A barrel mill fault data migration device and a fault data acquisition method

Abstract

The present application relates to the technical field of grinding equipment fault detection, and specifically provides a barrel type grinder fault data migration device and a fault data acquisition method. The fault data migration device comprises a slide rail base with slide rails, a power mechanism, two I-shaped supports, two end covers, multiple cabin sections, multiple sensor assemblies and a lining plate. The power mechanism and the two I-shaped supports are slidably arranged on the slide rail base. The multiple cabin sections and the two end covers are combined to form a closed cabin body, which is arranged between the two I-shaped supports. The cabin body is provided with a reinforcing rod around the periphery. By adjusting the size and strength of the cabin body, a real barrel type grinder is simulated to obtain corresponding fault data. In the barrel type grinder fault data acquisition method, the fault data is obtained through processes such as classification, preprocessing, self-adaptive modal decomposition and feature extraction of the collected fault data. The present application does not damage the real barrel type grinder, has low cost and is easy to operate.

Description

Technical Field

[0001] This invention relates to the field of grinding equipment fault detection technology, specifically providing a fault data migration device and fault data acquisition method for a cylindrical mill. Background Technology

[0002] In the current industrial production process, which is constantly moving towards intelligence and efficiency, the intelligent development of the cylindrical mill, as a key piece of equipment, has become an inevitable trend, and the requirements for safety and stable production are becoming increasingly stringent. To keep pace with this trend, many domestic studies have focused on using neural network training technology to carefully construct intelligent fault identification systems and early warning mechanisms for cylindrical mills. However, this innovative exploration is fraught with challenges. Obtaining fault data is extremely difficult, requiring not only a large amount of consumables and a long working time, but also a very long fault reproduction cycle. For fault detection of core components, sometimes even destructive experiments on the cylindrical mill are necessary, consuming enormous human and material resources. The generated data not only requires huge storage space, but also lacks clear characteristic features. Furthermore, traditional experimental cylindrical mills have significant limitations; key parameters cannot be flexibly adjusted, testing can only be conducted on a single model, and the acquired data is limited to testing purposes, making it difficult to directly transfer and apply to real cylindrical mills. This undoubtedly poses significant obstacles to the intelligent development of cylindrical mills. Summary of the Invention

[0003] To address the aforementioned problems, this invention provides a fault data migration device and a fault data acquisition method for a cylindrical mill. By setting up a fault data migration device to simulate a real cylindrical mill, fault data of the cylindrical mill can be fully acquired without damaging the real cylindrical mill. This method is low-cost and easy to operate.

[0004] The fault data migration device for a cylindrical mill provided by this invention is used to simulate the structure of a real cylindrical mill. The fault data migration device includes: It includes a slide rail base, a power mechanism, two I-beam supports, two end caps, multiple compartments, multiple sensor assemblies, and a liner. Two I-beam supports are respectively located at the left and right ends of the slide rail base and can move along the slide rail; two end caps are rotatably connected between the two I-beam supports via bearings, and a large gear is connected to the outer circumference of any one end cap; multiple compartments are spliced ​​end to end to form a hollow compartment, and two end caps are located at the left and right ends of the compartment to close the compartment; a liner is located on the inner surface of the compartment; the power mechanism is located on the slide rail base and can move along the slide rail; the power mechanism includes a small gear, which meshes with a large gear to transmit power; Adjust the distance between the two I-beam supports, and simultaneously increase or decrease the number of chamber sections and change the diameter of the chamber sections to change the size of the chamber; multiple reinforcing rods are installed on the chamber sections to adjust the strength of the chamber; sensor assemblies are installed at preset positions to collect mill operating data.

[0005] Preferably, the power mechanism includes a motor, power mechanism fixing screws, reducer bracket, reducer, and pinion; The reducer bracket is connected to the slide rail by the power mechanism fixing screw. The motor and reducer are mounted on the reducer bracket. The output shaft of the motor is connected to the input shaft of the reducer, and the output shaft of the reducer is connected to the pinion.

[0006] Preferably, the fault data migration device further includes two bearing assemblies, each bearing assembly including a bearing housing, a bearing, a bearing inlet pipe, and a bearing outlet pipe; one bearing assembly is connected to each I-beam support, and a bearing is provided in the bearing assembly for connecting the end cap; The bearing inlet and outlet oil pipes are used to supply lubricating oil to the bearing.

[0007] Preferably, the fault data migration device further includes a lubricating oil detection component, which includes a lubricating oil inlet pipe, a lubricating oil outlet pipe, and a core detection device. One end of the lubricating oil inlet pipe is connected to the bearing inlet pipe, and the other end passes through the core detection device and is connected to the lubricating oil injection system. One end of the lubricating oil outlet pipe is connected to the bearing outlet pipe, and the other end passes through the core detection device and is connected to the lubricating oil tank.

[0008] Preferably, the head includes a first head flange, a head neck, a head disc, and a second head flange connected in sequence; the first head flange is connected to a corresponding bearing via a bearing connector, and the second head flange is connected to the compartment.

