Fault monitoring mechanism for coal mill

Abstract

The invention relates to the field of coal mill fault monitoring, in particular to a fault monitoring mechanism for a coal mill. Comprising a shell arranged on the outer side of a roller of the coal mill in a sleeving mode, a second sound collection module arranged on the periphery of a structural part of the coal mill and an early warning management and control system used for analyzing the working state of the coal mill according to information collected by a first sound collection module and the second sound collection module, and a hollow ball is arranged in the shell; a suspended metal ball is arranged in the center of the interior of the hollow ball, a first sound collection module is arranged on the outer side of the hollow ball, and when the coal mill works and vibrates, the first sound collection module collects sound characteristics generated when the metal ball shakes and impacts the hollow ball. Through the hollow ball and the metal ball arranged in the hollow ball, a state of generating impact under the condition of vibration can be formed, information of converting vibration into sound can be obtained, and the collection is placed in a closed state, so that interference of an external environment can be avoided.

Description

Technical Field

[0001] This invention relates to the field of coal mill fault monitoring, and more particularly to a fault monitoring mechanism for coal mills. Background Technology

[0002] A coal mill is a processing device that crushes coal into pulverized coal particles that meet combustion requirements. Conventional coal mills typically use methods such as grinding or ball milling. For ball mills, multiple balls are placed inside a drum; the coal is crushed by the impact of these balls. Air circulation within the drum blows up the sized particles, which are then expelled through channels. Due to the flammable and explosive nature of pulverized coal, a safety control system is usually required to monitor the coal mill's operation and prevent potential safety risks.

[0003] In the safety management system of a coal mill, various auxiliary sensors are typically used to monitor the mill's operating parameters. Acoustic monitoring devices are usually used to collect noises generated during the mill's operation. The baseline formed by these noises reflects the equipment's working condition. If an anomaly occurs, the resulting noises will affect the fluctuation of the baseline, which is crucial for the safety management of the coal mill. However, acoustic monitoring devices typically collect sound by installing various types of microphone arrays or sound sensors. This method collects all nearby sounds that meet the collection frequency. Since coal mills usually operate in complex environments such as workshops and construction sites, using this method alone will result in various complex information in the collected data. It is difficult to completely remove this environmental information in subsequent data processing, affecting the safety management system's monitoring of the coal mill's operating status and easily leading to inaccurate safety risk assessments. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides the following technical solution: The fault monitoring mechanism for a coal mill includes: a shell fitted on the outside of the coal mill drum, a second sound acquisition module installed around the structural components of the coal mill, and an early warning and control system for analyzing the working status of the coal mill based on the information collected by the first and second sound acquisition modules.

[0005] Specifically, the outer shell has a hollow sphere inside, and a suspended metal ball is placed at the center of the hollow sphere. A sound acquisition module is placed on the outside of the hollow sphere. When the coal mill vibrates during operation, the sound acquisition module collects the sound characteristics generated by the metal ball shaking and hitting the hollow sphere.

[0006] As an improvement to the above technical solution, the outer shell is arranged in a ring shape. The outer shell includes a shell one and a shell two arranged symmetrically. The outer rings of the shell one and the shell two are provided with flange rings. The shell one and the shell two are fixed by bolts passing through the flange rings.

[0007] As an improvement to the above technical solution, the interior of the outer shell is provided with several mounting slots arranged in a ring. The mounting slots include a main cavity and a side cavity covering the outside of the main cavity. The hollow sphere is disposed inside the main cavity, and the sound acquisition module is disposed inside the side cavity.

[0008] As an improvement to the above technical solution, the main cavity and the side cavity are separated by a ring-shaped partition. The surface of the partition is provided with several embedding slots. The sound acquisition module is ring-shaped, and a microphone array is provided on its inner ring sidewall. Each microphone unit of the microphone array is inserted into the interior of an embedding slot.

[0009] As an improvement to the above technical solution, the mounting groove includes a positioning groove formed above the main cavity. The positioning groove is connected to the main cavity through a wire groove. The wire groove is vertically arranged and faces the center of the hollow sphere. There is a connecting wire in the wire groove. The other end of the connecting wire is connected to a positioning block. The positioning block is set inside the positioning groove. The connecting wire passes through the wire groove and the main cavity and enters the interior of the hollow sphere to be fixed to the metal sphere.

[0010] As an improvement to the above technical solution, the interior of the main cavity is provided with several support parts, at least six of which are divided into two groups of at least three, respectively located on both sides of the hollow sphere, with the ends of the support parts contacting the surface of the hollow sphere.

