Machine tool spindle temperature measurement method and system

Abstract

The present application relates to precision machine tool control technical field, especially machine tool spindle temperature measurement method and system, method includes: real-time acquisition spindle drive motor current signal and switching frequency data, analysis spindle each component's heat power distribution;Based on the physical structure layout of each component of the spindle, the spindle is decomposed into a thermal node network, combined with the heat power distribution, the temperature rise prediction value of each node is obtained;The actual temperature value of the key node is obtained, when the deviation of the temperature rise prediction value and the actual temperature value of any node exceeds the set threshold, the corresponding node is compensated and corrected;According to the corrected thermal node network, the periodic temperature rise gradient in the future machining path is predicted, and based on the workpiece geometric error obtained by in-situ measurement, the thermal resistance parameter in the thermal node network is corrected reversely. Through the present application, the problems such as temperature rise prediction distortion caused by disconnection of mechanical parameters, lack of temperature perception in closed area and dynamic working condition response lag of traditional temperature measurement technology are effectively solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of precision machine tool control, and in particular to a machine tool spindle temperature measurement method and system. BACKGROUND

[0002] The temperature rise of a machine tool spindle due to bearing friction, motor loss and other factors during high-speed machining can cause thermal deformation, directly causing machining errors, and has become a core problem restricting precision manufacturing. Traditional temperature measurement techniques generally have three major defects: disconnection of mechanical system thermal characteristics, internal thermal state perception blind area, and dynamic response lag, resulting in insufficient thermal error compensation accuracy.

[0003] In traditional temperature measurement techniques, the installation of contact sensors can disrupt the dynamic balance of the spindle, and the temperature in closed areas such as the bearing interior cannot be monitored. Non-contact infrared temperature measurement is susceptible to interference from cooling liquid atomization and is difficult to capture the spatial and temporal gradient changes in the temperature field. Fiber Bragg grating sensors can achieve distributed measurement, but require customized modification of the spindle structure and are affected by long-term stability due to vibration. Indirect thermal estimation models based on experimental data lack mechanical parameter coupling, resulting in significant prediction errors under dynamic conditions such as variable speed and variable load, and are unable to locate local hot spots in real time.

[0004] The information disclosed in this BACKGROUND section is only intended to enhance the understanding of the general background of the present disclosure and is not intended to be recognized or implied in any form that this information constitutes prior art known to those of ordinary skill in the art. SUMMARY

[0005] The present application provides a machine tool spindle temperature measurement method and system, which can effectively solve the problems in the background art.

[0006] To achieve the above purpose, the technical solution adopted by the present application is:

[0007] A machine tool spindle temperature measurement method, the method comprising:

[0008] Real-time acquisition of current signal and switching frequency data of the spindle drive motor, and analysis of the heat power distribution of each component of the spindle;

[0009] Based on the physical structure layout of each component of the spindle, the spindle is decomposed into a thermal node network, and in combination with the heat power distribution, the temperature rise prediction value of each node is obtained;

[0010] The actual temperature value of the key node is obtained, and when the deviation between the temperature rise prediction value and the actual temperature value of any node exceeds a set threshold, temperature compensation correction is performed on the corresponding node;

[0011] According to the modified thermal node network, the periodic temperature rise gradient in the future machining path is predicted, and the thermal resistance parameters in the thermal node network are corrected in reverse based on the workpiece geometric error obtained by in-situ measurement.

[0012] Further, the spindle is decomposed into a thermal node network based on the physical structural layout of each component of the spindle, including:

[0013] According to the actual physical contact area of the bearing inner and outer ring raceway contact surface, the shaft core and the tool clamping surface, the spindle is divided into three core thermal nodes of bearing thermal node, shaft core conduction thermal node and tool thermal node;

[0014] Based on the core thermal nodes, thermal conduction topology links are generated between adjacent nodes, wherein the link density between the bearing thermal node and the shaft core conduction thermal node is determined by the proportion of the bearing roller contact pressure distribution;

[0015] The real-time rotational speed of the spindle and the flow state of the cooling liquid are converted into the thermal resistance parameters of the bearing thermal node and the tool thermal node, and are updated to the thermal conduction topology links;

[0016] According to the deviation of the actual temperature value and the current temperature rise prediction value, the link density and the thermal resistance parameters are adjusted in reverse until the prediction accuracy of the thermal node network meets the preset convergence condition.

[0017] Further, the spindle temperature measurement is based on the thermal node network, including:

[0018] The actual temperature value of the core thermal node is extracted from the thermal conduction topology link, and is time-domain aligned with the current harmonic characteristics of the spindle driving motor;

[0019] The actual temperature value is input into a pre-trained time series prediction model to generate a temperature rise trend curve of each core thermal node in the future machining period, and the amplitude and phase of the temperature rise trend curve are dynamically adjusted according to the current task executed by the spindle;

[0020] Based on the workpiece geometric error and the temperature rise trend curve, the maximum deviation area of the thermal resistance parameters in the thermal node network is located, and the directional reinforcement of the link density is triggered;

[0021] The link reinforcement result is synchronized to the thermal conduction topology link, and the above steps are repeatedly executed at a preset time interval until the residual error of the temperature rise prediction value is less than the convergence threshold value for three consecutive times.

[0022] Further, triggering the directional reinforcement of the link density includes:

[0023] When any axial deviation of the workpiece geometric error is detected to continuously exceed the tolerance range, a temperature gradient change feature of a corresponding axial direction in the temperature rise trend curve is extracted and mapped and matched with the thermal resistance parameter of the thermal node network;

[0024] According to the matching result, a weak area of a thermal conduction topological link in the shaft core conduction thermal node or the tool thermal node is located, and a target link density increment of the weak area is calculated in combination with a current bearing roller contact pressure distribution;

[0025] Based on the target link density increment, a distribution mode of the thermal conduction topological link of the weak area is dynamically adjusted, and convergence of the workpiece geometric error after adjustment is verified in the next machining period;

[0026] If the convergence does not meet the expectation, the target link density increment is iteratively improved until a residual error of the temperature rise prediction value of the thermal node network and the workpiece geometric error fall in a preset interval.

