized running state dataset.

3. The method for early warning of electrical faults in coal conveyor belts based on the Internet of Things as described in claim 1, characterized in that, The process involves collecting environmental monitoring data of the environment surrounding the coal conveyor belt, preprocessing the environmental monitoring data to obtain a standardized environmental monitoring dataset, including: collecting environmental monitoring data of the environment surrounding the coal conveyor belt, wherein the environmental monitoring data includes at least dust concentration data, humidity data, temperature data, air pressure data, wind speed data, and light intensity data; and removing outliers and duplicate values ​​from the environmental monitoring data to obtain initial environmental monitoring data. The KNN nearest neighbor algorithm is used to fill in the missing values ​​in the initial environmental monitoring data to obtain complete environmental monitoring data; The complete environmental monitoring data is Z-score standardized to obtain a standardized environmental monitoring dataset.

4. The method for early warning of electrical faults in coal conveyor belts based on the Internet of Things as described in claim 1, characterized in that, The DBSCAN density clustering algorithm is used to perform anomaly pattern analysis on the standardized operating status dataset and the standardized environmental monitoring dataset to obtain an electrical fault feature dataset, including: Set the neighborhood radius parameter and minimum number of sample points parameter for the DBSCAN density clustering algorithm; Calculate the neighborhood density of each data point in the standardized operational status dataset and the standardized environmental monitoring dataset; Data points are divided into core points, boundary points, and noise points based on neighborhood density. Clusters are formed by connecting core points, and boundary points are assigned to the clusters to which the nearest core point belongs. Unassigned data points are marked as outliers, forming an electrical fault feature dataset.

5. An Internet of Things-based electrical fault early warning system for coal conveyor belts, characterized in that, The early warning system includes the following modules: The operation data acquisition module 1 is used to acquire the operation status data of the electrical system of the coal conveyor belt, and to perform data preprocessing on the operation status data to obtain a standardized operation status dataset. Environmental data acquisition module 2 is used to collect environmental monitoring data of the environment where the coal conveyor belt is located, and to preprocess the environmental monitoring data to obtain a standardized environmental monitoring dataset. Anomaly pattern analysis module 3 is used to perform anomaly pattern analysis on the standardized operating status dataset and the standardized environmental monitoring dataset using the DBSCAN density clustering algorithm to obtain an electrical fault feature dataset. The early warning model construction module 4 is used to construct an electrical fault early warning model based on an LSTM long short-term memory network. The electrical fault feature dataset is input into the electrical fault early warning model for training to obtain electrical fault early warning results. The equipment control module 5 is used to generate electrical equipment maintenance instructions and fault isolation instructions based on the electrical fault early warning results, and to control the electrical equipment based on the electrical equipment maintenance instructions and fault isolation instructions.

6. The IoT-based electrical fault early warning system for coal conveyor belts as described in claim 5, characterized in that, The runtime data acquisition module 1 includes the following sub-modules: The operation data acquisition submodule is used to acquire the operating status data of the electrical system of the coal conveyor belt. The operating status data includes at least current waveform data, voltage fluctuation data, power factor data, temperature change data, vibration frequency data, and equipment start-up and shutdown status data. The data cleaning submodule 6 is used to remove outliers and duplicates from the running status data to obtain the first running status data; The data filling submodule 7 is used to fill in the missing values ​​in the first running state data using cubic spline interpolation to obtain the second running state data; The data standardization submodule 8 is used to perform minimum-maximum normalization processing on the second running state data to obtain a standardized running state dataset.

7. The IoT-based electrical fault early warning system for coal conveyor belts as described in claim 5, characterized in that, The early warning model construction module 4 includes the following sub-modules: Model initialization submodule 9 is used to construct an electrical fault early warning model based on an LSTM long short-term memory network. The electrical fault early warning model includes at least a forget gate, an input gate, and an output gate. The activation function configuration submodule is used to use the Tanh activation function as the activation function for setting the electrical fault early warning model; The parameter setting submodule is used to set the learning rate of the electrical fault early warning model to 0.0015 and the number of hidden layer nodes to a range of 20 to 100. The model training submodule 10 is used to divide the electrical fault feature dataset into a training set and a validation set in a 7:3 ratio, and input the training set into the electrical fault early warning model for iterative training until the loss function converges.

Images