Precise dynamic control method for motorized spindle

Abstract

The invention provides a precise dynamic control method for an electric spindle, belongs to the technical field of electric spindle control, and aims at obtaining electric spindle fault data and fault triggering factors from the Internet to realize expansion of the fault data. Data of a vibration sensor, a temperature sensor, a rotating speed sensor and an axis displacement sensor of the motorized spindle are collected and compared with the fault library; compared data are transmitted to a data analysis and prediction module, working state data of the electric spindle at the next moment are predicted, whether faults occur at the next moment or not is judged, if no faults occur, whether the electric spindle is in the optimal working state or not is judged, and accurate control is achieved. According to the method, the fault database is richer by expanding the fault data and then carrying out semantic splitting and recognition, the change of possible parameters of the motorized spindle at the next moment can be pre-judged in advance during later fault collection, and then possible faults are matched according to the change of the parameters, so that the fault collection accuracy is improved. And meanwhile, optimal state identification is carried out when no fault is detected, so that the Euclidean control motorized spindle is in the optimal working state in real time, and meanwhile, the fault is avoided.

Description

Technical Field

[0001] The invention relates to the technical field of electric spindle control, and in particular to a precise dynamic control method of an electric spindle. Background Art

[0002] As the manufacturing industry develops towards high precision and high efficiency, traditional mechanical spindles are gradually being replaced by electric spindles due to their complex structure and large energy transmission loss. However, electric spindles will encounter problems such as thermal deformation and increased vibration when running at high speeds, which seriously affect the processing accuracy and surface quality. Therefore, how to effectively control these dynamic factors has become the key to improving the performance of electric spindles.

[0003] At present, there are two main measures to solve the vibration control of high-speed electric spindles. The first measure is to adopt high-speed precision bearing technology, precision machining and assembly technology, high-precision dynamic balancing technology, lubrication and cooling and other measures in the structural design stage to ensure that the high-speed spindle has good dynamic performance and thermal performance; the second measure is to adopt a series of passive vibration control technologies in the operation stage to suppress the vibration of the electric spindle or improve the dynamic stiffness performance of the high-speed electric spindle. Due to the limitation of its own technical level, the first measure cannot achieve satisfactory results by taking these measures in the design stage alone, and cannot solve the vibration problem of the electric spindle in the machine tool processing process. This is mainly because there are many factors affecting the vibration of the high-speed electric spindle in the machine tool processing process, and the systematic errors and random errors formed in the processing and manufacturing and equipment stages cannot be eliminated. The second measure usually adopts dynamic characteristic analysis technology and uses vibration isolators to suppress the vibration of key components such as machine tool bearing bases, worktable bases and complete machine bases. Although in recent decades, with the continuous development of new vibration theories and vibration isolation materials, various new vibration isolators have emerged one after another, but the inherent defects of the passive vibration control method based on vibration isolators make it impossible to effectively suppress or eliminate machine tool vibration noise. In addition, with the continuous improvement of machining accuracy requirements and the complexity of the vibration of the electric spindle of the machine tool, especially for precision and ultra-precision machining machine tools, the passive vibration control technology has been difficult to meet the requirements of machine tool vibration isolation and vibration reduction. Therefore, the existing technology cannot effectively suppress the bending vibration and torsional vibration of the electric spindle, and then ensure the balanced vibration of the electric spindle. The existing control detection cannot perform advance identification and cannot achieve advance prediction and control. Therefore, it is necessary to design a precise dynamic control method for the electric spindle. Summary of the invention

[0004] The purpose of the present invention is to provide a precise dynamic control method for an electric spindle to solve technical problems that cannot be solved by existing electric spindles.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] A precise dynamic control method for an electric spindle, the method comprising the following steps:

[0007] Step 1: Obtain the electric spindle fault data and fault-causing factors from the Internet to obtain the electric spindle fault summary data;

[0008] Step 2: Perform semantic recognition on the fault summary data to obtain fault semantic data;

[0009] Step 3: Search the fault semantic data for identical or similar semantic data from the Internet, and then aggregate them to obtain aggregated semantic data;

[0010] Step 4: Split the aggregated semantic data to obtain the smallest unit semantic word, then search the Internet for words that are the same or similar to the smallest unit semantic word, and aggregate them to obtain Internet expansion fault semantic word data;

[0011] Step 5: Combine any two or more minimum unit semantic words to obtain a new fault, combine all combined faults to obtain combined fault data, and form a fault library;

[0012] Step 6: Collect data from the vibration sensor, temperature sensor, speed sensor and axis displacement sensor of the electric spindle and compare them with the fault library;

[0013] Step 7: The compared data is transmitted to the data analysis and prediction module to predict the working state data of the electric spindle at the next moment, and determine whether a fault occurs at the next moment. If so, proceed to step 8. If not, determine whether the optimal control state is achieved. If so, return to step 6. If not, proceed to step 9.

