Machine tool environment temperature change prediction device and prediction method
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- OKUMA CORP
- Filing Date
- 2022-03-08
- Publication Date
- 2026-06-02
Smart Images
Figure CN115128995B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to an apparatus and method for predicting changes in ambient temperature in a factory equipped with machine beds. Background Technology
[0002] When machining on a machine tool, thermal deformation occurs in various parts of the machine tool due to mechanical heat generated by spindle and feed axis movements, as well as temperature variations in the machine tool's operating environment and coolant. This thermal displacement causes a change in the relative position of the tool and workpiece, thus deteriorating the machining accuracy of the workpiece. Existing technology for preventing accuracy degradation caused by thermal displacement of the machine tool involves estimating the displacement amount based on a pre-programmed thermal displacement estimation formula, using temperature sensors installed in various parts of the machine tool's structure or the operating conditions of the spindle, feed axes, etc., and then adjusting the axis movement accordingly. This thermal displacement correction is effective and widely used.
[0003] However, especially for thermal displacement caused by changes in ambient temperature, it is difficult to accurately estimate the amount of thermal displacement. Even for machinery that has undergone thermal displacement correction, the error can increase when the ambient temperature varies greatly. Therefore, in production, high-precision machining is sometimes required during periods of low ambient temperature variation, while dimensional verification / correction is performed during periods of high ambient temperature variation to ensure accuracy. These countermeasures, which take into account the magnitude of ambient temperature variation, largely rely on the operator's experience and experience.
[0004] On the other hand, regarding methods for predicting changes in room temperature within a building, prior art, such as Patent Document 1, is shown. Patent Document 1 describes a method that involves obtaining external air temperature and room temperature data within a building, generating a regression equation for calculating room temperature based on the external air temperature as a predictive equation, and using this predictive equation to predict the shift in room temperature by predicting changes in the external air temperature.
[0005] Patent Document 1: Japanese Patent No. 6160945
[0006] However, the method in Patent Document 1 is a technique for estimating room temperature without indoor cooling or heating. In reality, air conditioning is frequently used in factories, requiring consideration not only of changes in outside temperature but also of the effects of the air conditioning system. Furthermore, factory environments are considered in various scenarios depending on the method of air conditioning use and the building's insulation performance, such as continuous air conditioning use, use only during daytime working hours, high insulation against outside temperatures, and low insulation against outside temperatures. Therefore, predicting room temperature changes within a factory is effective if the prediction is based on the characteristics of the factory environment.
[0007] On the other hand, it is a well-known technology to predict changes in room temperature in a factory by inputting information such as the thermal insulation performance of the factory building, the settings of the air conditioners used in the factory, and the heat generated by the operation of equipment other than machine tools, and then using computer simulation. It is believed that this is feasible, but it is considered difficult to accurately input all the information required for the calculation in the actual production site. Summary of the Invention
[0008] Therefore, the object of the present invention is to provide a machine tool ambient temperature change prediction device and prediction method, which can appropriately predict the ambient temperature based on the machine tool body and / or the ambient temperature and the external temperature of the factory where the machine tool is placed, based on data that can be easily obtained.
[0009] To achieve the above objectives, the present invention provides a machine tool ambient temperature change prediction device, which predicts the ambient temperature changes in a factory where machine tools are installed, characterized by having:
[0010] An ambient temperature acquisition unit measures the body temperature and / or ambient temperature of a part of the machine tool that is not affected by the heat generated by the heat-generating part of the machine tool using a temperature sensor, and obtains the body temperature and / or ambient temperature of the part that is not affected by the heat generated by the heat-generating part of the machine tool as the ambient temperature.
[0011] The external temperature acquisition unit acquires the external temperature of the factory through temperature sensor measurements and / or meteorological data as the factory's external temperature.
[0012] The factory environment mode setting unit predefines the following classification rules and environmental temperature prediction model as factory environment modes. The classification rules classify the changing trend of the environmental temperature into several modes based on the data of the environmental temperature and the external temperature of the factory. The environmental temperature prediction model is different for each of the modes.
[0013] The prediction model generation unit, based on past data of ambient temperature and / or external factory temperature, selects a suitable factory environment pattern from the classification rules, and determines the parameters of the ambient temperature prediction model corresponding to the selected factory environment pattern; and
[0014] An ambient temperature change prediction unit predicts future changes in ambient temperature using an ambient temperature prediction model generated by the prediction model generation unit.
[0015] Another aspect of the present invention is characterized in that, in the above structure, it further includes an external temperature prediction data acquisition unit, which acquires future prediction data of the external temperature of the factory as external temperature prediction data.
[0016] The environmental temperature change prediction unit predicts future changes in the environmental temperature using the external air temperature prediction data and the environmental temperature prediction model.
[0017] Another aspect of the present invention is characterized in that, in the above structure, the classification rule in the factory environment mode setting unit is a classification rule based on the magnitude of the past changes in ambient temperature, the degree of correlation between the factory external temperature obtained by processing the past factory external temperature with a time delay and the changes in ambient temperature, and the periodicity of the past changes in ambient temperature.