[0027] Further, obtaining the temperature rise prediction value of each node includes:

[0028] Based on a thermal conductivity coefficient of the main shaft material and a proportion of the bearing roller contact pressure distribution, the thermal resistance parameter of each thermal conduction topological link is determined;

[0029] According to a real-time rotating speed of the main shaft and a cooling liquid flow state, a convection heat transfer parameter of each core thermal node along the thermal conduction topological link is obtained;

[0030] The heat power distribution is injected into the corresponding core thermal node, and a temperature rise prediction value of each core thermal node is calculated based on the thermal resistance parameter and the convection heat transfer parameter;

[0031] When a deviation between the temperature rise prediction value and the actual temperature value exceeds a set threshold value, temperature compensation correction of the core thermal node is performed.

[0032] Further, determining the thermal resistance parameter of each thermal conduction topological link includes:

[0033] A bearing inner and outer ring raceway contact area of the main shaft is decomposed into a plurality of contact sub-areas, and each contact sub-area corresponds to a Hertz contact pressure distribution range of a single bearing roller;

[0034] According to the thermal conductivity coefficient of the main shaft material and an assembly pre-tightening force detection value, a contact thermal resistance reference value of each contact sub-area is calculated;

[0035] Based on a contact pressure distribution of the bearing roller along the circumferential direction, a conduction thermal resistance parameter between adjacent contact sub-areas is dynamically allocated;

[0036] If the deviation of the temperature rise prediction value corresponding to the conduction thermal resistance parameter from the actual temperature value exceeds a set threshold, the boundary division of the contact sub-area is adjusted in reverse, and the conduction thermal resistance parameter is recalculated.

[0037] Further, a temperature compensation correction of the core thermal node is performed, including:

[0038] When it is detected that the deviation of the temperature rise prediction value of the bearing thermal node or the tool thermal node exceeds a threshold, the geometric characteristics and spindle load state of the current machining path are extracted, and a compensation gain coefficient positively correlated with the curvature of the machining trajectory is generated;

[0039] According to the compensation gain coefficient, the conduction thermal resistance weight of the corresponding node is dynamically increased, and the convection heat transfer parameter of the adjacent node is simultaneously reduced;

[0040] Based on the conduction thermal resistance weight and the convection heat transfer parameter, the temperature rise prediction value of the core thermal node is recalculated, and the compensation effect is verified through in-situ temperature measurement;

[0041] If the residual error of the temperature rise prediction value after compensation is still higher than the threshold, the proportional relationship between the compensation gain coefficient and the conduction thermal resistance weight is iteratively adjusted until the prediction accuracy of the thermal node network recovers to a set range.

[0042] Further, the heat power distribution of each component of the spindle is analyzed, including:

[0043] Based on the stator resistance parameter of the spindle motor, the stator copper loss power is calculated, and the rotor eddy current loss power caused by harmonic current is calculated;

[0044] According to the product relationship of the real-time rotational speed and the load torque of the spindle, the bearing friction loss power is calculated in combination with the preset bearing friction coefficient;

[0045] Based on the switching frequency data of the inverter and the DC bus voltage, the total amount of switching loss is calculated;

[0046] Based on the physical structure thermal capacity characteristics of the spindle assembly, the stator copper loss power, the rotor eddy current loss power, the bearing friction loss power, and the total amount of switching loss are distributed to the bearing thermal node, the shaft core conduction thermal node, and the tool thermal node.

[0047] A machine tool spindle temperature measurement system, the system comprising:

[0048] An information acquisition module, which acquires the current signal and switching frequency data of the spindle driving motor in real time, and analyzes the heat power distribution of each component of the spindle;

[0049] The temperature rise prediction module decomposes the spindle into a thermal node network based on the physical structure layout of each component of the spindle, and obtains the temperature rise prediction value of each node in combination with the heat power distribution;

[0050] The node compensation module obtains the actual temperature value of the key node, and performs temperature compensation correction on the corresponding node when the deviation between the temperature rise prediction value and the actual temperature value of any node exceeds the set threshold value.

[0051] The thermal resistance correction module predicts the periodic temperature rise gradient in the future machining path according to the corrected thermal node network, and reversely corrects the thermal resistance parameter in the thermal node network based on the workpiece geometric error obtained by in-situ measurement.

[0052] Further, the temperature rise prediction module comprises:

[0053] The thermal resistance determination unit determines the thermal resistance parameter of each thermal conduction topology link based on the thermal conductivity coefficient of the spindle material and the proportion of the bearing roller contact pressure distribution;

[0054] The state acquisition unit acquires the convective heat transfer parameter of each core thermal node along the thermal conduction topology link according to the real-time rotating speed of the spindle and the cooling liquid flow state;

[0055] The temperature rise calculation unit injects the heat power distribution into the corresponding core thermal node, and calculates the temperature rise prediction value of each core thermal node based on the thermal resistance parameter and the convective heat transfer parameter;

[0056] The compensation detection unit performs temperature compensation correction on the core thermal node when it is detected that the deviation between the temperature rise prediction value and the actual temperature value exceeds the set threshold value.

[0057] The technical scheme of the present application can achieve the following technical effects:

[0058] The present application realizes high-precision reconstruction of the spindle global temperature field and thermal error feedforward compensation, reduces the workpiece geometric error in high-speed heavy-load machining, and reduces the sensor deployment requirement through dynamic optimization of the thermal resistance parameter, thereby improving the stability and service life of the machine tool spindle under complex working conditions.

[0059] The above description is only a summary of the technical scheme of the present application, in order to more clearly understand the technical means of the present application, the specific embodiments of the present application can be implemented in accordance with the content of the description, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0060] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings described below only represent some of the embodiments of the present application, and other drawings can also be obtained by those skilled in the art without any creative effort.

[0061] Figure 1 Flowchart of the machine tool spindle temperature measurement method;

[0062] Figure 2 Schematic diagram of decomposing the spindle into a thermal node network;

[0063] Figure 3 Schematic diagram of obtaining the temperature rise prediction value of each node. DETAILED DESCRIPTION

[0064] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments only represent some of the embodiments of the present application, but not all the embodiments.

