Processing device and method for determining a model for calculating winding temperatures

Abstract

A processing device (4) that determines a calculation model (10, 30) for estimating a temperature of a winding, wherein the calculation model (10, 30) is contained in an electronic thermal unit (100) of a motor (2) that includes a stator around which the winding is wound and a rotor, wherein the calculation model (10, 30) includes a winding temperature characteristic model (11, 21) that includes a winding-related parameter related to a temperature characteristic of the winding, and a predetermined temperature characteristic model (12, 31) that includes a predetermined parameter related to a characteristic of a temperature in the vicinity of the winding, which is detected by a temperature sensor (3) arranged in the vicinity of the winding, wherein the processing device (4) comprises: a temperature transition detection unit (220) configured to detect a first rising transition, which is a rising transition of a temperature of the winding, and a second rising transition, which is a rising transition of a temperature detected by the temperature sensor (3) in a state in which a voltage application for increasing the temperature of the winding to a predetermined temperature is performed; and a determination unit (230) configured to determine the predetermined temperature characteristic model (12, 31) by calculating the predetermined parameter based on the second rising transition, and further configured to determine the winding temperature characteristic model (11, 21) by calculating the winding-related parameter based on the first rising transition, wherein the determination unit (230) determines the predetermined temperature characteristic model (12, 31) by calculating the predetermined parameter based on the second rising transition via a stator temperature characteristic model (22, 32) that includes a stator-related parameter related to a temperature characteristic of the stator and has a correlation with the winding temperature characteristic model (11, 21) and the predetermined temperature characteristic model (12, 31),and determines the winding temperature characteristic model (11, 21) by calculating the winding-related parameter based on the first rising transition via the stator temperature characteristic model (22, 32).

Description

TECHNICAL FIELD

[0001] The present invention relates to a technique for adjusting a model parameter related to the electronic thermals of an engine. BACKGROUND OF THE INVENTION

[0002] Motors are used in various fields, and the speeds, motor loads, and the like that constitute their operating conditions also vary. In addition, the ambient atmosphere of a motor is not necessarily constant. When the ambient temperature of a motor rises, it is generally difficult for the motor to dissipate heat, and the operating conditions become more severe. When a motor operates in an overload environment, the motor winding temperature rises excessively, and the winding may burn out. To prevent such a winding burnout, there is a technique in which a temperature sensor such as a thermistor and a thermostat is embedded in a motor, and the winding temperature is directly detected by them to prevent overload operation of the motor (see, for example, Patent Document 1).However, in such a case, it is necessary to embed the temperature sensor into the motor, and it is difficult to adequately detect the temperature of the winding unless the temperature sensor is precisely located at a predetermined location.

[0003] In contrast, a technique related to electronic thermal analysis has been developed, in which a load situation is calculated from a current command applied to a motor without using a direct sensor such as a temperature sensor, and the overheating of a winding is determined. In such electronic thermal analysis, the overheating of the winding is determined by software. For example, in the technique disclosed in Patent Document 2, a winding resistance value is estimated based on parameters such as a voltage applied to a motor, a current, and an induced voltage of the motor, and a winding temperature is estimated from the estimated winding resistance value.In the method disclosed in Patent Document 3, the winding temperature at start-up is estimated from the value of the winding resistance measured at the start of the motor, and then the transition of the winding temperature is estimated based on the current flowing through the motor. PRIOR ART PATENT DOCUMENTS Patent Document 1: Japanese Unexamined Patent Publication JP H04 - 283 087 A. Patent Document 2: Japanese Unexamined Patent Publication JP 2011 - 015 584 A Patent Document 3: Japanese Unexamined Patent Publication JP H09 - 261 850 A

[0004] DE 11 2012 006 170 T5 describes a motor drive device for a multiphase motor with a coil, which has a function for estimating the motor coil temperature. Key features include: a motor coil ambient temperature estimation unit with a filter (phase advance and low-pass), a motor coil elevated temperature estimation unit (based on current), and a motor coil protection unit for current limitation.