[0014] Step 8: Issue a fault warning, and have the adaptive control system control and adjust the corresponding parameters that may cause the fault, and return to step 6;

[0015] Step 9: The adaptive control system enters a better control parameter mode and returns to step 6.

[0016] Furthermore, in step 1, when acquiring fault data, one fault corresponds to one or more fault-causing factors, and the fault and the fault-causing factors are stored as one piece of data, which is acquired from various websites or forums on the corresponding Internet through a crawler.

[0017] Furthermore, in step 2, semantic recognition is performed on each piece of fault information, firstly the semantics of the fault is recognized, and then the semantics of the fault-causing factor is recognized. When passing through, a mapping library is recommended to store the corresponding fault semantics and fault factor semantics.

[0018] Furthermore, in step 3, the semantics of each fault and the semantics of the fault factor are identified one by one, and put into the Internet to search for data with the same or similar meanings, wherein the fault semantics are put into the fault column, and the fault factors of the corresponding faults are searched at the same time. When searching for semantic data with the same or similar meanings as the fault factors, the searched fault factors are stored in the fault factor column, and the corresponding possible faults are listed and stored in the fault column, thereby completing the expansion of the fault data.

[0019] Furthermore, in step 4, semantic splitting is performed on the faults and fault-causing factors. When splitting, the minimum word is used as the splitting limit until it cannot be split any further. When splitting, the word is the smallest unit. The split words are then searched for identical or similar words on the Internet. When expanding semantic words, they are screened according to their application fields to avoid data expansion.

[0020] Furthermore, in step 5, when combining, first combine the words of the fault factors, and when searching for faults in the composed fault factors, all the faults originally associated with the composed words are used as the fault points of the composed fault factors, and then the composed fault factors and faults are re-matched to obtain a fault library.

[0021] Furthermore, in step 6, the vibration frequency, axial displacement, operating temperature and rotation speed of the electric spindle are collected, and the collected data are compared with the fault factors in the fault library to obtain a comparison result, which includes semantic similarity and similarity or identity of direction.

[0022] Furthermore, in step 7, the data analysis and prediction module is provided with an artificial intelligence neural network model, and the comparison result is input into the input section of the neural network. Then, the fault data corresponding to the compared or similar fault factors are used as the optimization reference object of the neural network, and then the boundaries are limited. Finally, the possible faults of the electric spindle in the future are output. If there is no fault, the working status is output according to the collected data.

[0023] Furthermore, in step 8, the control factors are compared with the control parameters according to the predicted fault factors, and then the adaptive control module adjusts the corresponding parameters, wherein the adjustment rule is to adjust the reverse fault factors of the fault, and at the same time, the possible fault data is transmitted to the management personnel.

[0024] Furthermore, in step 9, in the adaptive control module, the specific parameters of the electric spindle and the prepared material parameter data are originally input, and then based on the material characteristic data, it is determined whether the operation of the electric spindle has reached the material limit working state. If the working state of the electric spindle is not adjusted in the positive direction, if it is the optimal working state, the working state at this time is kept unchanged.

[0025] The present invention has the following beneficial effects due to the adoption of the above technical solution:

[0026] The present invention enriches the fault library by expanding the fault data and then performing semantic splitting and identification. In the later fault collection, the possible parameter changes of the electric spindle at the next moment can be predicted in advance, and then the possible faults can be matched according to the parameter changes. At the same time, the optimal state identification is performed when no fault is detected, so that the electric spindle can be controlled to be in the optimal working state in real time, while avoiding the occurrence of faults and realizing precise control. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is a flow chart of the method of the present invention;

[0028] Figure 2 It is a schematic diagram of the structure of the electric spindle of the present invention;

[0029] Figure 3 It is a schematic diagram of the heat dissipation structure of the present invention;

[0030] Figure 4 It is a schematic diagram of the gas path structure of the present invention;

[0031] Figure 5 It is a comparison chart of the speed and torque data of the present invention and the speed and torque data of the existing electric spindle.