[0018] Another aspect of the present invention is characterized in that, in the above structure, the environmental temperature prediction model defined by the factory environment mode setting unit is any one of constant temperature, a function related to time or day of the week, and a transfer function that takes the external temperature of the factory as input.
[0019] Another aspect of the present invention is characterized in that, in the above structure, when the change in ambient temperature is less than a predetermined threshold value, the factory environment mode setting unit sets the ambient temperature prediction model to the constant temperature.
[0020] If the magnitude of the change is above the threshold value and the correlation is higher than a predetermined coefficient threshold value, the factory environment mode setting unit sets the environmental temperature prediction model to a transfer function with the external air temperature of the factory as input.
[0021] When the magnitude of the change is above the magnitude threshold and the correlation is below the coefficient threshold, and the past changes in ambient temperature have a daily or weekly periodicity, the factory environment mode setting unit sets the ambient temperature prediction model as a function related to the time or day of the week.
[0022] If the magnitude of the change is above the magnitude threshold and the magnitude of the correlation is below the coefficient threshold, and the past changes in ambient temperature do not exhibit daily or weekly periodicity, the factory environment mode setting unit is set to be unable to determine the ambient temperature prediction model.
[0023] Another aspect of the present invention is characterized in that, in the above structure, a prediction result notification unit is further provided, which notifies the future change in the ambient temperature predicted by the ambient temperature change prediction unit.
[0024] To achieve the above objectives, the present invention provides a method for predicting environmental temperature changes in machine tools, which predicts environmental temperature changes in factories where machine tools are installed, characterized by performing the following steps:
[0025] The ambient temperature acquisition step involves measuring the body temperature and / or ambient temperature of a part of the machine tool that is not affected by the heat generated by the heat-generating part of the machine tool using a temperature sensor, and obtaining the body temperature and / or ambient temperature of the part that is not affected by the heat generated by the heat-generating part of the machine tool as the ambient temperature.
[0026] The external temperature acquisition step involves obtaining the external temperature of the factory through temperature sensor measurements and / or meteorological data.
[0027] The factory environment mode setting steps predefine the following classification rules and environmental temperature prediction model as factory environment modes. The classification rules classify the changing trend of the environmental temperature into several modes based on the data of the environmental temperature and the external temperature of the factory. The environmental temperature prediction model is different for each of the modes.
[0028] The prediction model generation step involves selecting a suitable factory environment pattern from the classification rules based on past ambient temperature and / or factory external temperature data, and determining the parameters of the ambient temperature prediction model corresponding to the selected factory environment pattern; and
[0029] The environmental temperature change prediction step uses the environmental temperature prediction model generated in the prediction model generation step to predict future changes in the environmental temperature.
[0030] Another aspect of the invention is characterized in that, in the above structure, an external temperature forecast data acquisition step is performed before the environmental temperature change prediction step, in which future forecast data of the factory's external temperature is acquired as external temperature forecast data.
[0031] In the environmental temperature change prediction step, the future environmental temperature change is predicted using the external temperature prediction data and the environmental temperature prediction model.
[0032] Another aspect of the present invention is characterized in that, in the above structure, the classification rule in the factory environment mode setting step is a classification rule based on the magnitude of past changes in ambient temperature, the degree of correlation between the factory external temperature obtained by time-delaying the past factory external temperature and the changes in ambient temperature, and the periodicity of past changes in ambient temperature.
[0033] Another aspect of the invention is characterized in that, in the above structure, the ambient temperature prediction model defined in the factory environment mode setting step is any one of constant temperature, time- or day-of-week related function, and transfer function that takes the external temperature of the factory as input.
[0034] Another aspect of the present invention is characterized in that, in the above structure, when the change in ambient temperature is less than a predetermined threshold, the ambient temperature prediction model is set to the constant temperature in the factory environment mode setting step.
[0035] If the magnitude of the change is above the threshold value and the correlation is higher than the predetermined coefficient threshold value, then in the factory environment model setting step, the environmental temperature prediction model is set as a transfer function with the external temperature of the factory as input.
[0036] If the magnitude of the change is above the magnitude threshold and the correlation is below the coefficient threshold, and the past environmental temperature changes have a daily or weekly periodicity, then in the factory environment model setting step, the environmental temperature prediction model is set as a function related to the time or day of the week.
[0037] If the magnitude of the change is above the magnitude threshold and the magnitude of the correlation is below the coefficient threshold, and the past changes in ambient temperature do not exhibit daily or weekly periodicity, then the ambient temperature prediction model cannot be determined in the factory environment mode setting step.
[0038] Another aspect of the present invention is characterized in that, in the above structure, a prediction result notification step is also performed, in which the future change in the ambient temperature predicted in the ambient temperature change prediction step is notified.