[0009] Preferably, the compartment includes a first compartment flange, a compartment neck, a second compartment flange, a compartment inner hole distributed circumferentially along the inner wall of the compartment neck, and a plurality of reinforcing rods distributed circumferentially along the compartment neck. The number, diameter, and position of the reinforcing rods are adjusted according to the required strength of the fault data migration device.

[0010] A method for acquiring fault data of a cylindrical mill includes: S1: Prepare a fault data migration device to classify fault conditions and acquire corresponding fault data; S2: Preprocess the fault data, including fault data classification and interpolation. S3: Expand fault data using system identification methods; S4: Using the adaptive variational mode decomposition method, the expanded fault data is decomposed to obtain multiple intrinsic mode functions; preset screening conditions are set to retain intrinsic mode functions that meet the screening conditions, and the features of the retained intrinsic mode functions are extracted; the features extracted from the same intrinsic mode function are concatenated to form a two-dimensional matrix, and the two-dimensional matrix is ​​classified according to the fault condition type and fault level.

[0011] Preferably, when preparing the fault data migration device, the strength calculation method of the fault data migration device is as follows: When a real cylindrical mill does not have reinforcing ribs: , When a real cylindrical mill has reinforcing ribs: , in, This represents the actual outer diameter of the cylindrical mill chamber. This represents the actual internal diameter of the cylindrical mill chamber. This indicates the ratio of the fault data migration device to the actual cylindrical mill. , and These represent the number, width, and height of the reinforcing ribs on a real cylindrical mill, respectively. Indicates the outer diameter of the cabin of the fault data migration device. This indicates the internal diameter of the fault data migration device. This indicates the number of reinforcing rods on the fault data migration device. This indicates the distance from the center of the reinforcing rod to the outside of the fault data migration device. This indicates the diameter of the reinforcing rod.

[0012] Preferably, the method for expanding fault data is to trim each signal segment corresponding to each normalized fault data to achieve fault data expansion.

[0013] Preferably, the types of failure conditions include cracking at the root of the end cover journal, excessive wear of the liner, bearing failure, insufficient grinding steel balls, severe gear wear, mill dry-hammering, and loose bolts; Fault levels include minor faults, moderate faults, and severe faults.

[0014] Compared with the prior art, the present invention can achieve the following beneficial effects: This invention simulates a real cylindrical mill by setting up a fault data migration device, indirectly obtaining fault data from the real cylindrical mill without damaging it. It is low-cost and easy to operate. Furthermore, the dimensions and other parameters of the fault data migration device are adjustable, thus enabling it to simulate various models and specifications of cylindrical mills, making it highly versatile and widely applicable.

[0015] This invention obtains multi-source fault data by combining coupled simulation experiments, real cylindrical mill experimental data, and fault data migration device experiments, thereby reducing the difficulty and cost of fault data acquisition and enabling the application of test data or simulation data. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the fault data migration device provided in an embodiment of the present invention; Figure 2 This is a front view of a fault data migration apparatus provided according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of a bearing assembly provided according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of the lubricating oil detection component provided in an embodiment of the present invention; Figure 5 This is a structural schematic diagram of a bearing connector provided according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of the end cap provided according to an embodiment of the present invention; Figure 7 This is a structural schematic diagram of a compartment provided according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the structure of a sensor assembly provided according to an embodiment of the present invention; Figure 9 This is a flowchart of a fault data acquisition method provided according to an embodiment of the present invention; Figure 10 This is a schematic diagram of the source of fault data provided in an embodiment of the present invention; Figure 11 This is a flowchart of fault data expansion provided according to an embodiment of the present invention.

[0017] The reference numerals in the figures include: 1. Slide rail base, 101. Slide rail, 102. Weight reduction hole, 103. Side lug, 104. Side lug hole, 2. Power mechanism, 201. Motor, 202. Power mechanism fixing screw, 203. Output shaft, 204. Reducer bracket, 205. Reducer, 206. Pinion, 207. I-beam bracket, 3. Lower fixing pin, 301. Upper fixing pin, 302. Bearing assembly, 4. Bearing seat, 401. Bearing, 402. Bearing oil inlet pipe, 403. Bearing oil outlet pipe, 404. Lubricating oil detection assembly, 5. Lubricating oil inlet pipe, 501. Lubricating oil outlet pipe, 502. Core detection device, 503. Bearing connector, 6. Bearing connecting section, 60. 1. Head connecting plate 602, head connecting hole 603, head 7, first head flange 701, head neck 702, head plate 703, second head flange 704, first head flange hole 705, second head flange hole 706, large gear 8, compartment 9, compartment neck 901, first compartment flange 902, second compartment flange 903, first compartment flange hole 904, second compartment flange hole 905, compartment inner hole 906, reinforcing rod 907, sensor assembly 10, sensor housing 1001, clamping bolt 1002, rubber protective block 1003, core sensor 1004, liner 11. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and do not constitute a limitation thereof. Similar elements in different embodiments are referred to by associated similar element reference numerals. In the following embodiments, many details are described to facilitate a better understanding of the invention. However, those skilled in the art will readily recognize that some features may be omitted in different situations, or may be replaced by other elements, materials, or methods. In some cases, some operations related to the invention are not shown or described in the specification. This is to avoid obscuring the core parts of the invention with excessive description. For those skilled in the art, detailed description of these related operations is not necessary; they can fully understand the related operations based on the description in the specification and general technical knowledge in the art.