[0065] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0066] Embodiment one;

[0067] As shown in Figure 1 The present application provides a machine tool spindle temperature measurement method, which comprises:

[0068] S10: Real-time acquisition of current signal and switching frequency data of the spindle driving motor, and analysis of the heat power distribution of each component of the spindle;

[0069] S20: Decomposition of the spindle into a thermal node network based on the physical structure layout of each component of the spindle, and obtaining the temperature rise prediction value of each node in combination with the heat power distribution;

[0070] S30: Acquisition of the actual temperature value of the key node, and temperature compensation correction of the corresponding node when the deviation between the temperature rise prediction value and the actual temperature value of any node exceeds the set threshold value;

[0071] S40: Prediction of the periodic temperature rise gradient in the future machining path according to the corrected thermal node network, and reverse correction of the thermal resistance parameter in the thermal node network based on the workpiece geometric error obtained by in-situ measurement.

[0072] Specifically, firstly, the operating parameters of the spindle drive motor are acquired in real time, including three-phase current signals and switching frequency data. These data reflect the motor's load condition and operating status, and can be used to estimate the heat power distribution of various components of the spindle. The current signal can be used to calculate the copper loss heat power of the motor windings, while the switching frequency data helps to assess the switching loss heat power of the inverter or controller. In addition, based on the spindle's structure and operating parameters, the frictional heat power of components such as bearings and couplings can be estimated to obtain the heat power distribution. Based on the physical structure layout of each component of the spindle, the spindle is divided into multiple thermal nodes, and a thermal equivalent network model, i.e., a thermal node network, is established. Each thermal node represents a key part of the spindle, such as the front bearing, motor rotor, motor stator, intermediate shaft section, rear bearing, and spindle housing. These thermal nodes are connected by thermal resistance and thermal capacitance to form a thermal network model. Combined with the aforementioned heat power distribution, the heat... The node network can predict the temperature rise of each node, thereby obtaining the temperature distribution of the spindle. To improve the accuracy of temperature prediction, several key nodes are selected in the spindle structure, and the actual temperature data of these locations are collected in real time by embedded or mounted temperature sensors. The collected actual temperature values ​​are compared with the predicted values. If the deviation exceeds a preset threshold, it is determined that there is an error in the current thermal network model, and the heat source or thermal resistance parameters of the corresponding nodes are compensated and corrected. The correction method can use algorithms such as extended Kalman filter or least squares method to optimize the model parameters. During the processing, combined with the processing path control system, the current load cycle caused by the processing task is obtained in advance, and the temperature rise gradient change of future cycles is predicted based on the current corrected thermal node network. In addition, combined with the geometric error of the workpiece after processing obtained by in-situ measurement, the thermal resistance error in the thermal network is identified in reverse, and the spindle thermal model is optimized again to make it more in line with the actual thermal behavior.

[0073] The technical solution of this invention achieves high-precision reconstruction of the full-domain temperature field of the spindle and thermal error feedforward compensation, reducing workpiece geometric errors in high-speed heavy-duty machining. At the same time, dynamic optimization of thermal resistance parameters reduces the need for sensor deployment, improving the stability and service life of the machine tool spindle under complex working conditions.

[0074] Furthermore, such as Figure 2 As shown, based on the physical structure layout of each component of the spindle, the spindle is decomposed into a network of hot nodes, including:

[0075] Based on the actual physical contact area of ​​the inner and outer raceway contact surfaces of the spindle bearing and the clamping surface between the spindle core and the tool, the spindle is divided into three core thermal nodes: bearing thermal node, spindle core thermal node, and tool thermal node.

[0076] Based on the core thermal node, a thermal conduction topology link is generated between adjacent nodes, wherein the link density between the bearing thermal node and the shaft core conduction thermal node is determined by the proportion of the bearing roller contact pressure distribution;

[0077] The main shaft real-time rotating speed and the cooling liquid flow state are converted into the thermal resistance parameters of the bearing thermal node and the tool thermal node, and are updated to the thermal conduction topology link;

[0078] According to the deviation between the actual temperature value and the current temperature rise prediction value, the link density and the thermal resistance parameter are adjusted reversely until the thermal node network prediction accuracy meets the preset convergence condition.

[0079] As a preferred embodiment of the above, on the basis of the traditional thermal node model, the structure contact physical parameters are further introduced, and the main shaft is divided into three core thermal nodes, i.e. bearing thermal node, shaft core conduction thermal node and tool thermal node. The bearing thermal node refers to the heat concentration area formed by the bearing parts at both ends of the main shaft, which usually has a stable raceway contact surface. The shaft core conduction thermal node is located in the middle section of the main shaft and mainly undertakes the heat conduction path of the motor heat to the outside through the shaft core. The tool thermal node is concentrated between the end of the main shaft and the tool clamping surface, which is an important interface for direct heat transfer between the main shaft and the workpiece. This thermal node division method is based on the main contact interface in the real physical structure, and thus has more practical significance and heat path representativeness. On the basis of this structure division, the heat conduction topology link is constructed by analyzing the heat transfer relationship between each thermal node. The topology connection strength between the bearing thermal node and the shaft core conduction thermal node, i.e. the link density, is set according to the proportion of the contact pressure distribution of the rollers inside the bearing and the inner and outer rings. The contact area and the unit heat flux density can be weighted according to the pressure distribution curve obtained by simulation or actual measurement under different load conditions of the main shaft, so as to generate a link density value with physical significance. This method effectively improves the response ability of the model to the change of the bearing heat conduction capacity under different operating conditions, especially in high load, high speed machining and other conditions prone to heat abnormities. In order to improve the real-time and dynamic adaptability of the thermal network, the current real-time speed information of the main shaft and the flow state of the cooling liquid are converted into dynamic thermal resistance parameters of the bearing thermal node and the tool thermal node. For example, when the main shaft rotates at high speed, the contact area of the bearing will expand due to the increase of centrifugal force, and the thermal resistance will decrease accordingly. For example, when the flow rate of the cooling liquid increases, the heat carried away increases, and the thermal resistance of the node also decreases. Timely update of these dynamic thermal resistance parameters in the heat conduction topology link ensures that the thermal model always reflects the current real heat conduction state. In the running process, the actual temperature data of the key nodes are continuously obtained through the foregoing temperature measurement mechanism, and are compared with the predicted values of the thermal node network model. When a significant deviation is found between the actual temperature and the predicted temperature, and the deviation exceeds the set deviation threshold, the link adaptive adjustment mechanism will be triggered. The mechanism will adjust the link density and thermal resistance parameters between the corresponding thermal nodes in the opposite direction according to the error direction and size. The adjustment process continues until the error between the predicted value and the actual value converges to the set allowable range, ensuring the prediction accuracy of the thermal node network.