[0039] Invention Effects
[0040] According to the present invention, the ambient temperature around the machine tool and the external temperature of the factory are obtained, and based on this data, the tendency of ambient temperature changes is classified into several patterns. These patterns are associated with several envisioned factory environment patterns, such as factory environments close to temperature-controlled rooms, factory environments with low insulation and thus greatly affected by external temperatures, and factory environments greatly affected by air conditioning or surrounding heat sources. Furthermore, since different ambient temperature prediction models are prepared for each pattern of the factory environment, it is first necessary to diagnose which pattern is suitable for the factory environment where the machine tool is located, and then determine the parameters of the ambient temperature prediction model in accordance with the obtained data, thereby enabling the appropriate prediction of the ambient temperature in accordance with the factory environment where the machine tool is located. That is, the ambient temperature can be appropriately predicted in accordance with the factory environment where the machine tool is located, based on data of the machine tool's body and / or surrounding temperature and the external temperature of the factory, which are readily available.
[0041] According to another technical solution of the present invention, in addition to the above-mentioned effects, future changes in ambient temperature are predicted by using external temperature prediction data and an ambient temperature prediction model. Therefore, the parameters of the prediction model can be determined in accordance with the future prediction data of the external temperature of the factory, and the ambient temperature can be appropriately predicted in accordance with the factory environment with low insulation and thus a large influence of external temperature.
[0042] According to another aspect of the present invention, in addition to the effects described above, by setting the classification rules to be based on the magnitude of past changes in the ambient temperature of the factory, the degree of correlation between changes in the ambient temperature outside the factory and changes in the ambient temperature of the factory obtained by processing past changes in the ambient temperature outside the factory with a time delay, and the periodicity of past changes in the ambient temperature of the factory, it is possible to appropriately classify the patterns of the factory environment based on the obtained data of the ambient temperature around the machine tool and the ambient temperature outside the factory.
[0043] According to another technical solution of the present invention, in addition to the above-mentioned effects, by setting the environmental temperature prediction model to any of the following—constant temperature, time- or day-of-week related functions, and transfer functions that take the outside temperature of the factory as input—to match the factory environment pattern obtained by classification, it is possible to generate an appropriate environmental temperature prediction model that matches the factory environment.
[0044] According to another aspect of the present invention, in addition to the effects described above, by setting the algorithm to use classification rules to determine the ambient temperature prediction model, it is possible to automatically determine the ambient temperature prediction model that matches the factory environment based on the obtained data of the ambient temperature around the machine tool and the external temperature of the factory, and to easily predict changes in ambient temperature.
[0045] According to another aspect of the present invention, in addition to the effects described above, by notifying the predicted changes in ambient temperature, it is possible to prevent problems with poor machining accuracy caused by thermal displacement due to changes in ambient temperature. Attached Figure Description
[0046] Figure 1 This is a structural diagram of a machine tool ambient temperature change prediction device.
[0047] Figure 2 This is a flowchart illustrating the algorithm for predicting environmental temperature changes.
[0048] Figure 3 This is a graph illustrating temperature changes in a factory environment where external air temperature has a significant impact.
[0049] Figure 4 This is a graph illustrating temperature changes in a factory environment where the influence of air conditioning or surrounding heat sources is significant.
[0050] Label Explanation
[0051] 1: Factory; 2: Machine tool; 3: Air conditioner; 4, 5, 6: Temperature sensor; 7: Ambient temperature change prediction device; 8: Ambient temperature acquisition unit; 9: External temperature acquisition unit; 10: Factory environment mode setting unit; 11: Prediction model generation unit; 12: External temperature prediction data acquisition unit; 13: Ambient temperature change prediction unit; 14: Prediction result notification unit. Detailed Implementation
[0052] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.
[0053] Figure 1 This is a structural diagram of the ambient temperature change prediction device for machine tools that utilizes the present invention.
[0054] Inside the building of Factory 1, there are machine tools 2 and air conditioners 3 that control the room temperature of Factory 1. Outside Factory 1, there are temperature sensors 4 that measure the outside temperature of the factory. Inside Factory 1, on the machine tool 2, there are temperature sensors 5 that measure the temperature of the part that is not affected by heat-generating parts such as the spindle (in this example, the column) and temperature sensors 6 that measure the ambient temperature.
[0055] Furthermore, the external temperature of the factory can also be obtained by installing temperature sensor 4 on a portion of the outer wall of factory 1 that is considered to have the same temperature as the external temperature. Alternatively, the external temperature can be obtained from meteorological data of the area where factory 1 is located via the internet, in which case temperature sensor 4 can be omitted. Alternatively, both meteorological data and data from temperature sensor 4 can be used, for example, by averaging the data. Temperature sensor 5 and temperature sensor 6 can also be installed using only one of them.
[0056] If the room temperature in factory 1 changes, machine tool 2 may experience thermal displacement, leading to a deterioration in machining accuracy. To predict this in advance, an ambient temperature change prediction device 7 is installed. The ambient temperature change prediction device 7 can be built into the NC device of machine tool 2 or assembled into an electronic device different from machine tool 2.
[0057] The ambient temperature change prediction device 7 is equipped with an ambient temperature acquisition unit 8 that uses temperature sensors 5 and 6 to acquire data on the ambient temperature change of the machine tool 2, an external temperature acquisition unit 9 that uses temperature sensors 4 and the like to acquire data on the external temperature of the factory, and a factory environment mode setting unit 10.