[0019] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined to form various implementations. Furthermore, the order of the steps or actions in the method description can be changed or adjusted in a manner readily apparent to those skilled in the art. Therefore, the various orders in the specification and drawings are merely for the clear description of a particular embodiment and do not imply a mandatory order, unless otherwise stated that a particular order must be followed.

[0020] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on this invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0021] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0022] The invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0023] like Figure 1 and Figure 2 As shown, this invention provides a fault data migration device for a cylindrical mill, used to characterize the structure of a real cylindrical mill. The fault data migration device includes: a slide rail base 1, a power mechanism 2, two I-beam supports 3, two bearing assemblies 4, two lubricating oil detection assemblies 5, two bearing connectors 6, two end caps 7, a large gear 8, multiple chambers 9, multiple sensor assemblies 10, and a liner 11. The size and strength of the fault data migration device can be varied to simulate the structure of a real cylindrical mill.

[0024] The slide rail base 1 is located at the bottom of the fault data migration device. Its structure includes two slide rails 101, weight-reducing holes 102, four side ears 103, and a side ear hole 104 on each side ear 103. In cross-section, the slide rail base 1 has an upper panel and "legs" on both sides of the lower surface of the upper panel, with both "legs" extending along the long side of the upper panel. The two slide rails 101 are located on the upper surface of the upper panel and within the side "legs". The weight-reducing holes 102 are located at the center of the upper panel for weight reduction. The four side ears 103 are located at the four corners of the slide rail base 1, and the fault data migration device is mounted on the workbench through the side ear holes 104 on the side ears 103. To ensure the normal operation of the slide rails 101, the height of the side ears 103 does not exceed the lowest point of the slide rails 101.

[0025] The power mechanism 2 is located on the upper surface of the slide rail base 1. The power mechanism 2 includes a motor 201, a motor mount 202, power mechanism fixing screws 203, an output shaft 204, a reducer bracket 205, a reducer 206, and a pinion 207. The motor mount 202 is mounted on the reducer bracket 205. The motor mount 202 consists of a main board and a mounting plate, which are perpendicular to each other and integrally formed. The mounting plate has connection holes through which the motor mount 202 is mounted on the reducer bracket 205. After installation, the plane of the main board is perpendicular to the plane of the reducer bracket 205. The motor 201 is a stepper motor, fixed to the motor mount 202 with four screws. The reducer 206 is a worm gear structure, fixed to the reducer bracket 205. The output shaft 204 of the motor 201 and the input shaft of the reducer 206 are connected via a coupling, which is equipped with a measuring device. The reducer bracket 205 has two mounting holes. The power mechanism 2 is movably fixed within the slide rails 101 on both sides of the slide rail base 1 via these mounting holes, the power mechanism fixing screw 203, and the hexagonal head nut. Specifically, the hexagonal head nut, which mates with the power mechanism fixing screw 203, is positioned within the slide rail 101. When the power mechanism fixing screw 203 rotates, there is no relative rotation between the hexagonal head nut and the power mechanism fixing screw 203. The output shaft of the reducer 206 is connected to a pinion 207.

[0026] On the left and right sides of the upper surface of the slide rail base 1, there are two I-beam brackets 3. The two I-beam brackets 3 have the same structure, and the planes on which the two I-beam brackets 3 are located are parallel to each other and perpendicular to the straight line of the two slide rails 101. The cross-section of the I-beam bracket 3 is "I" shaped, with an upper beam and a lower beam. Fixing holes are provided on both sides of the lower beam, and the two sides of the I-beam bracket 3 are movably fixed within the two slide rails 101 by two lower fixing screws 301, so that the I-beam bracket 3 can move along the slide rails 101. The fixing principle is the same as that of the fixing screw 203 of the power mechanism. Since both I-beam brackets 3 can move along the slide rails 101, the distance between the two I-beam brackets 3 can be adjusted to change the size of the fault data migration device. Another pair of fixing holes are provided on both sides of the upper beam, and the bearing assembly 4 is fixed to the I-beam bracket 3 by two upper fixing screws 302.

[0027] like Figure 3 As shown, there are two bearing assemblies 4, both with identical structures, and each is connected to the I-beam bracket 3. Specifically, each bearing assembly 4 includes a bearing housing 401, a bearing 402, a bearing oil inlet pipe 403, and a bearing oil outlet pipe 404. The bearing housing 401 has a circular inner cavity and end connecting plates at both ends below the circular inner cavity that mate with the upper beam of the I-beam bracket 3. The bearing assembly 4 is fixed to the upper surface of the upper beam of the I-beam bracket 3 by two upper fixing pins 302 and the end connecting plates. A removable and replaceable bearing 402 is installed inside the circular inner cavity. The bearing 402 is equipped with a bearing oil inlet pipe 403 and a bearing oil outlet pipe 404, which provide lubricating oil to the bearing 402.