[0011] As an improvement to the above technical solution, connecting feet are provided on both sides of the lower end of the outer shell, and elastic elements are connected to the lower end face of the connecting feet, and a base plate is connected to the other side of the elastic elements.

[0012] As an improvement to the above technical solution, the early warning and control system includes a data processing module, a data analysis module, and an early warning module. The data processing module is used to preprocess and extract features from the sound information collected by the sound acquisition module 1 and the sound acquisition module 2. The data analysis module is used to analyze the features extracted by the data processing module and determine whether an early warning is needed. The early warning module is used to execute an early warning action based on the early warning signal given by the data analysis module.

[0013] As an improvement to the above technical solution, the data analysis module includes a data statistics module, a deep learning module, and a feature fusion module. The data statistics module is used to record the working parameters of the coal mill under normal working conditions and compare the features extracted by the sound acquisition module 1 and the sound acquisition module 2 to preliminarily determine whether there is a possibility of a fault. The feature fusion module is used to perform weighted fusion of the features extracted by the sound acquisition module 1 and the sound acquisition module 2 and input the fused features into the deep learning module. When the data statistics module determines that there is a possibility of a fault, the deep learning module performs data analysis based on the input features to determine whether the model has a fault.

[0014] As an improvement to the above technical solution, the method by which the data statistics module initially determines the possibility of a fault relies on the following steps: The operating parameters of the coal mill under normal operating conditions are collected, and a behavioral baseline under normal operating conditions is established.

[0015] The features extracted from sound acquisition module one and sound acquisition module two are constructed into a line state and compared with the baseline to determine whether it exceeds the normal fluctuation range. If it exceeds the threshold, it is considered that there is a possibility of failure, and it is judged as a suspected anomaly, and subsequent steps are executed.

[0016] The beneficial effects of this invention are: By using a hollow sphere and a metal sphere inside it, a state of impact can be created when the sphere vibrates. This impact produces relatively clear sound information. This information is collected by an external sound acquisition module, thus obtaining information about the vibration converted into sound. Moreover, this acquisition is placed in a closed state, which can avoid interference from the external environment, ensure the accuracy of the collected data, and reduce the complexity of data processing. Attached Figure Description

[0017] Figure 1 This is a three-dimensional structural diagram of the present invention; Figure 2 This is an exploded structural diagram of the present invention; Figure 3 for Figure 2 Enlarged structural diagram at point A in the middle; Figure 4 for Figure 3 Enlarged structural diagram at point B; Figure 5 This is a front view structural diagram of the present invention; Figure 6 for Figure 5 Isometric side sectional view of the structure at point aa; Figure 7 for Figure 6Enlarged structural diagram at point C; Figure 8 This is a schematic diagram of the early warning and control system of the present invention.

[0018] Figure label: 10. Base plate; 11. Elastic components; 20. Outer shell; 21. Shell 1; 22. Shell 2; 23. Mounting slot; 231. Main cavity; 232. Side cavity; 233. Cable groove; 234. Positioning slot; 235. Partition plate; 236. Embedded groove; 237. Support part; 24. Hollow ball; 25. Metal ball; 251. Connecting thread; 252. Positioning block; 26. Sound acquisition module 1; 261. Microphone unit; 27. Sound acquisition module 2; 28. Connecting foot; 29. ​​Flange ring; 30. Early warning and control system; 31. Data processing module; 32. Data analysis module; 321. Data statistics module; 322. Deep learning module; 323. Feature fusion module; 33. Early warning module. Detailed Implementation

[0019] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0020] In the safety management system of a coal mill, various auxiliary sensors are typically used to monitor the mill's operating parameters. Acoustic monitoring devices are usually used to collect noises generated during the mill's operation. The baseline formed by these noises reflects the equipment's working condition. If an anomaly occurs, the resulting noises will affect the fluctuation of the baseline, which is crucial for the safety management of the coal mill. However, acoustic monitoring devices typically collect sound by installing various types of microphone arrays or sound sensors. This method collects all nearby sounds that meet the collection frequency. Since coal mills usually operate in complex environments such as workshops and construction sites, using this method alone will result in various complex information in the collected data. It is difficult to completely remove this environmental information in subsequent data processing, affecting the safety management system's monitoring of the coal mill's operating status and easily leading to inaccurate safety risk assessments.