[0058] In the factory environment mode setting unit 10, multiple modes classified based on the tendency of ambient temperature change are predefined as factory environment modes, and an ambient temperature prediction model (hereinafter referred to as "prediction model") is used to predict ambient temperature changes according to each mode (factory environment mode setting step). An algorithm is input into the factory environment mode setting unit 10, which diagnoses and classifies the factory environment mode based on the ambient temperature obtained by the ambient temperature acquisition unit 8 and the external temperature of the factory obtained by the external temperature acquisition unit 9.
[0059] In addition, the ambient temperature change prediction device 7 is equipped with a prediction model generation unit 11, an external temperature prediction data acquisition unit 12, an ambient temperature change prediction unit 13, and a prediction result notification unit 14.
[0060] The prediction model generation unit 11 selects the factory environment mode set by the factory environment mode setting unit 10 based on the past temperature data stored by the ambient temperature acquisition unit 8 and the external temperature acquisition unit 9, and the algorithm input to the factory environment mode setting unit 10, and determines the parameters of the prediction model corresponding to the selected factory environment mode.
[0061] External temperature forecast data acquisition department 12 obtains meteorological forecast data through the Internet or other means, or obtains future external temperature forecast data using other forecasting units.
[0062] The ambient temperature change prediction unit 13 uses the prediction model generated by the prediction model generation unit 11 and the external temperature prediction data obtained by the external temperature prediction data acquisition unit 12, or it uses the prediction model and past temperature data accumulated in the ambient temperature acquisition unit 8 and the external temperature acquisition unit 9 without using the external temperature prediction data, to predict the future ambient temperature change of the machine tool 2.
[0063] The prediction result notification unit 14 notifies the operator using machine tool 2 of the ambient temperature change predicted by the ambient temperature change prediction unit 13. If the predicted ambient temperature change is larger than the specified range, thus anticipating a deterioration in the machining accuracy of machine tool 2, the prediction result notification unit 14 will issue an alarm. The prediction result notification unit 14 can display the results on a screen or send email notifications to other terminals.
[0064] Next, use Figure 2 The flowchart illustrates the method for predicting environmental temperature changes performed by the environmental temperature change prediction device 7.
[0065] First, in S1, data on changes in the ambient temperature outside the factory and changes in the ambient temperature of the machine tool 2 are acquired by the ambient temperature acquisition unit 8 and the external temperature acquisition unit 9 (ambient temperature acquisition step and external temperature acquisition step). The acquisition period is based on a period of one week to one month.
[0066] Next, in S2, the factory environment mode setting unit 10 calculates the range of change in ambient temperature during the calculation period and determines whether it exceeds a predetermined range threshold. The range threshold is determined considering factors such as the required precision in the machining of the machine tool 2. The range threshold is based on approximately 1°C to 3°C. If the range of change is smaller than the range threshold, it is determined in S3 to be a factory environment close to a constant temperature room. In this case, in the prediction model generation unit 11, the prediction model for ambient temperature change is defined according to the average temperature (constant temperature) during the past measurement period, as shown in the following formula (1) (prediction model generation step).
[0067] [Mathematical Expression 1]
[0068]
[0069] θ m.i Measured values of ambient temperature changes
[0070] E(θ m.i ): The average value of the measured changes in ambient temperature.
[0071] Predicted values of ambient temperature changes
[0072] In the judgment of S2, if the change in ambient temperature during the period is above the change range threshold, the factory environment mode setting unit 10 performs a first-order delay processing on the external temperature of the factory while changing the time constant in S4, calculates the correlation coefficient related to the change in ambient temperature, and performs processing to find the value of the maximum correlation coefficient. First, for the measured external temperature of the factory θa.i, the external temperature of the factory after the first-order delay processing is calculated by the following formula (2).
[0073] [Mathematical Expression 2]
[0074]
[0075] Δt: The period of temperature measurement
[0076] T: The time constant of the first-order delay (assumed value)
[0077] θ a.i Measurement of external temperature at the factory
[0078] θ aT.i The external temperature of the factory obtained by applying a first-order delay to the time constant T.
[0079] Next, the correlation coefficient between the factory external air temperature θaT.i obtained by delaying the process using the time constant T in equation (2) and the measured value θm.i of the ambient temperature change is calculated. While changing the time constant T of the first-order delay in equation (2), the maximum correlation coefficient rmax is obtained by the following equation (3).
[0080] [Mathematical Expression 3]
[0081]
[0082] Standard deviation of measured values of ambient temperature change
[0083] The standard deviation of the factory's external temperature obtained by performing first-order delay processing
[0084] The measured values of ambient temperature changes and the covariance of the factory's external air temperature obtained after first-order delay processing.
[0085] Next, in S5, the factory environment mode setting unit 10 determines whether the maximum correlation coefficient rmax obtained by equation (3) is below a predetermined first coefficient threshold. This first coefficient threshold is preset, but is based on a level of 0.8.