[0028] like Figure 4 As shown, each bearing assembly 4 is equipped with a lubricating oil detection component 5. The lubricating oil detection component 5 includes a lubricating oil inlet pipe 501, a lubricating oil outlet pipe 502, and a core detection device 503. One end of the lubricating oil inlet pipe 501 is connected to the bearing inlet pipe 403, and the other end passes through the core detection device 503 and connects to the lubricating oil injection system. One end of the lubricating oil outlet pipe 502 is connected to the bearing outlet pipe 404, and the other end passes through the core detection device 503 and connects to the lubricating oil tank. The core detection device 503 has functions for detecting the lubricating oil temperature and the lubricating oil dielectric constant. Furthermore, the portions of the lubricating oil inlet pipe 501 and the lubricating oil outlet pipe 502 inside the core detection device 503 are metal pipes, while the portions outside the core detection device 503 are Teflon flexible hoses. The lubricating oil detection component 5 supplies lubricating oil to the bearing 402 and monitors the lubricating oil status in real time.

[0029] like Figure 1 and Figure 2 As shown, in this embodiment of the invention, multiple interconnected sections 9 are provided to form the body of the fault data migration device. Figure 7As shown, each compartment 9 has the same structure, including a compartment neck 901, a first compartment flange 902, a second compartment flange 903, first compartment flange holes 904 and second compartment flange holes 905, a compartment inner hole 906, and a reinforcing rod 907. The compartment neck 901 is a large-diameter tubular structure, with its left and right ends connected to the first compartment flange 902 and the second compartment flange 903, respectively. The first compartment flange 902 has multiple first compartment flange holes 904 evenly arranged circumferentially, and the second compartment flange 903 has multiple second compartment flange holes 905 evenly arranged circumferentially. Multiple compartments 9 are connected end to end and linked through corresponding first compartment flange holes 904 and second compartment flange holes 905. The assembled multiple compartments 9 form a chamber, which is smoothly and continuously internally connected and used to contain the material to be ground. Multiple compartment internal holes 906 are evenly arranged circumferentially inside the neck 901 of the compartment section. These internal holes 906 are non-through holes used to connect the liner plates 11. The shape and material of the liner plates 11 are not limited, and the liner plates 11 are replaceable. They are used to protect the inner wall of the compartment and prevent wear. Additionally, in certain specific cases, it is necessary to divide the internal space of the compartment into multiple sub-spaces to accommodate different types of materials being ground. In this case, a bulkhead can be clamped between two adjacent compartment sections 9 to form sub-spaces.

[0030] There are multiple flange holes 904 in the first compartment and multiple flange holes 905 in the second compartment. When connecting adjacent compartments 9, some of the flange holes 904 and 905 can be used, while the remaining flange holes 904 and 905 are used to connect reinforcing rods 907 or are not used. By adjusting the number, thickness, and connection position of the reinforcing rods 907, the strength of the fault data migration device can be adjusted to closely approximate the strength of a simulated real mill. In this embodiment of the invention, when splicing compartments 9, a bolt is installed at intervals of one or two flange holes 904 (or flange holes 905) in the first compartment for connection, and the interval holes are used to connect reinforcing rods 907.

[0031] like Figure 6As shown, the fault data migration device also includes two end caps 7, located at the left and right ends of the compartment to seal it. The two end caps 7 have identical structures, each comprising a first end cap flange 701, an end cap neck 702, an end cap plate 703, and a second end cap flange 704 connected in sequence. One end of the end cap neck 702 is connected to one end of the end cap plate 703, or they can be integrated into a single structure. Both the end cap neck 702 and the end cap plate 703 are disc-shaped, with the diameter of the end cap neck 702 being much smaller than that of the end cap plate 703. The first end cap flange 701 is located on the other end face of the end cap neck 702, and a plurality of first end cap flange holes 705 are evenly arranged circumferentially on the first end cap flange 701. The second end cap flange 704 is located on the other end of the end cap plate 703, and a plurality of second end cap flange holes 706 are evenly arranged circumferentially on the second end cap flange 704. Two second end flange holes 706 are used to connect to the outermost compartment 9 of the hull (first compartment flange hole 904 or second compartment flange hole 905) to seal the hull. Two first end flanges 701 on both sides of the hull are rotatably connected to the bearings 402 of the bearing assembly 4 via a bearing connector 6.

[0032] To enable the cabin rotation, a large gear 8 is installed on the end cap 7 closest to the power mechanism 2. Specifically, the large gear 8 is located on the outer periphery of the end cap disk 703 of this end cap 7. The power mechanism 2 is moved so that its small gear 207 can mesh with the large gear 8. The power output from the motor 201 is transmitted to the cabin through the reducer 206, causing the cabin to rotate.

[0033] like Figure 5 As shown, the bearing connector 6 includes a bearing connecting section 601 and a head connecting plate 602. The bearing connecting section 601 is tubular, and the head connecting plate 602 is located at one end of the bearing connecting section 601. The head connecting plate 602 has a plurality of head connecting holes 603 evenly arranged circumferentially. The bearing connecting section 601 is used to connect the bearing 402 of the bearing assembly 4, specifically, the bearing connecting section 601 and the inner ring of the bearing 402 are fitted together. The bearing connector 6 is connected to the first head flange 701 of the head 7 through the head connecting holes 603 on the head connecting plate 602. After installation, the housing can rotate.