[0080] Further, the main shaft temperature measurement based on the thermal node network comprises:

[0081] The actual temperature value of the core thermal node is extracted from the heat conduction topology link, and is time-domain aligned with the current harmonic characteristics of the main shaft driving motor;

[0082] input the actual temperature value into the pre-trained time series prediction model to generate the temperature rise trend curve of each core thermal node in the future machining period, and dynamically adjust the amplitude and phase of the temperature rise trend curve according to the current task executed by the spindle;

[0083] Based on the workpiece geometric error and the temperature rise trend curve, the maximum deviation area of the thermal resistance parameter in the thermal node network is located, and the directional reinforcement of the link density is triggered;

[0084] The link reinforcement result is synchronized to the thermal conduction topology link, and the above steps are repeatedly executed at a preset time interval until the residual error of the temperature rise prediction value for three consecutive times is less than the convergence threshold.

[0085] As a preferred embodiment of the above embodiment, first, the thermal conduction topology link is established based on the spindle structure and the heat conduction path, and a plurality of key parts are taken as core thermal nodes. The nodes are connected by thermal resistance to simulate the actual heat flow distribution relationship. In the running process, the real-time temperature data of each core thermal node is obtained through the temperature sensor arranged on each core thermal node. At the same time, in order to more accurately reflect the dynamic change of the spindle thermal load, the three-phase current signals of the spindle driving motor are synchronously collected, and the current harmonic components are extracted by using digital signal processing technology, especially the characteristic harmonic components reflecting the change of the thermal load. The above temperature data and current harmonic characteristics are aligned based on a unified time reference in the main controller to ensure the time domain consistency of the input data. The thermal node temperature sequence after time domain alignment is taken as input and sent into a time series prediction model constructed and pre-trained based on long short-term memory neural network (LSTM) to output the temperature rise trend curve of each thermal node in the future machining period. In order to improve the task adaptability of the trend curve, the amplitude and phase of the model prediction result are dynamically adjusted based on the information of the current task executed by the spindle, so that the prediction result is closer to the thermal behavior under the current actual working condition. In order to optimize the accuracy of temperature rise prediction, after collecting the workpiece machining result, the geometric error is analyzed and used as a model reverse correction signal for joint analysis with the temperature rise trend curve to identify the area with the largest deviation of the thermal resistance parameter in the current thermal node network. After locating the deviation area, the directional reinforcement of the thermal conduction network is triggered by increasing the link density in the area. The reinforcement operation is based on the network connection weight updating mechanism to improve the response ability to abnormal changes of thermal resistance. The link density reinforcement result is synchronized back to the original thermal conduction topology link in real time for updating the subsequent thermal conduction modeling. In the machining process, the above process is repeatedly executed at a fixed time interval, and the residual error of the trend curve prediction value and the actual sampling temperature is evaluated. When the residual error for three consecutive times is less than the set convergence threshold, it is considered that the model has stabilized, and the temperature rise monitoring maintenance state is entered.

[0086] Further, the directional reinforcement of the link density includes:

[0087] When any axial deviation of the workpiece geometric error is detected to continuously exceed the tolerance range, the temperature gradient change feature of the corresponding axial in the temperature rise trend curve is extracted and mapped and matched with the thermal resistance parameters of the thermal node network;

[0088] According to the matching result, the weak area of the thermal conduction topological link in the shaft core conduction thermal node or the tool thermal node is located, and the target link density increment of the weak area is calculated in combination with the current bearing roller contact pressure distribution;

[0089] Based on the target link density increment, the distribution mode of the thermal conduction topological link of the weak area is dynamically adjusted, and the convergence of the adjusted workpiece geometric error in the next machining cycle is verified;

[0090] If the convergence does not meet the expectation, the target link density increment is iteratively improved until the residual of the temperature rise prediction value of the thermal node network and the workpiece geometric error fall in the preset interval.

[0091] As a preferred embodiment of the above-mentioned embodiment, three-dimensional geometric data of the workpiece is continuously collected during the main shaft machining process, and error analysis is performed on the machined workpiece using a high-precision probe or an online measurement device, with particular attention paid to dimensional deviations in the axial direction (such as the Z-axis, X-axis, or Y-axis). When the geometric error value in any axial direction is detected to exceed the set tolerance range for multiple consecutive machining cycles, the initiation of the link density strengthening mechanism is triggered. At this time, the temperature rise trend curve related to the axial direction is extracted, and the temperature gradient change characteristics in the time domain and spatial domain are analyzed, including the position, direction, and rate of change of the temperature gradient abrupt change point. A mapping relationship is established between the temperature gradient change characteristics obtained in the temperature rise trend curve and the thermal resistance parameters in the thermal node network, and the main path or thermal node causing the temperature gradient anomaly is identified. If the gradient anomaly is concentrated in the axial core conduction path inside the main shaft, the axial core conduction thermal node is prioritized. If the gradient change is concentrated in the cutting contact zone near the front end of the main shaft, the influence of the tool thermal node is considered. On this basis, the contact pressure distribution of the current bearing roller is further introduced to correct and compensate the physical connection strength in the thermal path, thereby more accurately locating the weak area in the thermal conduction topology link that is most significantly affected by the change in thermal resistance. After determining the weak area, the target link density increment that should be performed in this area is calculated based on the thermal modeling simulation results and the geometric error response curve. The link density is a representation parameter of the weight or equivalent heat conduction capacity in the thermal conduction link in this embodiment. This target increment is used to simulate more intensive heat flow transmission in the weak area to compensate for the underestimation of the original model for the high thermal resistance in this area. Then, the target increment is used as the basis for dynamically adjusting the distribution pattern of the thermal conduction topology link in this area, such as by increasing the connection weight between nodes, adding redundant paths, optimizing node distribution density, and other methods to complete the structural strengthening of the link. After the strengthening measures are implemented, the verification stage is entered, and new workpiece geometric errors are continuously collected in the next machining cycle and compared with those of the previous cycle to evaluate the effect of the thermal conduction link adjustment on the improvement of machining accuracy. If it is determined that the convergence does not meet the preset standard (for example, the error fluctuation is still greater than the set threshold, or the error directionality has not changed), the target link density increment is iteratively improved, and the link distribution adjustment is performed again until the workpiece geometric error and the temperature rise prediction value residual are both within the set tolerance interval.