[0032] In the accompanying drawings, 1-housing, 2-front bearing mounting seat, 3-front bearing, 4-chuck assembly, 5-rotating shaft, 6-sleeve, 7-stator assembly, 8-rotor assembly, 9-heat sink, 10-encoder, 11-rear bearing, 12-rear bearing mounting seat, 13-rear seat, 14-cylinder power part, 15-power ring, 16-mounting seat, 17-push rod release mechanism, 18-push rod. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and preferred embodiments. However, it should be noted that many details listed in the specification are only for the purpose of enabling the reader to have a thorough understanding of one or more aspects of the present invention, and these aspects of the present invention can be implemented even without these specific details.

[0034] like Figure 1 As shown, a precise dynamic control method of an electric spindle comprises the following steps:

[0035] Step 1: Obtain the electric spindle fault data and fault inducing factors from the Internet to obtain the electric spindle fault summary data. When obtaining fault data, one fault corresponds to one or more fault inducing factors, and the fault and the fault inducing factor are stored as a piece of data, which is obtained from various websites or forums on the corresponding Internet through crawlers.

[0036] Step 2: Perform semantic recognition on the fault summary data to obtain fault semantic data. Perform semantic recognition on each fault information, first recognize the semantics of the fault, then recognize the semantics of the fault-causing factor, and recommend a mapping library to store the corresponding fault semantics and fault factor semantics.

[0037] Step 3: Search the fault semantic data for the same or similar semantic data from the Internet, and then summarize them to obtain the summarized semantic data. Identify the semantics of each fault and the semantics of the fault factor one by one, and put them into the Internet to search for data with the same or similar meanings. The fault semantics are put into the fault column, and the fault factors of the corresponding faults are searched at the same time. When searching for semantic data with the same or similar meanings as the fault factors, the fault factors found are stored in the fault factor column, and the corresponding possible faults are listed and stored in the fault column to complete the expansion of the fault data.

[0038] Step 4: Split the aggregated semantic data to obtain the smallest unit semantic word, and then search the Internet for words that are the same or similar to the smallest unit semantic word, and aggregate to obtain Internet expanded fault semantic word data. Semantically split the fault and the fault-causing factors. When splitting, use the smallest word as the splitting limit until it cannot be split any further. When splitting, the word is the smallest unit, and then search the Internet for the same or similar words after the split. When expanding the semantic words, screen them according to the application field of the words to avoid data expansion.

[0039] Step 5: Combine any two or more minimum unit semantic words to obtain a new fault, combine all combined faults to obtain combined fault data, and form a fault library. When combining, first combine the words of the fault factors, and when searching for faults in the composed fault factors, the faults originally associated with the composed words are used as the fault points of the composed fault factors, and then the composed fault factors and faults are re-associated to obtain the fault library.

[0040] Step 6: Collect the vibration sensor, temperature sensor, speed sensor and shaft displacement sensor data of the electric spindle and compare them with the fault library. Collect the vibration frequency, shaft displacement, working temperature and rotation speed of the electric spindle, compare the collected data with the fault factors in the fault library, and obtain the comparison results, which include semantic similarity and similarity or identity of direction.

[0041] Step 7: The compared data is transmitted to the data analysis and prediction module, and the working state data of the electric spindle at the next moment is predicted, and whether there is a fault at the next moment is judged. If yes, it goes to step 8, and if not, it goes to the optimal control state. If yes, it returns to step 6, and if not, it goes to step 9. The data analysis and prediction module is provided with an artificial intelligence neural network model, and the comparison result is input into the input of the neural network, and then the fault data corresponding to the compared or similar fault factors are used as the optimization reference object of the neural network, and then the boundary is limited, and finally the possible fault of the electric spindle in the next moment is output, and if there is no fault, the working state is output according to the collected data.

[0042] Step 8: Issue a fault warning, and have the adaptive control system control and adjust the corresponding parameters that may cause the fault, and return to step 6. According to the predicted fault factors, the control factors are compared with the control parameters, and then the adaptive control module adjusts the corresponding parameters. The adjustment rule is to adjust the reverse fault factors of the fault, and at the same time, the possible fault data is transmitted to the management personnel.

[0043] Step 9: The adaptive control system controls to enter a better control parameter mode and returns to step 6. In the adaptive control module, the specific parameters of the electric spindle and the prepared material parameter data are originally input, and then according to the material characteristic data, it is judged whether the working state of the electric spindle has reached the material limit working state. If the working state of the electric spindle is not adjusted in the positive direction, if it is the optimal working state, the working state at this time is kept unchanged.