[0034] The fault data migration device in this embodiment of the invention simulates the structure of a real cylindrical mill. Actual cylindrical mills vary in size, strength, and wear. To simulate the real cylindrical mill condition, the fault data migration device in this embodiment is adjusted as follows: By adjusting the distance between the two I-beam supports 3 to increase or decrease the number of chamber sections 9, combined with changing the diameter of the chamber sections, the proportional dimensions of the fault data migration device can be altered to achieve a preset ratio with the real cylindrical mill. By increasing or decreasing the number of reinforcing rods 907, and changing the position and diameter of the reinforcing rods 907, the strength of the fault data migration device can be adjusted to meet the required strength. To ensure that the wear condition of the device matches the wear condition of a real cylindrical mill, the wear condition of a real cylindrical mill can be installed, and the liner 11 can be artificially subjected to equivalent wear before being installed on the fault data migration device. The acquired fault data migration device can simulate the state of a real cylindrical mill, thereby simulating the faults of a real cylindrical mill based on the fault data migration device, realizing the "migration" of fault data from the real cylindrical mill to the fault data migration device, making data collection easier. Based on the fault data collected from the fault data migration device, the actual fault situation of the cylindrical mill is inferred.

[0035] In addition, to collect various data or signals at the required location on the fault data migration device, this embodiment of the invention sets up multiple sensor components 10 on the fault data migration device to obtain corresponding data at the target location and infer the fault. Specifically, such as... Figure 8 As shown, the sensor assembly 10 includes a sensor housing 1001, a clamping bolt 1002, a rubber protective block 1003, and a core sensor 1004. A threaded hole is formed at the upper end of the sensor housing 1001, through which the clamping bolt 1002 is screwed. The lower end of the clamping bolt 1002 is connected to the rubber protective block 1003, and the lower surface of the rubber protective block 1003 is connected to the core sensor 1004, through which data is acquired. The main material of the sensor housing 1001 is manganese silicon steel, with copper sheets inlaid or plated with copper at the mounting position. During installation, the copper-plated (copper sheet) position of the sensor housing 1001 can be soldered to the outer side of the polished compartment 9, or connected to the outer side of the bearing assembly 4, and can be removed with a hot air gun. The sensor assembly 10 is powered by an internal power supply, and data is transmitted to the terminal via Bluetooth. The type of sensor in the core sensor 1004 is selected according to requirements, including vibration sensors and vibration and noise sensors. In this embodiment of the invention, the vibration sensors are evenly arranged in a circular pattern on the outer side of the compartment 9, and one vibration and one noise sensor are respectively installed on the bearing assemblies 4 at both ends.

[0036] like Figure 9 As shown, based on the above-mentioned fault data migration device, this embodiment of the invention provides a method for acquiring fault data of a cylindrical mill, including the following steps: like Figure 10 As shown, S1: Prepare a fault data migration device to classify fault conditions and obtain corresponding fault data.

[0037] Based on the parameters of a real cylindrical mill, a fault data migration device with the aforementioned structure is fabricated according to a predetermined ratio. The parameters of the real cylindrical mill include its length, width, diameter, wall thickness, and other dimensional data, as well as performance parameters such as strength, wear condition, and rotational speed. Based on the length-to-width ratio and dimensions of the real cylindrical mill, a certain number of sections are selected and spliced ​​together to form a chamber. The outer ring diameter of the bearing assembly is determined according to the chamber dimensions. The diameter of the reinforcing rod is determined based on the ratio of the mill's wall thickness to its diameter. The strength of the fault data migration device is adjusted using the principle of equivalent moment of inertia.

[0038] The strength of the chamber of a real cylindrical mill can be adjusted by adding reinforcing ribs, which are typically installed on the outer wall of the chamber. To ensure the accuracy of the strength of the fault data migration device, this invention divides the strength calculation method into two categories: cases where the simulated real cylindrical mill has reinforcing ribs and cases where it does not. When the real cylindrical mill does not have reinforcing ribs: , When a real cylindrical mill has reinforcing ribs: , in, This represents the actual outer diameter of the cylindrical mill chamber. This represents the actual internal diameter of the cylindrical mill chamber. This indicates the ratio of the fault data migration device to the actual cylindrical mill. , and These represent the number, width, and height of the reinforcing ribs on a real cylindrical mill, respectively. Indicates the outer diameter of the cabin of the fault data migration device. This indicates the internal diameter of the fault data migration device. This indicates the number of reinforcing rods on the fault data migration device. This indicates the distance from the center of the reinforcing rod to the outer surface of the fault data migration device. This indicates the diameter of the reinforcing rod.