[0092] Further, as shown in Figure 3 the temperature rise prediction value of each node is obtained, including:

[0093] Based on the thermal conductivity coefficient of the main shaft material and the proportion of the bearing roller contact pressure distribution, the thermal resistance parameters of each thermal conduction topology link are determined;

[0094] According to the real-time speed of the main shaft and the flow state of the cooling liquid, the convective heat transfer parameters of each core thermal node along the thermal conduction topology link are obtained;

[0095] injecting heat generation power distribution into corresponding core thermal nodes, calculating temperature rise prediction values of each core thermal node based on thermal resistance parameters and convection heat transfer parameters;

[0096] When the deviation between the temperature rise prediction value and the actual temperature value exceeds the set threshold, temperature compensation correction of the core thermal node is performed.

[0097] As a preferred embodiment of the above embodiment, first, the thermal conduction topology link is established based on the thermal physical properties of the spindle structure material and the real-time load state. In the modeling process, the thermal conductivity coefficient of the material used for the spindle is extracted from the material database, combined with the bearing arrangement structure of the spindle, and the spindle is divided into several thermal nodes by finite element modeling or equivalent thermal resistance network method. Each two nodes are connected by thermal conduction link to form a topology network. Subsequently, the contact pressure distribution between the bearing roller and the raceway is obtained by the pressure sensor or the load estimation module. According to the proportional relationship between the pressure and the contact area, the contact thermal resistance change between each thermal node is calculated to further improve the real-time distribution model of each thermal resistance parameter in the thermal conduction topology link. According to the real-time speed information of the spindle and the flow state of the cooling liquid, compare the built-in convection heat transfer model library to extract the heat transfer coefficient of each core thermal node under the current machining condition. This step fully considers the influence of the relative motion between the surface layer and the cooling liquid on the heat exchange efficiency when the spindle rotates, so as to determine the local convection heat transfer parameters of each node along the thermal conduction topology link. According to the real-time power output of the spindle motor and the cutting force distribution, a heat generation power injection model is constructed. The heat generated by electromagnetic loss, friction heat, cutting heat and other factors is injected into each corresponding core thermal node in proportion, for example, the heat generated at the driving end of the spindle is mainly injected into the bearing driving node area, and the cutting heat is injected into the contact area at the front end of the spindle. Then, the heat conduction is solved. Based on the determined thermal resistance parameters and convection heat transfer parameters, the temperature rise prediction values of each core thermal node are calculated by a steady-state or transient heat network iterative solution algorithm. After the above calculation process is completed, the current predicted temperature rise value is compared with the actual temperature value collected in real time. The actual temperature value is collected by the embedded thermocouple or infrared temperature sensor, and is processed by filtering and error correction to ensure accuracy. If it is found that the deviation between the temperature rise prediction value of any core thermal node and the corresponding actual temperature value exceeds the set threshold, the temperature compensation correction mechanism is triggered. The compensation mechanism mainly includes adjusting the dynamic coefficient of the thermal resistance or heat transfer parameter, re-fitting the heat generation power distribution, or using a fusion type neural network correction model to correct the model error, so that the next period prediction value tends to be the actual measured value, thereby improving the overall temperature prediction accuracy and robustness.

[0098] Further, the thermal resistance parameters of each thermal conduction topology link are determined, including:

[0099] The bearing inner and outer ring raceway contact area of the spindle is divided into several contact sub-areas, and each contact sub-area corresponds to the Hertz contact pressure distribution range of a single bearing roller.

[0100] According to the thermal conductivity coefficient of the main shaft material and the assembly pre-tightening force detection value, the contact thermal resistance reference value of each contact sub-area is calculated;

[0101] Based on the contact pressure distribution of the bearing roller along the circumference, the conduction thermal resistance parameters between adjacent contact sub-areas are dynamically allocated;

[0102] If the deviation of the temperature rise prediction value corresponding to the conduction thermal resistance parameter and the actual temperature value exceeds the set threshold value, the boundary division of the contact sub-area is reversely adjusted, and the conduction thermal resistance parameter is recalculated.