[0005] JP H06 - 54 572 A describes a thermal protection device for a motor, which has the following features: detection of the motor winding current, the flange temperature and the rotor speed by means of sensors; identification of the winding resistance and the torque constant; conversion of these parameters into winding and rotor temperature; state estimation (state observer) of the temperatures based on a motor model; comparison of measured and estimated temperatures for fault detection; output of a protection signal to the controller and display of the temperature and fault states.

[0006] JP 2017 - 51 089 A describes a drive device for a motor with a coil and magnet, comprising a current driver and a control unit. The control unit generates current setpoints via two commutation calculations, calculated based on setpoints and feedback. A correction value is determined and added to compensate for the current driver's transfer function (F(s)-1-1). SUMMARY OF THE INVENTION PROBLEMS TO BE SOLVED BY THE INVENTION

[0007] To prevent winding overheating through the use of electronic thermal engineering (thermal engineering), it is preferable that the motor specifications are clear. That is, if physical parameters such as the winding resistance and the motor's induced voltage are clear in relation to the winding temperature, the winding temperature can be estimated more accurately. However, for a motor driven by a driver, the motor's physical parameters are not necessarily clear. Therefore, suppressing winding overheating through electronic thermal engineering must be done after securing a margin and is more of an overprotective measure.

[0008] In contrast, it is also possible to suppress overheating of the winding of a motor by attaching a temperature sensor to the winding and detecting the temperature. However, as described above, since the detection result depends greatly on the arrangement of the temperature sensor in the motor, it is not easy to realize winding protection with high accuracy. Even if an advantageous arrangement can be made, there is a predetermined heat capacity between the winding and the temperature sensor, and moreover, there is a response delay in the temperature sensor itself, so the detection of the temperature sensor itself has a delay. Therefore, in the case of using a temperature sensor, it is not easy to detect a rapid temperature change due to a large current in time.

[0009] The disclosure of the present application has been made in view of such a problem, and an object of the present disclosure is to provide a technique for accurately estimating the temperature of a winding with respect to the electronic thermals of a motor. MEANS TO SOLVE THE PROBLEM

[0010] In the disclosure of the present application, to solve the above problem, a configuration is adopted in which a winding temperature characteristic model and a predetermined temperature characteristic model for the electronic thermal circuit of a motor are determined by using a rising transition of a winding temperature obtained when a predetermined voltage is applied and a rising transition of a temperature detected by a temperature sensor arranged to detect the winding temperature. By reflecting the detection value of the temperature sensor in the electronic thermal circuit in this way, the temperature of the motor winding can be more accurately estimated by the electronic thermal circuit.

[0011] In particular, the present disclosure provides a processing device according to claim 1, which determines a calculation model for estimating a temperature of a winding, the calculation model being included in the electronic thermals of a motor including a stator around which the winding is wound and a rotor, the calculation model including a winding temperature characteristic model including a winding-related parameter related to a temperature characteristic of the winding, and a predetermined temperature characteristic model including a predetermined parameter related to a characteristic of a temperature in the vicinity of the winding detected by a temperature sensor arranged in the vicinity of the winding, the processing device including: a temperature transition detection unit configured to detect a first rising transition,which is a rising transition of a temperature of the winding, and a second rising transition, which is a rising transition of a temperature detected by the temperature sensor in a state in which the application of a voltage for raising the temperature of the winding to a predetermined temperature is performed; and a determination unit configured to determine the predetermined temperature characteristic model by calculating the predetermined parameter based on the second rising transition, and further configured to determine the winding temperature characteristic model by calculating the winding-related parameter based on the first rising transition.

[0012] The processing device of the present disclosure is configured to suppress an excessive temperature rise of a winding by electronic thermal technology with respect to a motor by disposing a temperature sensor near the winding to detect the winding temperature of the motor. A known temperature sensor can be used as the temperature sensor. The temperature sensor in the motor is preferably disposed at a position where a temperature detection unit of the temperature sensor comes into contact with the winding. However, the arrangement of the temperature sensor is not limited to a specific arrangement, as long as the temperature sensor is near the winding and can detect the temperature near the winding.