[0039] Real-world failures in cylindrical mills include end cover journal root cracking, excessive liner wear, bearing failure, insufficient grinding balls, severe gear wear, mill dry-running, and loose bolts. Bearing failures include pitting, peeling, and wear on the bearing surface. Accordingly, this invention classifies the failure conditions of the fault data migration device into seven types: end cover journal root cracking, excessive liner wear, bearing failure, insufficient grinding balls, severe gear wear, mill dry-running, and loose bolts. Each failure condition is further categorized into three levels: minor, moderate, and severe. In this embodiment, a minor failure involves some parts exhibiting less than 0.05% permanent shape change or some operating parameters exceeding design values ​​by less than 10%, with no abnormal noise, requiring no emergency shutdown and being included in the routine maintenance plan. A moderate failure involves some parts exhibiting 0.05%-0.1% permanent shape change or some operating parameters exceeding design values ​​by 10%-30%, with abnormal noise, requiring a scheduled shutdown for specialized repair. In the event of a severe malfunction, some parts will experience permanent shape changes of more than 0.1% or some operating parameters will exceed the design value by 30%, and abnormal noise will be severe. The machine must be stopped immediately for handling.

[0040] For the seven fault conditions mentioned above, the fault data migration device performs the following simulations respectively: 1. The simulation method for cracking at the root of the end cover journal is as follows: the end cap is cyclically loaded using a fatigue load tester until cracks appear at the root of the journal, and the end cap with cracks is used in the fault data migration device. 2. The simulation method for excessive wear of the liner is as follows: the liner is pre-treated for wear, or the liner is machined according to the shape of the liner after wear, and the pre-treated liner or the liner machined according to the shape is installed on the fault data migration device. 3. The simulation of bearing failure is as follows: scratches or pitting marks are applied to the bearing surface and the bearing is reinstalled; 4. The simulation method for insufficient grinding steel balls is to reduce the number of grinding steel balls to decrease their proportion. 5. The simulation method for severe gear wear is as follows: select severely worn gears for installation; 6. The simulation method of grinding mill empty impact is: reduce the injection of the material being ground in order to increase the ratio of grinding steel balls to the material being ground; 7. The simulation of bolt loosening is as follows: loosen the bolt at the corresponding position.

[0041] Because real cylindrical mills have theoretical speed requirements during operation, during fault experiments, each fault condition must be operated at both high and low speeds. The high and low speeds are the highest and lowest speeds commonly used in engineering, controlled by the equipment's gear settings. In the following three experiments, sensor components were used to acquire multiple sets of fault data for the seven fault conditions and three fault levels mentioned above. Figure 10 (Corresponding to 42 datasets). The three types of experiments are a coupled simulation experiment based on discrete element-finite element-multibody dynamics, a fault data migration device experiment, and a real cylindrical mill experiment. The reason for having three types of experiments is that different fault conditions are suitable for different acquisition methods. For example, the fault condition of cracking at the root of the end cover journal is not suitable for acquisition from the real cylindrical mill experiment, but is suitable for acquisition from the fault data migration device experiment.

[0042] S2: Preprocess the collected fault data.

[0043] After collecting and acquiring fault data, the actual fault data of the cylindrical mill has been transferred to the fault data transfer device. This process was completed without damaging or affecting the actual operation of the cylindrical mill. The fault data is then used to extrapolate the actual fault conditions of the cylindrical mill.

[0044] The acquired fault data undergoes preprocessing, including fault data classification and interpolation. The fault data is divided into two categories based on its characteristics. The first category consists of fault data exhibiting obvious periodicity and fluctuation, such as vibration data from accelerometers and acoustic fingerprint data from sound sensors; this is termed periodic fault data. The second category consists of fault data with good smoothness, such as non-periodic or slowly changing data, like bearing position temperature variation data and lubricating oil wear index. This is referred to as smooth fault data. The process involves setting upper and lower bounds for fault data. Fault data is compared one by one with these bounds, and data exceeding the upper or lower bounds are selected; these are called out-of-bounds fault data. Interpolation is then performed on this out-of-bounds fault data to supplement the data. The differences between data points are calculated to obtain a complete difference sequence, and the mean and standard deviation of the difference sequence are determined. Interpolation is used to supplement data with particularly severe abrupt changes. If the absolute values ​​of the differences between two adjacent segments before and after a data point are both greater than the sum of the mean of the difference sequence and three times the standard deviation, then this data point is identified as an outlier, and its value is replaced with the average of the two normal data points before and after it.

[0045] S3: Expand fault data using system identification methods.

[0046] like Figure 11 As shown, this embodiment of the invention utilizes a structural dynamics physical model and a data-driven nonlinear parameter identification method, combined with a residual convolutional neural network, to model the conversion relationship between experimental data from a fault data migration device and real cylindrical mill experimental data. By comprehensively utilizing coupled simulation experiments and real cylindrical mill experimental data, the parameter variation law is fitted, achieving data expansion. Specifically, based on periodic fault data... A data conversion relationship between the fault data migration device and the actual cylindrical mill was established, and nonlinear parameter identification based on the structural dynamics physical model and data-driven methods was performed. A residual convolutional neural network was constructed and trained on the residual part that the theoretical model could not predict, so as to obtain the relationship between the residual and frequency and correct the error of the theoretical model itself.