[0103] As a preferred embodiment of the above embodiment, firstly, the bearing structure closely related to heat conduction in the main shaft is subjected to micro thermal modeling processing, specifically, the contact area between the inner ring and the outer ring raceway of the main shaft bearing is divided into several contact sub-areas along the axial and radial directions, each contact sub-area corresponds to a single bearing roller Hertz contact area with the raceway during operation, the division is based on the real-time position information of the roller on the raceway, and is dynamically refined in combination with the instantaneous load contact trajectory of the roller under the running state of the main shaft, so as to better fit the actual heat conduction path; subsequently, based on the thermal conductivity coefficient of the main shaft and bearing material, the contact thermal resistance reference value of each contact sub-area is calculated considering the pre-tightening force detection value in the assembly process, the pre-tightening force detection value can be obtained through a sensor or a pressure assembly displacement and force feedback curve in the assembly process, and is converted into one of the parameters affecting the thermal conductivity of the contact surface through a force-thermal coupling model, this step effectively combines the influence of the actual assembly state of the main shaft assembly on the change of thermal resistance, improves the reality and dynamic adaptability of the modeling; then, a contact pressure distribution model of the bearing roller along the circumference is introduced, which can be constructed by an embedded force sensor, finite element analysis or load inversion algorithm, so as to realize real-time evaluation of the contact force of the roller at different positions on the raceway, according to the contact pressure distribution, the conduction thermal resistance parameters between adjacent contact sub-areas are dynamically allocated, for example, when the contact pressure of a roller in a certain area is significantly higher than that in the adjacent area, it indicates that the contact area is expanded and the heat flux density is higher, the corresponding thermal resistance should be reduced, then a lower thermal resistance value is allocated in the topology link to reflect the enhancement of the heat conduction capacity; on the basis of the heat conduction model, the temperature rise prediction calculation is continued, and the predicted temperature value is compared with the actual temperature measurement value, when it is detected that the temperature rise prediction value corresponding to one or a group of conduction thermal resistance parameters deviates from the actual temperature value by more than a set threshold value, a reverse adjustment mechanism is automatically triggered, which re-divides the boundary of the related contact sub-area based on the error gradient and the contact heat flux distribution, possible adjustment methods include expanding or reducing the coverage area of the high pressure area, merging the low efficiency heat channel area, etc., after the boundary re-division is completed, the conduction thermal resistance parameter of the area is recalculated, and the heat conduction topology link is updated to improve the accuracy and consistency of subsequent temperature rise prediction.

[0104] Further, the temperature compensation correction of the core thermal node is performed, including:

[0105] When the temperature rise prediction value deviation of the bearing thermal node or the tool thermal node is detected to exceed the threshold value, the geometric characteristics and spindle load state of the current machining path are extracted, and a compensation gain coefficient positively correlated with the curvature of the machining trajectory is generated;

[0106] According to the compensation gain coefficient, the conduction thermal resistance weight of the corresponding node is dynamically increased, and the convection heat transfer parameter of the adjacent node is simultaneously reduced;

[0107] Based on the conduction thermal resistance weight and the convection heat transfer parameter, the temperature rise prediction value of the core thermal node is recalculated, and the compensation effect is verified through in-situ temperature measurement;

[0108] If the residual error of the temperature rise prediction value after compensation is still higher than the threshold value, the proportional relationship between the compensation gain coefficient and the conduction thermal resistance weight is iteratively adjusted until the prediction accuracy of the thermal node network restores to the set range.

[0109] As a preferred embodiment of the above embodiment, first, the path geometric characteristics in the current machining task are extracted, including the curvature change of the tool motion trajectory, the feed speed, the depth of cut and other information. These parameters are obtained through numerical control system code analysis and real-time modeling of the tool trajectory. At the same time, the real-time load state data of the spindle corresponding to the path section is collected, including torque, power fluctuation and driving current amplitude. After fusion analysis of these multi-source information, a compensation gain coefficient is generated, which is positively correlated with the trajectory curvature. That is, in the machining section where the corner changes sharply or the load changes suddenly, the corresponding compensation gain coefficient will increase significantly, reflecting the response ability to the local thermal sensitive area temperature rise mutation. Then, according to the compensation gain coefficient, the thermal conduction topology attribute of the corresponding thermal node is dynamically adjusted, specifically, the conduction thermal resistance weight in the link connected with the node is increased, which means that the heat accumulation characteristics of the path are enhanced in the prediction model. At the same time, in order to maintain the heat balance logic, the convection heat transfer parameter of other nodes directly connected with the node is simultaneously reduced, simulating the situation that the local heat flow is more concentrated and the external heat dissipation ability is reduced. This kind of parameter coordination and adjustment mechanism can more truly reflect the transient heat accumulation phenomenon caused by local geometric load. Then, based on the updated conduction thermal resistance weight and convection heat transfer parameter, the temperature rise prediction calculation is performed again, and the temperature value of the core thermal node after compensation is obtained. In order to verify the compensation effect, the measured temperature value collected by the in-situ thermocouple or infrared temperature measurement is compared. When the residual error of the temperature rise prediction value after compensation decreases to within the set error tolerance, it is considered that the compensation is effective. If there is still a significant deviation, the residual error is used as a feedback signal to further iteratively adjust the proportional relationship between the compensation gain coefficient and the conduction thermal resistance weight, for example, to increase the nonlinear coupling term of the gain coefficient and the thermal resistance adjustment, until the prediction accuracy restores to the set range.

[0110] Further, the heat power distribution of each component of the main shaft is analyzed, including:

[0111] The stator copper loss power is calculated based on the stator motor resistance parameter, and the rotor eddy current loss power caused by harmonic current is calculated;

[0112] According to the product relationship of the real-time rotating speed of the main shaft and the load torque, and in combination with the preset bearing friction coefficient, the bearing friction loss power is calculated;

[0113] Based on the switching frequency data of the inverter and the DC bus voltage, the total switching loss is calculated;

[0114] Based on the physical structure heat capacity characteristics of the main shaft assembly, the stator copper loss power, the rotor eddy current loss power, the bearing friction loss power and the total switching loss are distributed to the bearing thermal node, the shaft core conduction thermal node and the tool thermal node.