[0013] The processing device then determines the winding temperature characteristic model and the predetermined temperature characteristic model to enable estimation of the winding temperature at the time of voltage application using the calculation model including the winding temperature characteristic model and the predetermined temperature characteristic model. The winding temperature characteristic model is a model for calculating the temperature characteristic of the winding that results when the thermal influence of the stator in the motor is virtually eliminated. Examples of winding-related parameters included in the winding temperature characteristic model include a thermal resistance, a thermal time constant, and the like related to the winding.Furthermore, the predetermined temperature characteristic model is a model for calculating the characteristic curve / characteristic of the temperature detected by the temperature sensor when the thermal influence of the stator in the motor is virtually eliminated. It is assumed that the winding temperature is taken into account during detection by the temperature sensor. Examples of the predetermined parameters included in the predetermined temperature characteristic model include a thermal resistance, a thermal time constant, and the like related to the temperature detection by the temperature sensor.

[0014] Here, the temperature transition detection unit detects two temperature transitions (rise transitions) related to the winding in a state where a voltage is applied to raise the temperature of the winding to a predetermined temperature. One rise transition is a first rise transition, which is a temperature transition actually occurring in the winding, and the other rise transition is a second rise transition, which is a rise transition of the temperature detected by the temperature sensor. To estimate the winding temperature using the calculation model including the predetermined temperature characteristic model and the winding temperature characteristic model using electronic thermal engineering, it is necessary to appropriately consider the influence of the stator temperature transition on the winding temperature.

[0015] Here, the first rising transition is a winding temperature transition, and the second rising transition is a temperature transition detected by the temperature sensor located near the winding. Therefore, both the first rising transition and the second rising transition are assumed to be influenced by the temperature transition of the motor stator, which can be considered essentially the same. This means that by using the first rising transition and the second rising transition, it is possible to estimate the winding temperature with electronic thermal engineering, taking into account the influence of the stator temperature transition.Therefore, the determination unit in the processing device determines the predetermined temperature characteristic model by calculating the predetermined parameter based on the second rising transition, and further determines the winding temperature characteristic model by calculating the winding-related parameter based on the first rising transition. The calculation model, which includes the predetermined temperature characteristic model and the thus determined winding temperature characteristic model, enables an estimation of the winding temperature that appropriately reflects the influence of the stator temperature transition, thus absorbing the temperature detection delay in the temperature sensor.

[0016] In the processing device, the determination unit determines the predetermined temperature characteristic model by calculating the predetermined parameter based on the second rising transition via a stator temperature characteristic model that includes a stator-related parameter related to a temperature characteristic of the stator and has a correlation with the winding temperature characteristic model and the predetermined temperature characteristic model, and further determines the winding temperature characteristic model by calculating the winding-related parameter based on the first rising transition via the stator temperature characteristic model. The stator temperature characteristic model is a model for calculating the temperature characteristic of the stator that results when the thermal influence of the winding in the motor is virtually eliminated.Examples of stator-related parameters included in the stator temperature characteristic model include thermal resistance, thermal time constant, and similar stator-related parameters. By adopting such a configuration, it is possible to determine the specified temperature characteristic model and the winding temperature characteristic curve model that appropriately reflect the influence of the stator's temperature transition.

[0017] In the processing device, the temperature transition detection unit can detect the first rising transition based on the resistance value of the winding. When the predetermined temperature characteristic model and the winding temperature characteristic model are determined, the rising transition of the winding temperature can be accurately measured by a measurement method other than the temperature sensor.

[0018] The processing device may further include a frequency response acquisition unit configured to acquire a frequency response in the motor when a voltage applied to the winding is an input and a current flowing through the winding is an output, and a resistance calculation unit configured to calculate a resistance value of the winding based on the frequency response. By focusing on the electrical characteristics of the winding and using the frequency response as described above, the voltage applied to the winding to acquire the resistance value of the winding can be reduced as much as possible, the fluctuation in the winding temperature caused by the voltage application can be suppressed, and the resistance value of the winding can be measured more accurately.

[0019] Furthermore, in the processing device, the voltage application may be performed in a first cycle. The resistance calculation unit may calculate the resistance value of the winding at the time of voltage application based on the frequency response acquired by the frequency response acquisition unit according to the current output of the motor at the time of inputting the voltage application in the first cycle. The temperature transition acquisition unit may detect the first rising transition based on the resistance value of the winding calculated by the resistance calculation unit. According to such a configuration, the fluctuation of the winding temperature can be suppressed when measuring the first rising transition, and the resistance value of the winding can be measured more accurately.

