Machine tool temperature rise and spindle thermal error look-ahead prediction method under complex unknown working conditions

Abstract

The application discloses a machine tool temperature rise and main shaft thermal error foresight prediction method under complex unknown working conditions, and comprises the following steps: S1, an improved machine tool temperature rise model and an improved main shaft thermal error model are established; S2, a machine tool temperature rise prediction model and a main shaft thermal error prediction model are established; S3, initial parameters of the machine tool temperature rise prediction model and the main shaft thermal error prediction model are determined; S4, the machine tool temperature rise and the main shaft thermal error under complex working conditions are predicted, and model parameters are updated; and S5, a change trend of the machine tool temperature rise and the main shaft thermal error under the whole complex working condition is obtained. The application can realize real-time monitoring of data and instant updating of the model, guarantee real-time performance and accuracy of model prediction, and obtain the change trend of the temperature rise and the thermal error under the whole complex working condition.

Description

Technical Field

[0001] This invention relates to the field of CNC machine tool error measurement and identification technology, and in particular to a method for predicting machine tool temperature rise and spindle thermal error under complex and unknown working conditions. Background Technology

[0002] Thermal error accounts for a significant proportion of the total error sources in machine tools. Thermal error compensation is an economical and effective means to improve the machining accuracy of machine tools. Its core lies in establishing a thermal error model with strong robustness and high predictive performance to predict and compensate for thermal errors.

[0003] Current CNC machine tool communication systems struggle to guarantee real-time performance. Furthermore, due to the nonlinear and variable nature of machine tool temperature rise and spindle thermal error changes under complex operating conditions, existing methods are insufficient for predicting these factors. Therefore, the precision and accuracy of online real-time compensation technology for CNC machine tools require further improvement. Summary of the Invention

[0004] To address the challenge of ensuring the real-time performance and accuracy of online compensation technology in complex machine tool operating conditions, this invention proposes a prospective prediction method for machine tool temperature rise and spindle thermal error under complex and unknown operating conditions. The machine tool continuously collects and monitors data in real time, while simultaneously updating model parameters dynamically. The updated model is then used to predict subsequent data, thereby achieving prospective prediction of machine tool temperature rise and spindle thermal error under all complex operating conditions. The prediction results are then used to achieve real-time compensation of the machined parts, improving the real-time performance and accuracy of online compensation and resolving the aforementioned problems.

[0005] This invention discloses a method for predicting machine tool temperature rise and spindle thermal error under complex and unknown operating conditions, comprising the following steps:

[0006] S1. Research on machine tool temperature rise and spindle thermal error under complex working conditions. Based on the machine tool temperature rise and spindle thermal error model, analyze the correspondence between the actual measured data of machine tool temperature rise and spindle thermal error and spindle speed. Add spindle speed to the prediction model. In order to reduce the lag effect of changes in machine tool temperature rise and spindle thermal error relative to changes in spindle speed, add time coefficient to the prediction model to establish an improved machine tool temperature rise model and an improved spindle thermal error model.

[0007] S2. Combine the improved machine tool temperature rise model and the improved spindle thermal error model with the unscented Kalman filter algorithm to establish the machine tool temperature rise prediction model and the spindle thermal error prediction model.

[0008] S3. Set the initial training time of the model, use the measurement data during the initial training time as input, and train the machine tool temperature rise prediction model and the spindle thermal error prediction model in combination with the unscented Kalman filter algorithm to determine the initial parameters of the machine tool temperature rise prediction model and the spindle thermal error prediction model.

[0009] S4. After the initial training time of the model, set the model update time period. Within the period, use the trained machine tool temperature rise prediction model and spindle thermal error prediction model to predict the machine tool temperature rise and spindle thermal error under complex working conditions, and update the model parameters respectively.