[0021] To address the above problems, the following implementation method is provided: Example 1

[0022] Please see Figures 1 to 8The system provides a fault monitoring mechanism for a coal mill, including: a housing 20 fitted around the outside of the coal mill drum, a second sound acquisition module 27 set around the structural components of the coal mill, and an early warning and control system 30 for analyzing the working status of the coal mill based on the information collected by the first sound acquisition module 26 and the second sound acquisition module 27.

[0023] Specifically, a hollow sphere 24 is provided inside the outer shell 20, and a suspended metal sphere 25 is provided at the center of the hollow sphere 24. A sound acquisition module 26 is provided on the outside of the hollow sphere 24. When the coal mill vibrates during operation, the sound acquisition module 26 collects the sound characteristics generated by the metal sphere 25 shaking and hitting the hollow sphere 24.

[0024] In this embodiment, the entire outer shell 20 is first fitted onto the roller of the coal mill. The outer shell 20 does not directly contact the roller; there is a small gap between them, typically 2-5mm. This way, during low-frequency normal operation of the coal mill, there will be no significant shaking, preventing the metal ball 25 from impacting the inner wall of the hollow ball 24. Data monitoring in this situation is performed using a coarse sound acquisition module 27. This module is installed on various key structural components, such as the motor, bearings, and rollers, collecting normal operation information from these locations. As the operating efficiency of the coal mill increases, the overall... Each roller will produce slight shaking. Normally, if an abnormality occurs during operation, the abnormality will also be reflected in the shaking of the roller. In both cases, the shaking will exceed the normal shaking situation, causing the outer wall of the roller to come into contact with the inner wall of the outer casing 20, which in turn causes the metal ball 25 to shake. The shaking metal ball 25 collides with the inner wall of the hollow ball 24. In this case, the sound acquisition module 26 can collect the sound information generated by the collision between the metal ball 25 and the hollow ball 24. Since the structural components are all located inside the outer casing 20, the external sound can be isolated, thereby ensuring the accuracy of the collected data.

[0025] During operation, data monitoring under normal conditions is carried out through sound acquisition module 27, while data monitoring under high-power operation and abnormal conditions is carried out simultaneously by sound acquisition module 1 26 and sound acquisition module 27. This method can reduce the dependence on the data of sound acquisition module 27 and ensure the accuracy of the analysis results of the early warning and control system 30.

[0026] To ensure the internal structure is removable, please refer to [link / reference]. Figure 1 and Figure 2 Specifically, the outer shell 20 is arranged in a ring shape. The outer shell 20 includes a first shell 21 and a second shell 22 arranged symmetrically. A flange ring 29 is provided on the outer ring of the first shell 21 and the second shell 22. The first shell 21 and the second shell 22 are fixed by bolts passing through the flange ring 29.

[0027] The outer shell 20 is divided into two symmetrically arranged structural components, and the various slots inside are also divided in two. The internal structure is fixed inside the shell 21 and the shell 22 by clamping the slots. When disassembly is required, the screws and nuts on the flange ring 29 can be loosened to remove the various structural components inside the shell 21 and the shell 22, which facilitates subsequent maintenance, replacement and other operations.

[0028] To further refine the internal structural design of the outer casing 20, please refer to Figures 1 to 4 The outer casing 20 has several ring-shaped mounting slots 23 inside. The mounting slots 23 include a main cavity 231 and a side cavity 232 covering the outside of the main cavity 231. The hollow sphere 24 is disposed inside the main cavity 231, and the sound acquisition module 26 is disposed inside the side cavity 232.

[0029] The hollow sphere 24 and the sound acquisition module 26 are fixed in place by the main cavity 231 and side cavity 232 within the mounting slot 23, respectively. They are installed using an embedded method via an annular slot. To further ensure the detachability of subsequent structures, please refer to [link / reference needed]. Figures 1 to 4 The main cavity 231 and the side cavity 232 are separated by a ring-shaped partition 235. The surface of the partition 235 is provided with several embedding slots 236. The sound acquisition module 26 is ring-shaped, and a microphone array is provided on its inner ring sidewall. Each microphone unit 261 of the microphone array is inserted into the interior of an embedding slot 236.

[0030] The main cavity 231 is separated from the side cavity 232 by the partition 235. This separation method can ensure the isolation between the sound acquisition module 26 and the hollow sphere 24. The microphone unit 261 is embedded separately through the embedding slot 236. The partition 235 needs to be made of sound-absorbing material. This method can separate the body of the sound acquisition module 26 from the sound acquisition part, thereby avoiding the sound from bouncing inside and causing abnormalities in the collected sound.