[0020] Then, in the processing device described above, the rotor of the motor can be driven to rotate at a predetermined constant speed when the voltage application is performed. During temperature detection by the temperature sensor, there is a strong tendency for the temperature of the winding to be detected near the detection unit of the temperature sensor. On the other hand, when a voltage is applied to the winding to achieve the second rising transition, the temperature detection by the temperature sensor may vary if the applied current flows unevenly into any phase (e.g., in a case where the motor is a three-phase AC motor, any of the U, V, and W phases) constituting the winding of the motor.Therefore, by rotating the rotor at a predetermined constant speed as described above, the applied current can flow substantially uniformly through the windings of the motor, and the second rising transition can be appropriately detected.

[0021] To solve the above problem, the present disclosure can also be understood from the aspect of a determination method for a calculation model of the winding temperature according to claim 6. That is, the present disclosure provides a method for determining a calculation model included in an electronic thermal circuit of a motor including a stator around which a winding is wound and a rotor, the calculation model estimating a temperature of the winding, the calculation model including a winding temperature characteristic model including a winding-related parameter related to a temperature characteristic of the winding, and a predetermined temperature characteristic model including a predetermined parameter related to a characteristic of a temperature in the vicinity of the winding, which is detected by a temperature sensor arranged near the winding.The method includes: a step of detecting a first rising transition, which is a rising transition of the temperature of the winding, and a second rising transition, which is a rising transition of the temperature detected by the temperature sensor in a state where application of a voltage for raising the temperature of the winding to a predetermined temperature is performed; and a step of determining the predetermined temperature characteristic model by calculating the predetermined parameter based on the second rising transition and further determining the winding temperature characteristic model by calculating the winding-related parameter based on the first rising transition.In the determining step, the predetermined temperature characteristic model is determined by calculating the predetermined parameter based on the second rising transition via a stator temperature characteristic model that includes a stator-related parameter related to a temperature characteristic of the stator and has a correlation with the winding temperature characteristic model and the predetermined temperature characteristic model. Furthermore, the winding temperature characteristic model is determined by calculating the winding-related parameter based on the first rising transition via the stator temperature characteristic model. Furthermore, the technical idea disclosed for the processing device described above can be applied to the determination method of the winding temperature calculation model as long as no technical discrepancy occurs. EFFECT OF THE INVENTION

[0022] With regard to the electronic thermals of the motor, the temperature of the winding can be accurately estimated. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a diagram showing the configuration of a calculation model including a winding temperature characteristic model and a predetermined temperature characteristic model; Fig. 2A is a diagram for explaining a correlation between the winding temperature characteristic model and a stator temperature characteristic model, and Fig. 2B is a diagram for explaining a correlation between the predetermined temperature characteristic model and a stator temperature characteristic model; Fig. 3 is a diagram showing the transition of a voltage applied to a motor when the calculation model is fitted to a motor, as well as the transitions of a winding temperature and a temperature detected by a temperature sensor at that time; Fig. 4 is a schematic configuration of a control system configured by including the engine; Fig. Fig. 5 is a first diagram showing a control structure implemented by a servo driver of the Fig. 4; Fig. Figure 6 is a second diagram showing the control structure implemented by the servo driver of the Fig. 4; Fig. 7 is a flowchart showing the flow of a method for adapting the calculation model to the motor, which is executed by the servo driver; and Fig. 8A and Fig. 8B are diagrams illustrating the transitions of a voltage applied to the motor. EMBODIMENT OF THE INVENTION <anwendungsbeispiel>

[0023] An example of a processing device that uses a model parameter to estimate the temperature of a motor winding in the electronic thermals, ie in the thermal behavior of the motor, in Motor 2 (see Fig. 4) with the electronic thermal system, is set with reference to the Fig. 1 to 3. Note that, in the embodiment of the present disclosure, the motor 2 only needs to have a configuration in which a winding is wound around a stator and a rotor is provided, and a specific configuration thereof is not limited to any particular configuration. Furthermore, a temperature sensor 3 is arranged inside the motor 2 to detect the winding temperature thereof. Specifically, the temperature sensor 3 is arranged such that a detection unit of the temperature sensor 3 is located near the winding, preferably, the detection unit is in contact with the winding.