[0086] When the correlation coefficient rmax exceeds the first coefficient threshold, it is determined in S6 that the factory environment is under the influence of the external temperature. At this time, in the prediction model generation unit 11, the prediction model of the ambient temperature change is defined as a transfer function with the external temperature of the factory as input. Specifically, the prediction model can be defined as follows: using the time constant Tm of the first-order delay that maximizes the correlation coefficient obtained by equations (2) and (3), the first-order delay of the predicted value of the external temperature of the factory is calculated using the following equation (4) with the same first-order delay as equation (2), and a linear transformation is performed using the following equation (5) (prediction model generation step).
[0087] [Mathematical Expression 4]
[0088]
[0089]
[0090] Δt: The period of temperature measurement
[0091] T m The time constant of the first-order delay (the time constant when the correlation coefficient rmax is at its maximum).
[0092] Predicted external temperature of the factory
[0093] The predicted external temperature of the factory obtained by applying a delay process using the time constant Tm.
[0094] Predicted values of ambient temperature changes
[0095] a, b: constants
[0096] In the judgment in S5, if the correlation coefficient rmax, which is the largest, is below the first coefficient threshold, the factory environment mode setting unit 10 calculates the periodicity index of daily temperature changes in S7. The calculation method, for example, involves dividing the ambient temperature change data by day and calculating the correlation coefficient between the daily data. At this time, the correlation coefficient between the data is calculated as 1 / 2 × (number of days) × (number of days - 1).
[0097] In S8, the factory environment mode setting unit 10 compares the smallest correlation coefficient among the aforementioned (1 / 2 × (number of days) × (number of days - 1) types) correlation coefficients with a pre-set second coefficient threshold to make a determination. Alternatively, the smallest correlation coefficient can be omitted, and the average of the calculated correlation coefficients can be compared. Furthermore, in the method of using correlation coefficients for determination, the magnitude of temperature change is ignored; therefore, an index equivalent to the distance between data points, such as the square root of the difference, can be calculated, and the magnitude of temperature change can be considered when making the determination. Alternatively, data can be divided into weekly periods instead of daily periods, and the correlation coefficients and distances between the data points can be calculated to determine periodicity. This method is effective for situations where air conditioning is used on weekdays (including nighttime) but the power is cut off on weekends.
[0098] In S8, if the correlation coefficient between the data is greater than the second coefficient threshold, there is a periodicity. Therefore, in S9, the factory environment is determined to be one where the influence of air conditioning or surrounding heat sources is significant, as the air conditioning is periodically turned ON / OFF daily or weekly. In this case, in the prediction model generation unit 11, a prediction model for the change in ambient temperature is defined using a function related to time or day of the week (prediction model generation step). Specifically, as shown in the following equation (6), a linear function that yields the average value of the measured values corresponding to each time and day of the week is used as the predicted value. Point sets of data for time and predicted values are generated, and the result obtained by interpolating between the points becomes a function representing the change in ambient temperature.
[0099] [Mathematical Expression 5]
[0100]
[0101] Predicted ambient temperature for a specific day and time of the week
[0102] The average value of the ambient temperature measured on a specific day and at a specific time during the week.
[0103] a, b: constants
[0104] On the other hand, in the determination in S8, if the correlation coefficient between the data is below the second coefficient threshold, the correlation between the temperature change outside the factory obtained by time-delaying the temperature change outside the factory and the ambient temperature change is low, and the ambient temperature change does not exhibit periodicity. In this case, in S10, it is determined that the factory environment has irregular temperature changes, and the prediction model generation unit 11 cannot determine the prediction model. In this situation, the prediction result notification unit 14 can notify the operator of the machine tool 2 that it is necessary to determine the cause of the irregular temperature changes and improve the factory environment.
[0105] Using the above method, in the factory environment mode setting unit 10, the factory environment mode is classified into S3: factory environment close to constant temperature, S6: factory environment with large influence of external temperature, S9: factory environment with large influence of air conditioning or surrounding heat sources, and S10: factory environment with irregular temperature changes, and the prediction model corresponding to each classification is determined.
[0106] Therefore, the prediction model generation unit 11 selects a suitable factory environment model based on past ambient temperature and / or external factory temperature data, and determines the parameters of the prediction model in the selected factory environment model.
[0107] Based on the prediction model whose parameters are determined in this way, in the ambient temperature change prediction unit 13, in S11, in the cases of S3 and S9, the ambient temperature change is predicted based on the past ambient temperature data accumulated in the ambient temperature acquisition unit 8 and the external temperature acquisition unit 9, and in the case of S6, the ambient temperature change is predicted based on the future prediction data of the factory's external temperature obtained from the external temperature prediction data acquisition unit 12 (ambient temperature change prediction step).
[0108] The prediction results in the ambient temperature change prediction unit 13 are notified by the prediction result notification unit 14 in S12 (prediction result notification step).
[0109] Finally, a specific example of using the present invention to diagnose the factory environment and predict ambient temperature based on measurement data is shown.