[0115] As a preferred embodiment of the above embodiment, firstly, for the stator copper loss power calculation of the main shaft motor part, the phase current and stator resistance parameters of the main shaft motor are collected, and the joule heat loss generated by the three-phase stator current in the load state is calculated in real time, that is, the stator copper loss power, in order to further improve the accuracy of the heat source modeling, the high-order harmonic components of the motor current signal are analyzed through Fourier transform, and the rotor eddy current loss power generated by the high-frequency harmonic excitation is derived combined with the rotor structure and material parameters, this part of the loss mainly affects the rotor and its surrounding thermal nodes, especially the high-frequency thermal excitation area in the shaft core; secondly, in terms of bearing heat source, the mechanical output power of the main shaft is calculated according to the product of the current real-time speed and load torque of the main shaft, and the bearing friction coefficient preset under the specific structure of the main shaft is introduced, and the friction loss power generated by the contact between the bearing roller and the inner and outer rings is calculated by proportional distribution method, this part of power is mainly concentrated in the bearing thermal nodes, especially the contact area of the inner ring raceway and the roller, which is one of the main reasons for the formation of local hot spots in the bearing; in order to comprehensively cover the heat sources of the electronic components in the main shaft system, the loss heat input of the inverter part is analyzed, the switching frequency data output by the inverter controller and the voltage value of the direct current bus are collected, combined with the switching loss model of the inverter device, the total switching loss power is derived, and the heat mainly concentrates on the electric control interface thermal node close to the tail of the motor, which can have a significant thermal effect on the heat dissipation structure at the rear end of the main shaft; after the quantification of the above-mentioned heat source power, combined with the physical structure characteristics of the main shaft assembly, especially the thermal capacity, heat conduction path and structure coupling relationship of each structural unit, the heat source power is spatially distributed, and the specific distribution method is that the stator copper loss is mainly distributed to the shaft core conduction thermal node; the rotor eddy current loss power is distributed to the shaft core and the tool thermal node according to the moment of inertia and the axial position distribution weight; the bearing friction loss is directly assigned to the contact thermal nodes of the upper and lower bearings; and the switching loss is distributed to the tail bearing thermal node and the proximal thermal node of the motor shell according to the distance and thermal capacity coefficient weight, and the distribution process is supported by the preset structure thermal coupling coefficient matrix, so that the heat source energy can be reasonably transmitted and accumulated in the thermal node network.

[0116] Embodiment two;

[0117] Based on the same inventive concept as the temperature measurement method of the machine tool spindle in the foregoing embodiment, the present application also provides a machine tool spindle temperature measurement system, which comprises:

[0118] An information acquisition module acquires current signals and switching frequency data of the main shaft driving motor in real time, and analyzes the heat power distribution of each component of the main shaft;

[0119] A temperature rise prediction module decomposes the main shaft into a thermal node network based on the physical structure layout of each component of the main shaft, and obtains the temperature rise prediction value of each node combined with the heat power distribution;

[0120] The node compensation module obtains the actual temperature value of the key node, and performs temperature compensation correction on the corresponding node when the deviation between the temperature rise prediction value and the actual temperature value of any node exceeds the set threshold.

[0121] The thermal resistance correction module predicts the periodic temperature rise gradient in the future machining path according to the corrected thermal node network, and reversely corrects the thermal resistance parameter in the thermal node network based on the workpiece geometric error obtained by in-situ measurement.

[0122] The above adjustment system in the application can effectively realize the machine tool spindle temperature measurement method, and the technical effects are as described in the above embodiment, which will not be repeated here.

[0123] Further, the temperature rise prediction module comprises:

[0124] The thermal resistance determination unit determines the thermal resistance parameter of each thermal conduction topology link based on the thermal conductivity coefficient of the spindle material and the proportion of the bearing roller contact pressure distribution;

[0125] The state acquisition unit acquires the convective heat transfer parameter of each core thermal node along the thermal conduction topology link according to the real-time rotating speed of the spindle and the flow state of the cooling liquid;

[0126] The temperature rise calculation unit injects the heat power distribution into the corresponding core thermal node, and calculates the temperature rise prediction value of each core thermal node based on the thermal resistance parameter and the convective heat transfer parameter;

[0127] The compensation detection unit performs temperature compensation correction on the core thermal node when it is detected that the deviation between the temperature rise prediction value and the actual temperature value exceeds the set threshold.

[0128] Similarly, the above optimization scheme of the system can also correspondingly realize the optimization effect of the method in embodiment one, which will not be repeated here.

[0129] Although the present application has been described in connection with specific features and embodiments thereof, it will be evident that various modifications and combinations can be made thereto without departing from the spirit and scope of the application. Accordingly, the description and drawings are to be regarded as illustrative in nature and are not to be regarded as limiting the scope of the application. Obviously, various modifications and changes can be made to the present application by those skilled in the art without departing from the scope of the present application. Thus, it is intended that the present application cover modifications and variations of this application provided they come within the scope of the appended claims and their equivalents.

Claims

1. A machine tool spindle temperature measurement method, characterized by, The method comprises: Real-time acquisition of spindle drive motor current signal and switching frequency data, analysis of spindle component heat power distribution; Based on the physical structure layout of the spindle components, the spindle is decomposed into a thermal node network, and the heat power distribution is combined to obtain the temperature rise prediction value of each node; Obtain the actual temperature value of the key node, when the deviation of any node between the temperature rise prediction value and the actual temperature value exceeds the set threshold, the corresponding node is temperature compensated and corrected; According to the corrected thermal node network, the periodic temperature rise gradient in the future machining path is predicted, and based on the workpiece geometric error obtained by in-situ measurement, the thermal resistance parameters in the thermal node network are corrected in reverse; Based on the physical structure layout of the spindle components, the spindle is decomposed into a thermal node network, comprising: According to the actual physical contact area of the spindle bearing inner and outer ring raceway contact surface, shaft core and tool clamping surface, the spindle is divided into bearing thermal node, shaft core conduction thermal node and tool thermal node three core thermal nodes; Based on the core thermal node, a thermal conduction topology link is generated between adjacent nodes, wherein the link density between the bearing thermal node and the shaft core conduction thermal node is determined by the proportion of the bearing roller contact pressure distribution; Convert the spindle real-time speed and cooling liquid flow state into the thermal resistance parameters of the bearing thermal node and the tool thermal node, and update them to the thermal conduction topology link; According to the deviation between the actual temperature value and the current temperature rise prediction value, the link density and the thermal resistance parameters are adjusted in reverse until the prediction accuracy of the thermal node network meets the preset convergence condition; Analyzing the heat power distribution of each component of the spindle, comprising: Based on the spindle motor stator resistance parameter, the stator copper loss power is calculated, and the rotor eddy current loss power caused by harmonic current is calculated; According to the product relationship of spindle real-time speed and load torque, combined with the preset bearing friction coefficient, the bearing friction loss power is calculated; Based on the switching frequency data of the inverter and the DC bus voltage, the total switching loss is calculated; Based on the physical structure thermal capacity characteristics of the spindle components, the stator copper loss power, the rotor eddy current loss power, the bearing friction loss power and the total switching loss are distributed to the bearing thermal node, the shaft core conduction thermal node and the tool thermal node.