[0024] Fig. Figure 1 shows a schematic structure of a calculation model 10 for calculating the winding temperature contained in the electronic thermals of the motor. Fig. 2A and Fig. 2B Diagrams for describing the derivation of the calculation model 10 and illustrate a correlation between a winding temperature characteristic model and a stator temperature characteristic model or a correlation between a given temperature characteristic model and a stator temperature characteristic model. Fig. Figure 3 is a diagram illustrating the transition of the voltage applied to the motor when calculation parameters described in the Fig. 1, ie, a winding-related parameter and a predetermined parameter, are determined, and the transition of the winding temperature at that time.

[0025] Before explaining the calculation model 10, a calculation model (hereinafter referred to as the "basic calculation model") 20 is described here that can estimate the temperature of a winding taking into account the main heat sources that can affect the winding temperature in the motor 2, i.e., the winding and a stator, which are structures with a relatively large heat capacity (see Fig. 2A). The basic calculation model 20 is a program for calculating the winding temperature of the motor 2 and outputs the motor winding temperature when the applied energy / power in the motor is given as input. It should be noted that the applied energy can be considered a so-called copper loss caused by the electrical resistance of the motor winding coil and is physically proportional to the square of the current flowing through a winding coil. As shown in Fig. 1, the basic calculation model 20 includes a winding temperature characteristic model 21 and a stator temperature characteristic model 22 as submodels that form the basic calculation model 20. The winding temperature characteristic model 21 is a model for calculating the temperature characteristic of the winding that results when the thermal influence of the stator in the motor is practically eliminated. The stator temperature characteristic model 22 is a model for calculating the temperature characteristic of the stator that results when the thermal influence of the winding in the motor is practically eliminated. As described above, the basic calculation model 20 includes both the winding temperature characteristic model 21 and the stator temperature characteristic model 22, and the sum of the outputs of the models is calculated as the winding temperature of the motor, as shown in Fig. 2A. Therefore, the winding temperature of the motor is calculated taking into account the correlation between the stator and the winding.

[0026] Next, the winding temperature characteristic model 21 will be described. The winding temperature characteristic model 21 is expressed by the following formula 1, which includes a thermal resistance Ra and a thermal time constant Ta related to the winding, which are parameters (winding-related parameters) related to the temperature characteristics of the winding. Note that the thermal resistance Ra is a value indicating the difficulty of heat transfer and is a parameter representing a temperature rise per amount of heat generated in a unit time. In this embodiment, the thermal resistance is assumed to be the one obtained when the motor winding is considered a thermally homogeneous object.The thermal time constant Ta is a parameter that represents the degree of responsiveness of the winding to temperature changes and is defined as the time required to change 63.2% of the temperature difference when the winding transitions from one initial thermal equilibrium state to another thermal equilibrium state. Winding temperature characteristic model = Ra / (Ta⋅s+1)

[0027] Next, the stator temperature characteristic model 22 will be described. The stator temperature characteristic model 22 is expressed by the following formula 2, which includes a thermal resistance Rb and a thermal time constant Tb related to the stator, which are parameters (stator-related parameters) related to the temperature characteristic of the stator. Note that the definition of the thermal resistance Rb is the same as the above-described definition of the thermal resistance Ra, and in this embodiment, the thermal resistance obtained when the stator of the motor is considered a thermally homogeneous object is adopted. The thermal time constant Tb is a parameter representing the degree of response of the stator to temperature changes and corresponds to the above-described definition of the thermal time constant Ta. Stator temperature characteristic model = Rb / (Tb⋅s+1)

[0028] Then, in the basic calculation model 20, an input (applied energy in the motor) is passed to the winding temperature characteristic model 21 and the stator temperature characteristic model 22. The outputs of the respective models are then added to obtain an output of the basic calculation model 20, that is, an estimated temperature of the motor winding. Note that when adding the outputs of the respective models, the values ​​obtained by multiplying the outputs of the models by predetermined gains can be added. By configuring the basic calculation model 20 in this way, the winding temperature of the motor is estimated taking into account the correlation between the stator and the winding.