[0010] S5. Using the updated model parameters, continue to predict the machine tool temperature rise and spindle thermal error in the next time period. Continuous real-time data measurement and analysis are used, and the model parameters are dynamically adjusted at the same time. Based on the updated model parameters, predict the temperature rise and spindle thermal error in the new time period. Repeat this process to obtain the trend of machine tool temperature rise and spindle thermal error under the entire complex working condition.

[0011] Preferably, the improved machine tool temperature rise model in S1 is as follows:

[0012]

[0013] The improved spindle thermal error model is as follows:

[0014]

[0015] Where n is the spindle speed, and D T and D Z Let A be an undetermined coefficient. T A Z B T B Z C T C Z denoted as the material correlation coefficient, t as time, Δt as the time coefficient, and e as the natural constant.

[0016] Preferably, the machine tool temperature rise prediction model in S2 is:

[0017]

[0018] The spindle thermal error prediction model is as follows:

[0019]

[0020] Among them, w k For process noise, k is time, and T is the time interval. k Let Z be the temperature at time k. k Let f(x) be the thermal error at time k. k-1 ) represents the state quantity at time k-1 after being transformed by the function f, which expresses the state quantity at time k.

[0021] Preferably, step S3 includes the following steps:

[0022] Set the initial training time T.W The measurement data during the initial training time is used as input, and the unscented Kalman filter algorithm is used to train the machine tool temperature rise prediction model and the spindle thermal error prediction model, thereby determining the initial parameters of the machine tool temperature rise prediction model and the spindle thermal error prediction model.

[0023] Preferably, step S4 includes the following steps:

[0024] S41, during the initial training time T W Then, set the model update time period ΔT, that is, the model update time point is T. W +iΔT, i=1, 2,...,n;

[0025] S42, When the model update time point is T W When +iΔT, use 0~T W The cumulative actual measurement data and actual spindle speed within the +iΔT time period affect the machine tool temperature rise prediction model parameters B. T D T C T , n and spindle thermal error prediction model parameter B Z D Z C Z Update T and n. W +iΔT~T W The machine tool temperature rise and spindle thermal error within the +(i+1)ΔT time period are predicted, while T is measured. W +iΔT~T W The actual machine tool temperature rise and spindle thermal error data within the +(i+1)ΔT time period.

[0026] Preferably, step S5 includes the following steps:

[0027] Repeat step S42, using continuous real-time data measurement and analysis, while dynamically adjusting the model parameters. Based on the updated model parameters, predict the machine tool temperature rise and spindle thermal error within a new time period. Repeat this cycle continuously to obtain the trend of machine tool temperature rise and thermal error under the entire complex working condition.

[0028] The beneficial effects of this invention are:

[0029] By using machine measurement, it is possible to monitor the real-time updates of data and models. Under complex and unknown working conditions, the model can be adjusted in a timely manner according to the changes in temperature rise and thermal error, ensuring the real-time performance and accuracy of model predictions, and obtaining the changing trends of temperature rise and thermal error under the entire complex working conditions. Attached Figure Description

[0030] Figure 1 This is a flowchart of the machine tool temperature rise and spindle thermal error look-ahead prediction method under complex and unknown working conditions according to an embodiment of the present invention;

[0031] Figure 2 This is a schematic diagram of the machine tool temperature rise prediction results according to an embodiment of the present invention;

[0032] Figure 3 This is a schematic diagram of the residual of the machine tool temperature rise prediction results in an embodiment of the present invention;

[0033] Figure 4 This is a schematic diagram of the z-axis thermal error prediction results according to an embodiment of the present invention;

[0034] Figure 5 This is a schematic diagram of the residual of the z-axis thermal error prediction result in an embodiment of the present invention. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided with reference to the accompanying drawings and embodiments.