[0031] To meet the design requirements for the suspended and free-impact states of the metal ball 25, please refer to... Figures 1 to 7 The mounting groove 23 includes a positioning groove 234 opened above the main cavity 231. The positioning groove 234 and the main cavity 231 are connected by a wire groove 233. The wire groove 233 is vertically arranged and faces the center of the hollow ball 24. There is a connecting wire 251 in the wire groove 233. The other end of the connecting wire 251 is connected to a positioning block 252. The positioning block 252 is set inside the positioning groove 234. The connecting wire 251 passes through the wire groove 233 and the main cavity 231 and enters the interior of the hollow ball 24 to be fixed to the metal ball 25.

[0032] The metal ball 25 is suspended from the center of the hollow ball 24 by the connecting wire 251, while the positioning block 252 at the top is used to provide the suspension connection for the metal ball 25, ensuring that the metal ball 25 can swing. The connecting wire 251 needs to be made of a fine filament material with high strength and good toughness, such as carbon fiber.

[0033] In the aforementioned design, the hollow sphere 24 is directly placed inside the main cavity 231. This design may result in a muffled sound, thus affecting the sound acquisition effect. To reduce the occurrence of this problem, please refer to [the relevant documentation / reference]. Figures 1 to 4 The main cavity 231 has a number of support parts 237 inside. There are at least six support parts 237, divided into two groups of at least three, which are respectively arranged on both sides of the hollow sphere 24. The ends of the support parts 237 contact the surface of the hollow sphere 24.

[0034] The hollow sphere 24 is positioned by the upper and lower support parts 237, ensuring its stability. Due to the symmetrical design of the outer shell 20, the hollow sphere 24 can be directly inserted into the shell, and then the other side of the shell 22 can be closed to secure it. There is also a considerable space between the hollow sphere 24 and the main cavity 231. The absorbent material-designed partition 235 prevents sound rebound from affecting the collection results. To further prevent sound rebound and interference from the external environment, the outer shell 20 can also be made of sound-absorbing material. For example, a layer of sound-absorbing material can be used on the inner surface, and a sound-insulating material on the outer surface. This reduces sound rebound internally, while the external design prevents sound from entering the interior.

[0035] In the aforementioned structural component design, it is necessary to ensure that the outer casing 20 can provide feedback when impacted by the coal mill drum. Therefore, the installation of the outer casing 20 cannot be fixed. Considering that the drum typically generates continuous impacts during operation, the support structure design of the outer casing 20 is as follows: Figure 1 As shown, specifically, connecting feet 28 are provided on both sides of the lower end of the outer casing 20, and elastic elements 11 are connected to the lower end face of the connecting feet 28. The other side of the elastic elements 11 is connected to the base plate 10.

[0036] The elastic element 11 ensures the elastic state of the outer shell 20. This design can immediately generate a shaking feedback when the roller hits the outer shell 20, so that the hollow ball 24 inside and the metal ball 25 collide and make a sound.

[0037] In the aforementioned technical design, it is necessary to ensure that all metal balls 25, connecting wires 251 and positioning blocks 252 are vertically arranged from bottom to top. This is to ensure that, in the case of a ring design, each metal ball 25 can generate a stable feedback action, thus ensuring the accuracy of sound acquisition in the ring structure. Example 2

[0038] To further utilize the data collected in Example 1, please refer to... Figure 8 The early warning and control system 30 includes a data processing module 31, a data analysis module 32, and an early warning module 33. The data processing module 31 is used to preprocess and extract features from the sound information collected by the sound acquisition module 1 26 and the sound acquisition module 27. The data analysis module 32 is used to analyze the features extracted by the data processing module 31 and determine whether an early warning is needed. The early warning module 33 is used to execute early warning actions based on the early warning signal given by the data analysis module 32.

[0039] The data processing module 31 typically preprocesses the input data by denoising it. The denoising method usually involves using filters. After denoising, the data is usually input into the input layer of the deep learning network for subsequent feature extraction.

[0040] In this embodiment, the feature extraction stage can use a convolutional neural network (CNN) for feature extraction, typically extracting key information such as energy distribution and frequency time-frequency components. In addition, to further improve the model's ability to recognize dynamic information, a recurrent neural network (RNN) can be used to establish temporal dependencies between the extracted features, thereby improving the ability to identify dynamic faults. The extracted features are then input into the data analysis module 32. The analysis is usually performed using a pre-designed model, and the model's output is used to determine whether a warning action needs to be taken. The warning module 33 typically uses various types of audible or luminous warning devices such as horns, buzzers, and warning lights to promptly alert relevant personnel to take action when a safety risk occurs.