[0029] In a case where the basic calculation model 20 is used to estimate the winding temperature in this way, it is necessary to specify the stator temperature characteristic model 22. However, to specify the stator temperature characteristic model 22, it is preferable to create a situation in which the winding has as little thermal influence on the stator as possible, but this is not easy. Therefore, in the present application, the temperature transition detected by the temperature sensor 3 arranged near the winding of the motor 2 is used. Fig. 2B shows a calculation model (hereinafter referred to as "winding environment calculation model") 30 capable of estimating the temperature in the vicinity of the winding, that is, the temperature detected by the temperature sensor 3. The winding environment calculation model 30 is a program for calculating the temperature detected by the temperature sensor 3, and outputs the temperature detected by the temperature sensor 3 when the power applied to the motor 2 is given as input. As shown in Fig. 2B, the winding environment calculation model includes a predetermined temperature characteristic model 31 and a stator temperature characteristic model 32 as sub-models that form the winding environment calculation model 30.

[0030] The predetermined temperature characteristic model 31 is a model for calculating the detected temperature characteristic of the temperature sensor 3 (temperature characteristic near the winding) when the thermal influence of the stator in the motor is practically eliminated. It can be assumed that the temperature characteristic of the winding is reflected in the detected temperature characteristic. The stator temperature characteristic model 32 is a model for calculating the temperature characteristic of the stator when the thermal influence of the winding in the motor is practically eliminated. Therefore, the stator temperature characteristic model 32 can be considered the same as the stator temperature characteristic model 22 in Fig. 2A. As described above, the winding environment calculation model 30 includes both the predetermined temperature characteristic model 31 and the stator temperature characteristic model 32, and the sum of the outputs of the respective models is calculated as the temperature detected by the temperature sensor 3, as shown in Fig. 2B.

[0031] The predetermined temperature characteristic model 31 is expressed by the following formula 3, which includes the temperature detected by the temperature sensor 3, i.e., a thermal resistance Rs, and a thermal time constant Ts, which are parameters (predetermined parameters) related to the temperature characteristic near, i.e., the surroundings of, the winding. Note that the definition of the thermal resistance Rs is the same as the above-described definition of the thermal resistance Ra, and in this embodiment, the thermal resistance obtained when the motor winding and the surrounding space are considered as a thermally homogeneous object is adopted. The thermal time constant Ts is a parameter representing the degree of response of the winding and the surrounding space to temperature changes and corresponds to the above-described definition of the thermal time constant Ta. given temperature characteristic model=Rs / (Ts⋅s+1)

[0032] Since the stator temperature characteristic model 32 can be considered the same as the stator temperature characteristic model 22 described above, a detailed description will be omitted. Then, in the winding environment calculation model 30, an input (applied energy in the motor) is passed to the predetermined temperature characteristic model 31 and the stator temperature characteristic model 32. Then, the outputs of the predetermined temperature characteristic model 31 and the stator temperature characteristic model 32 are added to obtain an output of the winding environment calculation model 30, that is, an estimated value of the temperature detected by the temperature sensor 3.

[0033] In comparison of the Fig. 2A and Fig. 2B, both the basic calculation model 20 for estimating the winding temperature and the winding environment calculation model 30 for estimating the temperature detected by the temperature sensor 3 contain the stator temperature characteristic models 22 and 32. Therefore, the Fig. 1 can be derived by inserting the stator temperature characteristic model 32 derived from the winding environment calculation model 30 into the stator temperature characteristic model 22 contained in the basic calculation model 20.

[0034] Based on the above, the calculation model 10 contained in the electronic thermal system is described with reference to Fig. 1. The calculation model 10 is a program for calculating the motor winding temperature in the electronic thermal circuit of the motor and outputs the motor winding temperature when the applied energy / power in the motor is given as input. As shown in Fig. 1, the calculation model 10 comprises a winding temperature characteristic model 11 and a predetermined temperature characteristic model 12 as sub-models, which form the calculation model 10, as well as the temperature sensor 3. The winding temperature characteristic model 11 is a model for calculating the temperature characteristic, ie temperature characteristic, of the winding when the thermal influence of the stator in the motor is practically eliminated, and is identical to that in Fig. 2A shown winding temperature characteristic model 21. The specified temperature characteristic model 12 is a model for calculating the detected temperature characteristic of the temperature sensor 3 (temperature characteristic near / around the winding) when the thermal influence of the stator in the motor is practically eliminated, and is identical to that shown in Fig. 2B shown predefined temperature characteristic model 31.