[0036] This application discloses a method for predicting machine tool temperature rise and spindle thermal error under complex and unknown working conditions. The method flow is as follows: Figure 1 As shown, it includes the following steps:

[0037] S1. Research on machine tool temperature rise and spindle thermal error under complex working conditions. Based on the machine tool temperature rise and spindle thermal error model, analyze the correspondence between the actual measured data of machine tool temperature rise and spindle thermal error and spindle speed. Add spindle speed to the prediction model. In order to reduce the lag effect of changes in machine tool temperature rise and spindle thermal error relative to changes in spindle speed, add time coefficient to the prediction model to establish an improved machine tool temperature rise model and an improved spindle thermal error model.

[0038] The improved machine tool temperature rise model is as follows:

[0039]

[0040] The improved spindle thermal error model is as follows:

[0041]

[0042] Where n is the spindle speed, and D T and D Z Let A be an undetermined coefficient. T A Z B T B Z C T C Z denoted as the material correlation coefficient, t as time, Δt as the time coefficient, and e as the natural constant.

[0043] S2. Combine the improved machine tool temperature rise model and the improved spindle thermal error model with the unscented Kalman filter algorithm to establish the machine tool temperature rise prediction model and the spindle thermal error prediction model.

[0044] The machine tool temperature rise prediction model is as follows:

[0045]

[0046] The spindle thermal error prediction model is as follows:

[0047]

[0048] Among them, w k For process noise, k is time, and T is the time interval. k Let Z be the temperature at time k. k Let f(x) be the thermal error at time k. k-1 ) represents the state quantity at time k-1 after being transformed by the function f, which expresses the state quantity at time k.

[0049] S3. Train the machine tool temperature rise prediction model and the spindle thermal error prediction model, and determine the initial parameters of the machine tool temperature rise prediction model and the spindle thermal error prediction model;

[0050] Set the initial training time T. W =15min, using the measurement data within the initial training time as input, and combining the unscented Kalman filter algorithm to complete the training of the machine tool temperature rise prediction model and the spindle thermal error prediction model, and determine the initial parameters of the machine tool temperature rise prediction model and the spindle thermal error prediction model.

[0051] S4. After the initial training time of the model, set the model update time period. Within the period, use the trained machine tool temperature rise prediction model and spindle thermal error prediction model to predict the machine tool temperature rise and spindle thermal error under complex working conditions, respectively. During this period, record the actual temperature rise and spindle thermal error data at the same time. When the update time point is reached, update the model coefficients using all the accumulated data.

[0052] S41, during the initial training time T W Then, set the model update time period ΔT, that is, the model update time point is T. W +iΔT, i=1, 2,...,n;

[0053] In this embodiment, the model update time period ΔT is set to 10 min, that is, the model update time points are 25 min, 35 min, 45 min, ..., 15+10n;

[0054] S42. When the model update time point is 15+10(i-1), the actual measurement data accumulated within the time period from 0 to 15+10(i-1) and the actual spindle speed are used to adjust the machine tool temperature rise prediction model parameters B. T D T C T , n and spindle thermal error prediction model parameter B Z D Z C Z The parameters n are updated, and the machine tool temperature rise and spindle thermal error within the time period of 15+10(i-1) to 15+10i are predicted. At the same time, the actual machine tool temperature rise and spindle thermal error data within the time period of 15+10(i-1) to 15+10i are measured.

[0055] When the model update time point is 15+10i, the actual measurement data accumulated during the time period from 0 to 15+10i and the actual spindle speed are used to adjust the machine tool temperature rise prediction model parameters B. T D T C T , n and spindle thermal error prediction model parameter B z D Z C Z The parameters n are updated, and the machine tool temperature rise and spindle thermal error within the time period of 15+10i~15+10*(i+1) are predicted. At the same time, the actual machine tool temperature rise and spindle thermal error data within the time period of 15+10i~15+10*(i+1) are measured.

[0056] S5. Using the updated model parameters, repeat S42 to continue predicting the machine tool temperature rise and spindle thermal error in the next time period. Use continuous real-time data measurement and analysis, and dynamically adjust the model parameters at the same time. Based on the updated model parameters, make forward predictions of temperature rise and spindle thermal error in the new time period, and continuously repeat to obtain the changing trends of machine tool temperature rise and spindle thermal error under the entire complex working condition.