[0110] Figure 3 and Figure 4These are graphs showing the measured and predicted values of external temperature changes and ambient temperature changes (indoor temperature changes within the factory) over a week in different factories and seasons. Observing the graphs, we can see that the external temperature follows a daily cycle, with higher temperatures during the day and lower temperatures at night. However, the external temperature is also affected by weather conditions, and the pattern of temperature increase varies depending on the date. For example… Figure 3 The sixth day Figure 4 The fifth day was rainy, so the temperature rise was smaller compared to other days.
[0111] On the other hand, when observing changes in ambient temperature, it can be seen that... Figure 3 On days when the external temperature changes little, the ambient temperature also changes little, but... Figure 4 The temperature in the central region rises to around 20°C during the day, independent of changes in external temperature. (Consideration) Figure 3 and Figure 4 The nature of temperature changes in the environment of the factories is different.
[0112] according to Figure 3 and Figure 4 The changes are based on Figure 2 The flowchart is used to derive the prediction model. In this example, the threshold for the change in ambient temperature in S2 is set to 3℃, the first threshold for the correlation coefficient between the external air temperature and the ambient temperature in S5 is set to 0.8, and the second threshold for the minimum correlation coefficient between the daily data in S8 is set to 0.7.
[0113] At this time, Figure 3 In the factory, the ambient temperature change over one week in S2 was 5.9℃, which was determined to be greater than the change threshold of 3℃. Next, in S4, while changing the time constant, a first-order delay was applied to the external temperature. When calculating the correlation coefficient with the ambient temperature change, the maximum correlation coefficient was found to be rmax = 0.95, which was determined in S5 to be greater than the first coefficient threshold of 0.8.
[0114] As a result, in S6, the factory environment was diagnosed as being greatly affected by external temperature, and the change in ambient temperature was defined as a transfer function with the external temperature as input. If the transfer function is actually calculated, then... Figure 3 The predicted results, represented by the dashed line, accurately reflect the actual changes in ambient temperature.
[0115] On the other hand, Figure 4In the factory, the ambient temperature variation over one week in S2 was 7.3℃, which was determined to be greater than the variation threshold of 3℃. Next, in S4, while varying the time constant, a first-order delay was applied to the external air temperature. When calculating the correlation coefficient with the ambient temperature variation, the maximum correlation coefficient was found to be rmax = 0.68, which was determined to be less than the first coefficient threshold of 0.8 in S5. Then, in S7, the ambient temperature variation data was divided into daily data, and the correlation coefficients between daily data were calculated. 21 correlation coefficients were obtained over 7 days, but the correlation coefficients ranged from 0.94 to 0.99. In S8, the minimum correlation coefficient of 0.94 was determined to be greater than the second coefficient threshold of 0.7. As a result, in S9, the factory environment was diagnosed as being significantly affected by air conditioning or surrounding heat sources, and the ambient temperature variation was defined as a function of time. If the average temperature of each time period over 7 days is taken and solved as a function of time, the following results are obtained. Figure 4 The dotted line-like prediction results show the trend of the measured environmental temperature changes within one day.
[0116] As described above, the ambient temperature change prediction device 7 for the machine tool 2 includes: an ambient temperature acquisition unit 8, which measures the body temperature of parts unaffected by the heat generation of the machine tool 2's heat-generating parts and the ambient temperature using temperature sensors 5 and 6, and acquires this collective temperature and ambient temperature as the ambient temperature; an external temperature acquisition unit 9, which acquires the external temperature of the factory using temperature sensor 4 and / or meteorological data as the external temperature of the factory; a factory environment mode setting unit 10, which pre-defines a classification rule that classifies the tendency of ambient temperature change into several modes based on ambient temperature and external temperature data, and a prediction model different for each mode as a factory environment mode; a prediction model generation unit 11, which selects a suitable factory environment mode from the classification rule based on past ambient temperature and / or external temperature data, and determines the parameters of the corresponding prediction model; and an ambient temperature change prediction unit 13, which predicts future ambient temperature changes using the prediction model generated by the prediction model generation unit 11. Thus, the ambient temperature change prediction device 7 performs... Figure 2 A method for predicting environmental temperature changes.
[0117] Based on this structure, the ambient temperature around the machine tool 2 and the external temperature of the factory are obtained. Based on this data, the tendency of ambient temperature changes is classified into several patterns. These patterns are associated with several envisioned factory environment patterns, such as factory environments close to constant temperature chambers, factory environments with low insulation and thus greatly affected by external temperatures, and factory environments greatly affected by air conditioning or surrounding heat sources. Furthermore, since different prediction models are prepared for each factory environment pattern, it is first necessary to diagnose which pattern is suitable for the factory environment where the machine tool 2 is located. Then, the parameters of the prediction model are determined in accordance with the obtained data, thereby enabling appropriate prediction of the ambient temperature based on the factory environment where the machine tool 2 is located. In other words, the ambient temperature can be appropriately predicted based on data from the machine tool 2 itself, its surrounding temperature, and the external temperature of the factory, in accordance with the factory environment where the machine tool 2 is located.