[0041] In one embodiment, the data analysis module 32 includes a data statistics module 321, a deep learning module 322, and a feature fusion module 323. The data statistics module 321 is used to record the working parameters of the coal mill under normal working conditions and compare the features extracted by the sound acquisition module 1 26 and the sound acquisition module 27 to preliminarily determine whether there is a possibility of a fault. The feature fusion module 323 is used to perform weighted fusion of the features extracted by the sound acquisition module 1 26 and the sound acquisition module 2 27 and input the fused features into the deep learning module 322. When the data statistics module 321 determines that there is a possibility of a fault, the deep learning module 322 performs data analysis based on the input features to determine whether the model has a fault.

[0042] In this embodiment, the data analysis is divided into two parts: the first part is preliminary comparison, and the second part is fault analysis. The preliminary comparison is performed by the data statistics module 321. Specifically, the method by which the data statistics module 321 initially determines the possibility of a fault depends on the following steps: Step 1: Collect the operating parameters of the coal mill under normal operating conditions and establish a behavioral baseline under normal operating conditions.

[0043] When a coal mill is in normal working condition, its sound parameters are constantly changing. However, under normal working conditions, whether at low or high power, its sound frequency should be in a relatively stable state. Therefore, the change of its parameters over time can be represented by a fluctuating baseline.

[0044] Step 2: Construct the features extracted from sound acquisition module 1 (26) and sound acquisition module 2 (27) into a line state, and compare it with the baseline to determine whether it exceeds the normal fluctuation range. If it exceeds the threshold, it is considered that there is a possibility of failure, and it is judged as a suspected anomaly, and then the subsequent steps are executed.

[0045] During operation, the changes in parameters collected by sound acquisition module 1 (26) and sound acquisition module 2 (27) over time can also be represented by a line. By comparing this line with the baseline, the fluctuation difference can be directly obtained. If the difference exceeds the threshold, it is considered that the fluctuation is large and there is a possibility of failure. It is judged as abnormal and the feature fusion part is executed. If the fluctuation is considered to be small and there is no possibility of failure, the subsequent steps are not started.

[0046] The feature fusion module 323 comprises two parts. The first part is the part that converts vibration information into sound information, acquired by the sound acquisition module 1 26. The second part is the part that directly acquires the sound information of each device during operation, acquired by the sound acquisition module 27. Because the information from the sound acquisition module 1 26 has high accuracy and low interference, its weight is set higher during fusion, typically 0.7. However, the sound acquisition module 27 contains more environmental information, and even after data cleaning, interference from environmental information cannot be completely avoided. Therefore, its weight is set lower during fusion, typically 0.3. These weights need to be adjusted according to the specific working environment. If the environmental impact of the working environment is low, the weight of the sound acquisition module 27 can be appropriately increased.

[0047] The deep learning module 322 is typically used to perform the final classification decision and map classification information. It consists of multiple layers, as detailed below: Fully connected layer: Integrates features extracted by CNN and RNN and performs classification mapping; Dropout layer: Prevents overfitting and improves the model's generalization ability; Softmax classifier: outputs the probability distribution of each class; Loss function: Cross-entropy loss function is used to guide model training and minimize prediction error; if there is a class imbalance problem, Focal Loss can be used to optimize the training effect.

[0048] The deep learning module 322 outputs the judgment of the device's operating parameters to determine whether the device is faulty. If it is determined to be faulty, it sends a message to the early warning module 33, which then performs an alarm action. If it is not determined to be faulty, it does not perform any action.

[0049] The above embodiments are merely illustrative of the technical solutions of the present invention and are not intended to limit it. Anyone skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A fault monitoring mechanism for a coal mill, characterized in that, include: A shell (20) is fitted on the outside of the coal mill drum. A hollow ball (24) is provided inside the shell (20). A suspended metal ball (25) is provided at the center of the hollow ball (24). A sound acquisition module (26) is provided on the outside of the hollow ball (24). Sound acquisition module 2 (27) is installed around the structural components of the coal mill. An early warning and control system (30) for analyzing the working status of a coal mill based on information collected by sound acquisition module 1 (26) and sound acquisition module 2 (27); When the coal mill vibrates during operation, the sound acquisition module (26) collects the sound characteristics generated by the metal ball (25) shaking and hitting the hollow ball (24).