[0057] In one embodiment, taking a CNC machine tool as an example, 29 temperature sensors T1-T29 are used to measure the main and secondary heat sources of the machine tool. Temperature sensors T1, T4, T12, and T22 are selected to collect data to illustrate the temperature rise prediction effect under complex working conditions. The results are as follows: Figure 2 , Figure 3 As shown in the figure, the gray shaded area represents the spindle speed. The maximum absolute errors at temperature points T1, T4, T12, and T22 are 0.374, 0.696, 0.384, and 0.864 degrees Celsius, respectively. The thermal error prediction is illustrated using the thermal error along the z-axis to demonstrate its effectiveness under complex operating conditions. The results are as follows... Figure 4 , Figure 5As shown in the figure, the gray shaded area represents the spindle speed, and the maximum absolute error in the z-axis direction is 1.727 micrometers. The results demonstrate that the method disclosed in this application can achieve prospective prediction of machine tool temperature and spindle thermal error under complex and unknown operating conditions.

[0058] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for predicting machine tool temperature rise and spindle thermal error under complex and unknown working conditions, characterized in that, Includes the following steps: S1. Add spindle speed and time coefficient to the machine tool temperature rise model and spindle thermal error model to establish an improved machine tool temperature rise model and an improved spindle thermal error model. The improved machine tool temperature rise model is as follows: The improved spindle thermal error model is as follows: in, Main spindle speed and For undetermined coefficients, , , , , , For material correlation coefficients, For time, For time coefficient, It is a natural constant; S2. Combine the improved machine tool temperature rise model and the improved spindle thermal error model with the unscented Kalman filter algorithm to establish the machine tool temperature rise prediction model and the spindle thermal error prediction model. The machine tool temperature rise prediction model is as follows: The spindle thermal error prediction model is as follows: in, For process noise, For a moment, for Temperature at any moment for Thermal error at any moment express The state quantity at time t is determined by the function Transformed expression State quantity at any given moment; S3. Train the machine tool temperature rise prediction model and the spindle thermal error prediction model, and determine the initial parameters of the machine tool temperature rise prediction model and the spindle thermal error prediction model; S4. Use the trained machine tool temperature rise prediction model and spindle thermal error prediction model to predict the machine tool temperature rise and spindle thermal error under complex working conditions, and update the model parameters. S41, during the initial training time Then, set the model update time period. That is, the model update time point is , ; S42, When the model update time point is At that time, utilize The cumulative actual measurement data and actual spindle speed within the time period affect the parameters of the machine tool temperature rise prediction model. , , , and spindle thermal error prediction model parameters , , , Update, Predict machine tool temperature rise and spindle thermal error within a time period, and simultaneously measure... Actual machine tool temperature rise and spindle thermal error data within the time period; S5. Using the updated model parameters, continue to predict the machine tool temperature rise and spindle thermal error, and repeatedly obtain the changing trends of machine tool temperature rise and spindle thermal error under the entire complex working condition.

2. The method for predicting machine tool temperature rise and spindle thermal error under complex and unknown working conditions according to claim 1, characterized in that, S3 includes the following steps: Set initial training time The measurement data during the initial training time is used as input, and the unscented Kalman filter algorithm is used to train the machine tool temperature rise prediction model and the spindle thermal error prediction model, thereby determining the initial parameters of the machine tool temperature rise prediction model and the spindle thermal error prediction model.

3. The method for predicting machine tool temperature rise and spindle thermal error under complex and unknown working conditions according to claim 2, characterized in that, S5 includes the following steps: Repeat step S42, using continuous real-time data measurement and analysis, while dynamically adjusting the model parameters. Based on the updated model parameters, predict the machine tool temperature rise and spindle thermal error within a new time period. Repeat this cycle continuously to obtain the trend of machine tool temperature rise and thermal error under the entire complex working condition.

Images