[0118] In particular, it also has an external temperature prediction data acquisition unit 12 that obtains future forecast data of the external temperature of the factory as external temperature prediction data, and an environmental temperature change prediction unit 13 that predicts future environmental temperature changes based on the external temperature prediction data and the prediction model. Therefore, it can determine the parameters of the prediction model in accordance with the future forecast data of the external temperature of the factory, and can appropriately predict the environmental temperature in accordance with the factory environment with low insulation and thus large influence of external temperature.
[0119] Furthermore, the classification rules in the factory environment mode setting unit 10 are based on the magnitude of past changes in ambient temperature, the degree of correlation between past external factory temperature and ambient temperature changes after time delay processing of past external factory temperature, and the periodicity of past ambient temperature changes. Therefore, the factory environment mode can be appropriately classified based on the acquired data of ambient temperature around the machine tool 2 and external factory temperature.
[0120] Furthermore, since the prediction model can be set to any of the following based on the factory environment pattern obtained from the classification: constant temperature, time or day of the week, and transfer function with the outside temperature as input, it is possible to generate an appropriate prediction model that matches the factory environment.
[0121] Furthermore, the factory environment model setting unit 10 sets the prediction model to constant temperature when the change in ambient temperature is below a predetermined threshold; sets it to a transfer function that takes the external air temperature as input when the change is above the threshold and the correlation is higher than a predetermined first coefficient threshold; sets it to a function related to time or day of the week when the change is above the threshold, the correlation is below the first coefficient threshold, and past ambient temperature changes exhibit daily or weekly periodicity; and sets it to undetermined when the change is above the threshold, the correlation is below the first coefficient threshold, and past ambient temperature changes do not exhibit daily or weekly periodicity. Therefore, based on the acquired data of the ambient temperature around the machine tool 2 and the external air temperature, it can automatically determine a prediction model that matches the factory environment, enabling simple prediction of ambient temperature changes.
[0122] Furthermore, by also having a notification unit 14 that notifies the ambient temperature change prediction unit 13 of the predicted future ambient temperature changes, it is possible to prevent problems with poor machining accuracy caused by thermal displacement due to ambient temperature changes.
[0123] In addition, the temperature sensors for measuring the machine tool body temperature, ambient temperature, and factory external temperature are not limited to one each; multiple sensors can be set and the average of multiple measurements can be taken.
[0124] The factory environment modes defined by the Factory Environment Mode Setting Department are not limited to the three modes mentioned above, and can be further subdivided and added.
Claims
1. A machine tool ambient temperature change prediction device, which predicts the ambient temperature changes in a factory where machine tools are installed, characterized in that, have: The ambient temperature acquisition unit measures the body temperature of a part of the machine tool that is not affected by the heat generated by the heat-generating part of the machine tool, or the body temperature and the ambient temperature, and obtains the body temperature or the body temperature and the ambient temperature as the ambient temperature. The external temperature acquisition unit acquires the external temperature of the factory through temperature sensor measurements and / or meteorological data as the factory's external temperature. The factory environment mode setting unit predefines the following classification rules and environmental temperature prediction model as factory environment modes. The classification rules classify the changing trend of the environmental temperature into several modes based on the data of the environmental temperature and the external temperature of the factory. The environmental temperature prediction model is different for each of the modes. The prediction model generation unit, based on past data of ambient temperature and / or external factory temperature, selects a suitable factory environment pattern from the classification rules and determines the parameters of the corresponding ambient temperature prediction model; and An ambient temperature change prediction unit predicts future changes in ambient temperature using an ambient temperature prediction model generated by the prediction model generation unit. The classification rules in the factory environment mode setting unit are based on the magnitude of past changes in ambient temperature, the degree of correlation between the factory's external temperature and the ambient temperature changes (obtained by time-delaying past external factory temperatures), and the periodicity of past ambient temperature changes. If the change in ambient temperature is less than a predetermined threshold, the factory environment mode setting unit sets the ambient temperature change tendency to a near-constant temperature factory environment, which is considered the first factory environment. When the change range is above the change range threshold and the correlation degree is higher than the predetermined coefficient threshold, the factory environment mode setting unit sets the ambient temperature change tendency to a factory environment where the external temperature has a significant impact as a second factory environment. When the magnitude of the change is above the magnitude threshold and the degree of correlation is below the coefficient threshold, and the past changes in ambient temperature have a daily or weekly periodicity, the factory environment mode setting unit sets the ambient temperature change tendency to a factory environment where the influence of air conditioning or surrounding heat sources, which are considered third-party factory environments, is significant. When the magnitude of the change is above the magnitude threshold and the magnitude of the correlation is below the coefficient threshold, and the past changes in ambient temperature do not exhibit daily or weekly periodicity, the factory environment mode setting unit sets the tendency of the ambient temperature change as a factory environment with irregular temperature changes, which is a fourth factory environment.
2. The machine tool ambient temperature change prediction device according to claim 1, characterized in that, The machine tool's ambient temperature change prediction device also includes an external temperature prediction data acquisition unit, which acquires future prediction data of the factory's external temperature as external temperature prediction data. The environmental temperature change prediction unit predicts future changes in the environmental temperature using the external air temperature prediction data and the environmental temperature prediction model.