[0044] like Figure 2-4 As shown, the electric spindle includes a spindle mechanism, a heat dissipation component and an electric drive component. The heat dissipation component is arranged at one end of the spindle mechanism and embedded in the spindle mechanism. The electric drive component is arranged at one end of the heat dissipation component and passes through the spindle mechanism and the heat dissipation component. The spindle mechanism mainly plays the role of driving the rotation, the heat dissipation component mainly dissipates the heat inside the spindle mechanism, and then the electric drive component mainly realizes the forward and backward movements.

[0045] The spindle mechanism includes a housing 1, a front bearing mounting seat 2, a front bearing 3, a motor assembly, a rotating shaft 5, a rear bearing 11 and a rear bearing mounting seat 12. The front bearing mounting seat 2 and the rear bearing mounting seat 12 are respectively arranged at the two ends of the housing 1, the motor assembly is arranged inside the housing 1, the front bearing 3 is arranged in the front bearing mounting seat 2, the rear bearing 11 is arranged in the rear bearing mounting seat 12, the rotating shaft 5 passes through the interior of the motor assembly, and the two ends are respectively arranged on the front bearing 3 and the rear bearing 11. The motor assembly includes a stator assembly 7, a rotor assembly 8 and a sleeve 6. The stator assembly 7 is arranged on the outside of the rotor assembly 8, the sleeve 6 is arranged between the stator assembly 7 and the rotor assembly 8, and the rotating shaft 5 is arranged inside the rotor assembly 8.

[0046] An encoder 10 is arranged on the rotating shaft 5 . The rotating shaft 5 is a hollow shaft. A chuck assembly 4 is arranged at the front end of the rotating shaft 5 .

[0047] The sheath 6 is a carbon fiber sheath, and both the front bearing 3 and the rear bearing 11 use magnetic bearings. The interference between the rotor carbon fiber sheath and the permanent magnet is 0.06mm, and the force on the permanent magnet is far lower than its tensile strength (the tensile strength of RuFeB is (80-140MPa).

[0048] The heat dissipation assembly includes a heat dissipation box 9 and a cooling structure. The heat dissipation box 9 is arranged at both ends of the cooling structure, and the cooling structure is arranged in the housing of the stator assembly 7. The cooling structure is arranged as a Z-shaped annular water jacket cooling structure, and the water jacket cooling structure is connected to the heat dissipation box 9 to keep the average temperature of the motor below 60 degrees Celsius. The stator housing Z-shaped annular water jacket cooling structure is adopted, and the size of the water channel is optimized to take away heat with maximum efficiency. The electric drive assembly includes a rear seat 13, a cylinder power part 14, a power ring 15, a mounting seat 16, a push rod release mechanism 17 and a push rod 18. The power ring 15 is arranged on the cylinder power part 14, and the push rod release mechanism 17 is connected to the power ring 15 and the push rod 18. One end of the push rod 18 is arranged on the power ring 15, and the mounting seat 16 is arranged at the rear end of the cylinder power part 14. The rear seat 13 is arranged on the heat dissipation assembly, and the push rod release mechanism 17 is arranged on the rear seat 13. The push rod 18 passes through the rear seat 13 and is arranged in the rotating shaft 5.

[0049] In the embodiment of the present invention, the electric spindle is designed by using several physical field joint simulation designs, including electromagnetic design, heat dissipation design and mechanical mechanics simulation design, and the method of harmonic pole clipping and harmonic injection is used to improve the torque density of the permanent magnet motor system. The coupling relationship between the motor and the controller is fully considered to achieve high-speed precision dynamic control and improve the overall efficiency of the motor system. Through the refined analysis of iron loss and optimization of size, the efficiency of the electric spindle is improved, and a high-efficiency and energy-saving electric spindle brand is created.

[0050] like Figure 5 As shown, Figure 5 B is the result data of this application, Figure 5 A is the data of the existing Xifeng brand motor electric spindle. The torque of this application is much larger than that of the existing electric spindle, and the speed is also higher, achieving a step-by-step breakthrough.

[0051] The rated voltage (Vdc) of the electric spindle is 208, the continuous power (kW) is 2.0, the maximum dynamic runout (μm) is 7 (150,000 rpm), the continuous torque (Nm) is 0.1273, the continuous load (Nm) is 2000, the rated speed (rpm) is 150000, the maximum speed (rpm) is 160000, the operating temperature range (℃) is -40~+65, the stall torque (Nm) is 7.47, the maximum clamp torque (Nm) is 2.3, the maximum leakage current (mA) is 0.01, the reference mass (kg) is 2.5, and the reference size (mm) is Φ50×200.

[0052] Matters not covered by the present invention are known technologies.