[0035] In the calculation model 10 configured as described above, a part which is Fig. 1 is surrounded by a dashed line 15, a configuration corresponding to that shown in Fig. 2B. Therefore, when the input (applied energy / power in the motor) is supplied to the winding temperature characteristic model 11 and the predetermined temperature characteristic model 12, the difference between the output of the temperature sensor 3 and the predetermined temperature characteristic model 12 is added to the output of the winding temperature characteristic model 11, and the obtained value is the output of the calculation model 10, that is, the estimated temperature of the motor winding. Note that when adding the outputs of the respective models, the values ​​obtained by multiplying the outputs of the models by predetermined gains may be added.

[0036] Next, the calculation of the thermal resistance Ra and the thermal time constant Ta used in the winding temperature characteristic models 11 and 21, the calculation of the thermal resistance Rs and the thermal time constant Ts used in the predetermined temperature characteristic models 12 and 31, and the calculation of the thermal resistance Rb and the thermal time constant Tb used in the stator temperature characteristic models 22 and 32 are described with reference to Fig. 3. It should be noted that these parameters Ra, Rb, Rs, Ta, Tb and Ts are collectively referred to as model parameters.

[0037] In the upper part of Fig. 3, voltage is applied to the motor 2 to increase the winding temperature to a predetermined temperature during the period from time T1 to time T2. Since the winding temperature of the motor increases and converges during the period from time T1 to time T2, as shown by the line L1 in the lower part of Fig. 3 (rise from temperature t0 to temperature t1), the transition of the winding temperature during this period is called the first rising transition L1. In addition, the transition of the temperature actually detected by the temperature sensor 3, which is arranged near the motor winding for detecting the temperature, is indicated by the line L2 (rises from temperature t0 to temperature t2 and converges). Since there is a certain heat capacity between the winding and the temperature sensor 3, the temperature of the second rising transition L2 is slightly lower than that of the first rising transition L1. In addition, a voltage V1 is applied so that a current flows only in a d-axis, so that the rotor of the motor 2 does not rotate when a voltage is applied. In this way, accidental co-rotation of the drive shaft of the motor 2 when the voltage is applied can be avoided.Alternatively, the voltage V1 can be applied so that the rotor of motor 2 rotates at a predetermined low speed (e.g., several tens of revolutions per minute) when the voltage is applied. In this way, the current can flow evenly in the winding of motor 2, and the influence of the arrangement of temperature sensor 3 on the temperature detected by temperature sensor 3 can be reduced. Note that the voltage V1 applied at this time can be arbitrary as long as the motor temperature is raised to a temperature suitable for calculating the model parameters, and can be, for example, a voltage corresponding to the rated power of the motor.

[0038] The model parameters Ra, Rb, Rs, Ta, Tb, and Ts are calculated based on the first rising transition L1 and the second rising transition L2 from time T1 to time T2. First, based on the second rising transition L2, the thermal resistance Rs and the thermal time constant Ts are calculated with respect to the predetermined temperature characteristic models 12 and 31, and the thermal resistance Rb and the thermal time constant Tb are calculated with respect to the stator temperature characteristic models 22 and 32. Specifically, the model parameters Rs, Ts, Rb, and Tb are calculated based on the time required for the temperature detected by the temperature sensor 3 to rise from t0 to t2, the input power, and the like using the least squares method. As a result, the predetermined temperature characteristic models 12 and 31 and the stator temperature characteristic models 22 and 32 are determined.Subsequently, the thermal resistance Ra and thermal time constant Ta related to the winding temperature characteristic models 11 and 21 are calculated based on the first rising transition L1. Note that this calculation uses previously calculated values ​​for the thermal resistance Rb and thermal time constant Tb related to the stator temperature characteristic models 22 and 32. Similarly, the specific calculation of the winding temperature characteristic models 11 and 21 is performed based on the time required for the temperature detected by the temperature sensor 3 to rise from t0 to t1, the input power, and the like, using the least squares method. As a result, the winding temperature characteristic models 11 and 21 are determined.