3. The machine tool ambient temperature change prediction device according to claim 1, characterized in that, The environmental temperature prediction model defined by the factory environment mode setting unit is any one of constant temperature, time-of-day or day-of-week function, and transfer function that takes the external temperature of the factory as input.
4. The machine tool ambient temperature change prediction device according to claim 3, characterized in that, In the case of the first factory environment, the factory environment mode setting unit sets the environmental temperature prediction model to the constant temperature. In the case of the second factory environment, the factory environment mode setting unit sets the environmental temperature prediction model to a transfer function that takes the external temperature of the factory as input. In the case of the third factory environment, the factory environment mode setting unit sets the environmental temperature prediction model as a function related to the time or day of the week. In the case of the fourth factory environment, the factory environment mode setting unit is set to be unable to determine the environmental temperature prediction model.
5. The machine tool ambient temperature change prediction device according to any one of claims 1 to 4, characterized in that, The machine tool's ambient temperature change prediction device also includes a prediction result notification unit, which notifies the future ambient temperature change predicted by the ambient temperature change prediction unit.
6. A method for predicting ambient temperature changes in machine tools, characterized in that, Perform the following steps: The ambient temperature acquisition step involves measuring the body temperature of a part of the machine tool that is not affected by the heat generated by the heat-generating part of the machine tool, or the body temperature and the ambient temperature, using a temperature sensor, and obtaining the body temperature or the body temperature and the ambient temperature as the ambient temperature. The external temperature acquisition step involves obtaining the external temperature of the factory through temperature sensor measurements and / or meteorological data. The factory environment mode setting steps predefine the following classification rules and environmental temperature prediction model as factory environment modes. The classification rules classify the changing trend of the environmental temperature into several modes based on the data of the environmental temperature and the external temperature of the factory. The environmental temperature prediction model is different for each of the modes. The prediction model generation step involves selecting a suitable factory environment pattern from the classification rules based on past data of the ambient temperature and / or the external temperature of the factory, and determining the parameters of the corresponding ambient temperature prediction model; and The environmental temperature change prediction step uses the environmental temperature prediction model generated in the prediction model generation step to predict future changes in the environmental temperature. The classification rules in the factory environment mode setting step are based on the magnitude of past changes in ambient temperature, the degree of correlation between the factory's external temperature and the ambient temperature changes obtained by time-delaying the past external temperature of the factory, and the periodicity of past ambient temperature changes. If the change in ambient temperature is less than a predetermined threshold, in the factory environment mode setting step, the tendency of the ambient temperature change is set to a near-constant temperature factory environment, which serves as the first factory environment. If the magnitude of the change is above the threshold value and the degree of correlation is higher than the predetermined coefficient threshold value, in the factory environment mode setting step, the tendency of the ambient temperature change is set to a factory environment where the external temperature has a significant impact as the second factory environment. If the magnitude of the change is above the magnitude threshold and the degree of correlation is below the coefficient threshold, and the past changes in ambient temperature have a daily or weekly periodicity, then in the factory environment mode setting step, the tendency of the ambient temperature change is set to a factory environment where the influence of the air conditioning or surrounding heat sources of the third factory environment is significant. If the magnitude of the change is above the magnitude threshold and the magnitude of the correlation is below the coefficient threshold, and the past changes in ambient temperature do not exhibit daily or weekly periodicity, then in the factory environment mode setting step, the tendency of the ambient temperature change is set as a factory environment with irregular temperature changes, which is considered a fourth factory environment.
7. The method for predicting ambient temperature changes in machine tools according to claim 6, characterized in that, Before the ambient temperature change prediction step, an external temperature prediction data acquisition step is performed. In this external temperature prediction data acquisition step, future prediction data of the factory's external temperature is obtained as the external temperature prediction data. In the environmental temperature change prediction step, the future environmental temperature change is predicted using the external temperature prediction data and the environmental temperature prediction model.
8. The method for predicting ambient temperature changes in machine tools according to claim 6, characterized in that, The environmental temperature prediction model defined in the factory environment mode setting step is any one of constant temperature, time- or day-of-week related functions, and transfer functions that take the external temperature of the factory as input.
9. The method for predicting ambient temperature changes in machine tools according to claim 8, characterized in that, In the case of the first factory environment, the environmental temperature prediction model is set to the constant temperature in the factory environment mode setting step. In the case of the second factory environment, during the factory environment mode setting step, the environmental temperature prediction model is set as a transfer function with the external temperature of the factory as input. In the case of the third factory environment, during the factory environment mode setting step, the environmental temperature prediction model is set as a function related to the time or day of the week. In the case of the fourth factory environment, the factory environment mode setting step is set to be unable to determine the environmental temperature prediction model.
10. The method for predicting ambient temperature changes in machine tools according to any one of claims 6 to 9, characterized in that, The system also performs a prediction result notification step, in which it notifies the system of the future changes in ambient temperature predicted in the ambient temperature change prediction step.