[0053] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A precise dynamic control method for an electric spindle, characterized in that: The method comprises the following steps: Step 1: Obtain the electric spindle fault data and fault-causing factors from the Internet to obtain the electric spindle fault summary data; Step 2: Perform semantic recognition on the fault summary data to obtain fault semantic data; Step 3: Search the fault semantic data for identical or similar semantic data from the Internet, and then aggregate them to obtain aggregated semantic data; Step 4: Split the aggregated semantic data to obtain the smallest unit semantic word, then search the Internet for words that are the same or similar to the smallest unit semantic word, and aggregate them to obtain Internet expansion fault semantic word data; Step 5: Combine any two or more minimum unit semantic words to obtain a new fault, combine all combined faults to obtain combined fault data, and form a fault library; Step 6: Collect data from the vibration sensor, temperature sensor, speed sensor and shaft displacement sensor of the electric spindle and compare them with the fault library; Step 7: The compared data is transmitted to the data analysis and prediction module to predict the working state data of the electric spindle at the next moment, and determine whether a fault occurs at the next moment. If so, proceed to step 8. If not, determine whether the optimal control state is achieved. If so, return to step 6. If not, proceed to step 9. Step 8: Issue a fault warning, and have the adaptive control system control and adjust the corresponding parameters that may cause the fault, and return to step 6; Step 9: The adaptive control system enters a better control parameter mode and returns to step 6.

2. The precise dynamic control method of an electric spindle according to claim 1, characterized in that: In step 1, when acquiring fault data, one fault corresponds to one or more fault-causing factors, and the fault and the fault-causing factors are stored as one piece of data, which is acquired from various websites or forums on the corresponding Internet through a crawler.

3. The precise dynamic control method of an electric spindle according to claim 1, characterized in that: In step 2, semantic recognition is performed on each fault information. First, the semantics of the fault is recognized, and then the semantics of the fault-causing factor is recognized. A mapping library is recommended to store the corresponding fault semantics and fault factor semantics.

4. The precise dynamic control method of an electric spindle according to claim 1, characterized in that: In step 3, the semantics of each fault and the semantics of the fault factor are identified one by one, and put into the Internet to search for data with the same or similar meanings. Among them, the fault semantics are put into the fault column, and the fault factors of the corresponding faults are searched at the same time. When searching for semantic data with the same or similar meanings as the fault factors, the fault factors found are stored in the fault factor column, and the corresponding possible faults are listed and stored in the fault column to complete the expansion of the fault data.

5. The precise dynamic control method of an electric spindle according to claim 1, characterized in that: In step 4, semantic splitting is performed on the faults and fault-causing factors. When splitting, the smallest word is used as the splitting limit until it cannot be split any further. When splitting, the word is the smallest unit. The split words are then searched for identical or similar words on the Internet. When expanding semantic words, they are screened according to their application fields to avoid data expansion.

6. The precise dynamic control method of an electric spindle according to claim 1, characterized in that: In step 5, when combining, first combine the words of the fault factors, and when searching for faults in the composed fault factors, all the faults originally associated with the composed words are used as the fault points of the composed fault factors, and then the composed fault factors and faults are re-matched to obtain a fault library.

7. The precise dynamic control method of an electric spindle according to claim 1, characterized in that: In step 6, the vibration frequency, axis displacement, operating temperature and rotation speed of the electric spindle are collected, and the collected data are compared with the fault factors in the fault library to obtain a comparison result, which includes semantic similarity and similarity or identity of direction.

8. The precise dynamic control method of an electric spindle according to claim 1, characterized in that: In step 7, the data analysis and prediction module is provided with an artificial intelligence neural network model, and the comparison result is input into the input section of the neural network. Then, the fault data corresponding to the compared or similar fault factors are used as the optimization reference object of the neural network, and then the boundaries are limited. Finally, the possible faults of the electric spindle in the future are output. If there is no fault, the working status is output according to the collected data.

9. The precise dynamic control method of an electric spindle according to claim 1, characterized in that: In step 8, according to the predicted fault factors, the control factors are compared with the control parameters, and then the adaptive control module adjusts the corresponding parameters. The adjustment rule is to adjust the reverse fault factors of the fault, and at the same time, the possible fault data is transmitted to the management personnel.

10. The precise dynamic control method of an electric spindle according to claim 1, characterized in that: In step 9, in the adaptive control module, the specific parameters of the electric spindle and the prepared material parameter data are originally input, and then based on the material characteristic data, it is judged whether the working state of the electric spindle has reached the material limit working state. If the working state of the electric spindle is not adjusted in the positive direction, if it is the optimal working state, the working state at this time is kept unchanged.

Images