[0047] By comprehensively utilizing coupled simulation experiments, fault data migration device experiments (implemented using a fault data migration device), and real cylindrical mill experiments, the parameters involved in the data conversion relationship between the fault data migration device and the real cylindrical mill are fitted to obtain the first fitting formula: , in, This represents data obtained from coupled simulation calculations of a real cylindrical mill in the frequency domain. The coefficients of the first fitting formula are represented. The value of must be such that The value should be as close as possible to The value, and The value should be as close as possible to The value, This represents the data obtained from the frequency domain coupled simulation calculation of the fault data migration device. This represents the theoretical law governing parameter changes between the fault data migration device and the actual cylindrical mill. This represents the residual data relationships that theoretical patterns fail to reflect. This represents data obtained from actual cylindrical mill experiments. This represents the data obtained from the fault data migration device experiment.

[0048] The same method was used to identify parameters for the data conversion relationship between the coupled simulation experiment and the actual cylindrical mill experiment, and the second fitting formula was obtained. , in, This describes the theoretical laws governing parameter changes between coupled simulation experiments and actual cylindrical mills. The coefficients of the second fitting formula, This describes the theoretical laws governing parameter variations between the finite element analysis and actual cylindrical mill analysis. This represents the residual data relationships that the theoretical laws fail to reflect.

[0049] and The data is obtained through training a residual neural network. During the training process, all parameters of the residual convolutional neural network are optimized in the training of the first equation of the first fitting formula and the first equation of the second fitting formula. In the training of the second equation of the first fitting formula and the second equation of the second fitting formula, the parameters of the first few layers of the convolutional network are fixed, and only the last two layers of the network are adjusted. If the fault mode is not suitable for data acquisition on a real cylindrical mill, only the first equation of the first fitting formula and the first equation of the second fitting formula need to be trained.

[0050] For smooth fault data The data independent variables need to be transformed into the time domain and the differences between the two ends need to be taken before performing the above parameter identification and fitting operations. Specifically, for smooth fault data... A one-dimensional time-domain sequence is constructed with sampling time as the independent variable. First-order difference calculation is performed on adjacent points before and after the time-domain sequence to obtain new data. The above parameter identification and fitting operation is then performed on the new data.

[0051] Before data conversion, power relationships can be used to correct fault data under different fault conditions to reflect the impact of power variations on the results under the same fault condition. The relevant expressions are as follows: , in, This represents the data for the studied type of cylindrical mill under a certain work level. This indicates the data for the cylindrical mill at its rated power. Indicates working hours. Indicates the power amplification relationship. Indicates as working hours Relationship with power amplification The changing parameters to be fitted, This represents the result of Gaussian kernel function RBF interpolation applied to the residuals.

[0052] After the transformation, all fault data are normalized in the time domain for subsequent analysis. Each signal segment corresponding to each normalized fault data is trimmed; in this embodiment, it is trimmed to approximately 30 seconds to expand the fault data.

[0053] S4: Using the adaptive variational mode decomposition method, the expanded fault data is decomposed to obtain multiple intrinsic mode functions; preset screening conditions are set to retain intrinsic mode functions that meet the screening conditions, and the features of the retained intrinsic mode functions are extracted; the features extracted from the same intrinsic mode function are concatenated to form a two-dimensional matrix, and the two-dimensional matrix is ​​classified according to the fault condition category and fault level.

[0054] Adaptive variational mode decomposition method is used to transform periodic fault data. Each signal segment is decomposed into multiple intrinsic mode functions (IMFs), and preset screening conditions are used to retain IMFs that meet the screening conditions. In this embodiment of the invention, the screening conditions are set based on kurtosis values ​​and correlation with the original signal. IMFs are selected for retention based on these screening conditions, while high-frequency and ultra-high-frequency data are filtered out. Specifically, IMFs with high kurtosis values ​​and high correlation coefficients with the original signal are retained. In particular, IMFs with kurtosis values ​​in the top 10 and correlation coefficients in the top 30 of the original signal are retained.

[0055] It should be noted that, generally speaking, periodic fault data has a higher sampling frequency and greater fluctuations, while temperature-related fault data... The fault data is relatively smooth, with a low sampling frequency and inconspicuous frequency domain characteristics, so there is no need to use the adaptive variational mode decomposition method for processing.

[0056] Features such as kurtosis, peak factor, root mean square, center frequency, energy entropy, sample entropy, and skewness of each retained intrinsic mode function are extracted. Multiple IMF features obtained from a signal decomposition are concatenated into a two-dimensional matrix, which is then classified according to fault condition and fault level.

[0057] An adaptive node allocation method is used to compress and archive the original data, so that it can be reproduced promptly when needed, while retaining most of the information (e.g., over 90%). This can significantly reduce data storage space and improve data processing efficiency.