2. The fault monitoring mechanism for a coal mill according to claim 1, characterized in that: The outer shell (20) is arranged in a ring shape. The outer shell (20) includes a shell one (21) and a shell two (22) arranged symmetrically. The outer rings of the shell one (21) and the shell two (22) are provided with flange rings (29). The shell one (21) and the shell two (22) are fixed by bolts passing through the flange rings (29).

3. The fault monitoring mechanism for a coal mill according to claim 1, characterized in that: The housing (20) has several ring-shaped mounting slots (23) inside. The mounting slots (23) include a main cavity (231) and a side cavity (232) covering the outside of the main cavity (231). The hollow sphere (24) is located inside the main cavity (231), and the sound acquisition module (26) is located inside the side cavity (232).

4. The fault monitoring mechanism for a coal mill according to claim 3, characterized in that: The main cavity (231) and the side cavity (232) are separated by a ring-shaped partition (235). The surface of the partition (235) is provided with a number of embedding slots (236). The sound acquisition module (26) is ring-shaped, and a microphone array is provided on its inner ring sidewall. Each microphone unit (261) of the microphone array is inserted into an embedding slot (236).

5. The fault monitoring mechanism for a coal mill according to claim 3, characterized in that: The mounting groove (23) includes a positioning groove (234) opened above the main cavity (231). The positioning groove (234) and the main cavity (231) are connected by a wire groove (233). The wire groove (233) is vertically arranged and faces the center of the hollow sphere (24). There is a connecting wire (251) in the wire groove (233). The other end of the connecting wire (251) is connected to a positioning block (252). The positioning block (252) is set inside the positioning groove (234). The connecting wire (251) passes through the wire groove (233) and the main cavity (231) and enters the interior of the hollow sphere (24) to be fixed with the metal ball (25).

6. The fault monitoring mechanism for a coal mill according to claim 3, characterized in that: The main cavity (231) is provided with a number of support parts (237). There are at least six support parts (237), divided into two groups of at least three, which are respectively arranged on both sides of the hollow sphere (24). The ends of the support parts (237) contact the surface of the hollow sphere (24).

7. The fault monitoring mechanism for a coal mill according to any one of claims 1-6, characterized in that: Connecting feet (28) are provided on both sides of the lower end of the outer shell (20). An elastic element (11) is connected to the lower end face of the connecting feet (28). A base plate (10) is connected to the other side of the elastic element (11).

8. The fault monitoring mechanism for a coal mill according to any one of claims 1-6, characterized in that: The early warning and control system (30) includes a data processing module (31), a data analysis module (32), and an early warning module (33). The data processing module (31) is used to preprocess and extract features from the sound information collected by the sound acquisition module 1 (26) and the sound acquisition module 2 (27). The data analysis module (32) is used to perform data analysis on the features extracted by the data processing module (31) and determine whether an early warning is needed. The early warning module (33) is used to execute an early warning action based on the early warning signal given by the data analysis module (32).

9. The fault monitoring mechanism for a coal mill according to claim 8, characterized in that: The data analysis module (32) includes a data statistics module (321), a deep learning module (322), and a feature fusion module (323). The data statistics module (321) is used to record the working parameters of the coal mill under normal working conditions and compare the features extracted by the sound acquisition module 1 (26) and the sound acquisition module 2 (27) to preliminarily determine whether there is a possibility of a fault. The feature fusion module (323) is used to perform weighted fusion of the features extracted by the sound acquisition module 1 (26) and the sound acquisition module 2 (27) and input the fused features into the deep learning module (322). When the data statistics module (321) determines that there is a possibility of a fault, the deep learning module (322) performs data analysis based on the input features to determine whether there is a fault in the model.

10. The fault monitoring mechanism for a coal mill according to claim 9, characterized in that: The method by which the data statistics module (321) initially determines the possibility of a fault relies on the following steps: Collect the operating parameters of the coal mill under normal operating conditions and establish a behavioral baseline under normal operating conditions; The features extracted from the sound acquisition module 1 (26) and the sound acquisition module 2 (27) are constructed into a line state and compared with the baseline to determine whether it exceeds the normal fluctuation range. If it exceeds the threshold, it is considered that there is a possibility of failure, and it is judged as a suspected abnormality, and subsequent steps are executed.

Images