2. The machine tool spindle temperature measurement method of claim 1, wherein, Based on the thermal node network, the spindle temperature measurement comprises: Extract the actual temperature value of the core thermal node from the thermal conduction topology link, and time-domain align with the current harmonic characteristics of the spindle drive motor; Input the actual temperature value into the pre-trained time series prediction model to generate the temperature rise trend curve of each core thermal node in the future machining period, and dynamically adjust the amplitude and phase of the temperature rise trend curve according to the current task of the spindle; Based on the workpiece geometric error and the temperature rise trend curve, the maximum deviation area of the thermal resistance parameters in the thermal node network is located, and the directional reinforcement of the link density is triggered; Synchronize the link reinforcement result to the thermal conduction topology link, and repeat the above steps at a preset time interval until the residual error of the temperature rise prediction value is less than the convergence threshold for three times in a row.

3. The machine tool spindle temperature measurement method of claim 2, wherein, Triggering directional reinforcement of the link density, comprising: When any axial deviation of the workpiece geometric error is detected to continuously exceed the tolerance range, a temperature gradient variation feature of a corresponding axial direction in the temperature rise trend curve is extracted and mapped and matched with the thermal resistance parameter of the thermal node network; According to the matching result, a weak area of a heat conduction topological link in the shaft core conduction thermal node or the tool thermal node is located, and a target link density increment of the weak area is calculated in combination with a current bearing roller contact pressure distribution; Based on the target link density increment, a distribution mode of the heat conduction topological link of the weak area is dynamically adjusted, and convergence of the workpiece geometric error in a next machining cycle is verified after adjustment; If the convergence does not reach an expectation, the target link density increment is iteratively improved until a residual error of the temperature rise prediction value of the thermal node network falls in a preset interval synchronously with the workpiece geometric error.

4. The machine tool spindle temperature measurement method of claim 1, wherein, Obtaining a temperature rise prediction value of each node includes: Based on a thermal conductivity coefficient of a spindle material and a proportion of a bearing roller contact pressure distribution, the thermal resistance parameter of each heat conduction topological link is determined; According to a real-time rotating speed of the spindle and a cooling liquid flow state, a convection heat transfer parameter of each core thermal node along the heat conduction topological link is obtained; The heat generation power distribution is injected into the corresponding core thermal node, and a temperature rise prediction value of each core thermal node is calculated based on the thermal resistance parameter and the convection heat transfer parameter; When a deviation of the temperature rise prediction value from the actual temperature value exceeds a set threshold value, temperature compensation correction of the core thermal node is performed.

5. The machine tool spindle temperature measurement method of claim 4, wherein, Determining the thermal resistance parameter of each heat conduction topological link includes: The bearing inner and outer ring raceway contact areas of the spindle are decomposed into a plurality of contact sub-areas, and each contact sub-area corresponds to a Hertz contact pressure distribution range of a single bearing roller; According to the thermal conductivity coefficient of the spindle material and an assembly pre-tightening force detection value, a contact thermal resistance reference value of each contact sub-area is calculated; Based on a contact pressure distribution of the bearing roller along the circumference, a conduction thermal resistance parameter between adjacent contact sub-areas is dynamically allocated; If a deviation of the temperature rise prediction value corresponding to the conduction thermal resistance parameter from the actual temperature value exceeds a set threshold value, the boundary division of the contact sub-area is reversely adjusted, and the conduction thermal resistance parameter is recalculated.

6. The machine tool spindle temperature measurement method of claim 4, wherein, Performing temperature compensation correction of the core thermal node includes: When the temperature rise prediction value of the bearing thermal node or the tool thermal node is detected to deviate beyond a threshold value, geometric features and a spindle load state of a current machining path are extracted, and a compensation gain coefficient positively correlated with a machining trajectory curvature is generated; According to the compensation gain coefficient, a conduction thermal resistance weight of the corresponding node is dynamically improved, and the convection heat transfer parameter of the adjacent node is synchronously reduced; Based on the conduction thermal resistance weight and the convection heat transfer parameter, the temperature rise prediction value of the core thermal node is recalculated, and a compensation effect is verified through in-situ temperature measurement; If a residual error of the temperature rise prediction value after compensation is still higher than the threshold value, a proportional relationship between the compensation gain coefficient and the conduction thermal resistance weight is iteratively adjusted until a prediction accuracy of the thermal node network is restored to a set range.

7. Machine tool spindle temperature measurement system, characterized in that The machine tool spindle temperature measurement method as claimed in claim 1, wherein the system comprises: an information acquisition module, which acquires current signals and switching frequency data of a spindle driving motor in real time, and analyzes heat generation power distribution of each component of the spindle; a temperature rise prediction module, which decomposes the spindle into a thermal node network based on physical structure layout of each component of the spindle, and obtains temperature rise prediction values of each node in combination with the heat generation power distribution; a node compensation module, which obtains actual temperature values of key nodes, and performs temperature compensation correction on corresponding nodes when a deviation between the temperature rise prediction values and the actual temperature values of any node exceeds a set threshold value; a thermal resistance correction module, which predicts a periodic temperature rise gradient in a future machining path according to the corrected thermal node network, and reversely corrects thermal resistance parameters in the thermal node network based on workpiece geometric errors obtained through in-situ measurement.

8. The machine tool spindle temperature measurement system of claim 7, wherein, The temperature rise prediction module comprises: a thermal resistance determination unit, which determines thermal resistance parameters of each thermal conduction topology link based on a thermal conductivity coefficient of spindle material and a proportion of bearing roller contact pressure distribution; a state acquisition unit, which acquires convective heat transfer parameters of each core thermal node along the thermal conduction topology link according to a real-time rotating speed of the spindle and a cooling liquid flow state; a temperature rise calculation unit, which injects the heat generation power distribution into corresponding core thermal nodes, and calculates temperature rise prediction values of each core thermal node based on the thermal resistance parameters and the convective heat transfer parameters; a compensation detection unit, which performs temperature compensation correction on the core thermal nodes when detecting that a deviation between the temperature rise prediction values and the actual temperature values exceeds a set threshold value.

Images