[0058] Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

[0059] The specific embodiments of the present invention described above do not constitute a limitation on the scope of protection of the present invention. Any other corresponding changes and modifications made in accordance with the technical concept of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A fault data migration device for a cylindrical mill, characterized in that, The fault data migration device is used to simulate the structure of a real cylindrical mill. The fault data migration device includes: It includes a slide rail base, a power mechanism, two I-beam supports, two end caps, multiple compartments, multiple sensor assemblies, and a liner. Two I-beam supports are respectively disposed at the left and right ends of the slide rail base and can move along the slide rail; two end caps are rotatably connected between the two I-beam supports via bearings, and a large gear is connected to the outer periphery of any one of the end caps; multiple compartments are spliced ​​end to end to form a hollow compartment, and two end caps are disposed at the left and right ends of the compartment to close the compartment; the liner is disposed on the inner surface of the compartment; the power mechanism is disposed on the slide rail base and can move along the slide rail; the power mechanism includes a small gear, which meshes with the large gear to transmit power; Adjust the distance between the two I-beam supports, and simultaneously increase or decrease the number of compartments and change the diameter of the compartments to change the size of the chamber; multiple reinforcing rods are provided on the compartments to adjust the strength of the chamber; the sensor assembly is installed at a preset position to collect mill operating data.

2. The fault data migration device for a cylindrical mill as described in claim 1, characterized in that, The power mechanism includes a motor, power mechanism fixing screws, reducer bracket, reducer, and pinion; The reducer bracket is connected to the slide rail by the power mechanism fixing screw. The motor and the reducer are mounted on the reducer bracket. The output shaft of the motor is connected to the input shaft of the reducer, and the output shaft of the reducer is connected to the pinion.

3. The cylindrical mill fault data migration device as described in claim 1, characterized in that, The fault data migration device also includes two bearing assemblies, each bearing assembly including a bearing housing, a bearing, a bearing inlet pipe, and a bearing outlet pipe; one bearing assembly is connected to each of the I-beam supports, and each bearing assembly contains a bearing for connecting the end cap; The bearing inlet pipe and the bearing outlet pipe are used to provide lubricating oil to the bearing.

4. The cylindrical mill fault data migration device as described in claim 3, characterized in that, The fault data migration device also includes a lubricating oil detection component, which includes a lubricating oil inlet pipe, a lubricating oil outlet pipe, and a core detection device. One end of the lubricating oil inlet pipe is connected to the bearing inlet pipe, and the other end passes through the core detection device and is connected to the lubricating oil injection system. One end of the lubricating oil outlet pipe is connected to the bearing outlet pipe, and the other end passes through the core detection device and is connected to the lubricating oil tank.

5. The fault data migration device for a cylindrical mill as described in claim 1, characterized in that, The head includes a first head flange, a head neck, a head disc, and a second head flange connected in sequence; the first head flange is connected to a corresponding bearing via a bearing connector, and the second head flange is connected to the compartment.

6. The cylindrical mill fault data migration device as described in claim 1, characterized in that, The compartment includes a first compartment flange, a compartment neck, and a second compartment flange connected in sequence, a compartment inner hole distributed circumferentially along the inner wall of the compartment neck, and a plurality of reinforcing rods distributed circumferentially along the compartment neck. The number, diameter, and position of the reinforcing rods are adjusted according to the required strength of the fault data migration device.

7. A method for acquiring fault data of a cylindrical mill, characterized in that, include: S1: Prepare a fault data migration device as described in any one of claims 1 to 5, classify fault conditions and obtain corresponding fault data; S2: Preprocess the fault data, the preprocessing including fault data classification and interpolation; S3: Expand the fault data using system identification methods; S4: The extended fault data is decomposed using the adaptive variational mode decomposition method to obtain multiple intrinsic mode functions; Preset filtering conditions, retain intrinsic mode functions that meet the filtering conditions, and extract the features of the retained intrinsic mode functions; Features extracted from the same intrinsic mode function are spliced ​​together to form a two-dimensional matrix, and the two-dimensional matrix is ​​classified according to the type of fault condition and the fault level.

8. The method for acquiring fault data of a cylindrical mill as described in claim 7, characterized in that, When preparing the fault data migration device, the strength calculation method for the fault data migration device is as follows: When a real cylindrical mill does not have reinforcing ribs: ， When a real cylindrical mill has reinforcing ribs: ， in, This represents the actual outer diameter of the cylindrical mill chamber. This represents the actual internal diameter of the cylindrical mill chamber. This indicates the ratio of the fault data migration device to the actual cylindrical mill. , and These represent the number, width, and height of the reinforcing ribs on a real cylindrical mill, respectively. Indicates the outer diameter of the cabin of the fault data migration device. This indicates the internal diameter of the fault data migration device. This indicates the number of reinforcing rods on the fault data migration device. This indicates the distance from the center of the reinforcing rod to the outside of the fault data migration device. This indicates the diameter of the reinforcing rod.

9. The method for acquiring fault data of a cylindrical mill as described in claim 7, characterized in that, The method for expanding the fault data is as follows: each signal segment corresponding to each normalized fault data is trimmed to achieve fault data expansion.

10. The method for acquiring fault data of a cylindrical mill as described in claim 7, characterized in that, The types of fault conditions include cracking at the root of the end cover journal, excessive wear of the liner, bearing failure, insufficient grinding steel balls, severe gear wear, mill cavitation, and loose bolts. The fault levels include minor faults, moderate faults, and severe faults.

Images