[0039] The winding temperature characteristic model 11 and the predetermined temperature characteristic model 12 are created using the model parameters calculated in this way, and the calculation model 10 including both the winding temperature characteristic model 11 and the predetermined temperature characteristic model 12 is determined. By using the calculation model 10 obtained in this way, it is possible to estimate the winding temperature of the motor by electronic thermal analysis using the detection value of the temperature sensor 3. In the estimation, since the winding temperature is calculated by the calculation model 10 in a form that absorbs the detection delay included in the temperature sensor 3, it is possible to realize an estimation of the winding temperature with high accuracy and low delay, and thus it is possible to effectively prevent overheating of the winding of the motor 2. <Erstes Beispiel>

[0040] Fig. 4 is a schematic configuration diagram of a control system including a servo driver 4, which also functions as the processing device of the embodiment. The control system includes a network 1, the motor 2, the servo drive 4, and a standard programmable logic controller (PLC) 5. Note that the motor 2 includes the temperature sensor 3 as described above. The control system is a system for driving and controlling a load device (not shown) together with the motor 2. The motor 2 and the load device are controlled by the control system. Examples of load devices here include various mechanical devices (e.g., an arm of an industrial robot or a conveyor). The motor 2 is installed in the load device as an actuator that drives the load device.Motor 2, for example, is an AC servomotor with a stator around which a winding is wound and a rotor. An encoder (not shown) is attached to motor 2, and a parameter signal related to the operation of motor 2 is fed back to servo driver 4 via the encoder.

[0041] Parameter signal (hereinafter referred to as feedback signal) that is fed back includes, for example, position information about the rotational position (angle) of the rotating shaft of the motor 2, information about the rotating speed of the rotating shaft and the like.

[0042] The servo driver 4 receives an operation command signal regarding the operation (movement) of the motor 2 from the standard PLC 5 via the network 1 and receives the feedback signal output from the encoder connected to the motor 2. The servo driver 4 calculates a command value related to servo control for driving the motor 2, that is, the operation of the motor 2, based on the operation command signal from the standard PLC 5 and the feedback signal from the encoder, and supplies a drive current to the motor 2 so that the operation of the motor 2 follows the command value. Note that alternating current is used for the supply current, which is supplied to the servo driver 4 from an AC power supply 7. In this example, the servo driver 4 is a type that receives three-phase alternating current, but it may also be a type that receives single-phase alternating current.It should be noted that the servo control performed by the servo driver 4 is feedback control using a position controller 41, a speed controller 42, and a current controller 43 included in the servo driver 4; details of which will be described later with reference to FIG. Fig. 5 described.

[0043] As in Fig. As shown in Figure 4, the servo driver 4 includes the position controller 41, the speed controller 42, and the current controller 43. The position controller 41, the speed controller 42, and the current controller 43 execute processes to perform the servo control described above. The servo driver 4 includes an electronic thermal unit 100 (see Fig. 5) to protect the motor 2 from damage due to overload. The electronic thermal unit 100 estimates the winding temperature of the motor 2 and determines the overload condition of the motor 2 based on the estimated winding temperature. Therefore, the servo control performed by the servo driver 4 and the protective control of the motor 2 performed by the electronic thermal unit 100 are based on the values ​​specified in the Fig. 5. The control structure is formed by executing a predetermined control program in the servo driver 4, which includes a predetermined arithmetic device, a memory, and the like.

[0044] The position controller 41 performs, for example, proportional control (P control). Specifically, a speed command is calculated by multiplying the position deviation, i.e., the deviation between the position command reported by the standard PLC 5 and the detected position, by a position proportional gain Kpp. It should be noted that the position controller 41 has the position proportional gain Kpp as a control parameter in advance. Next, the speed controller 42 performs, for example, proportional-integral control (PI control). Specifically, a torque command is calculated by multiplying the integral amount of the speed deviation, i.e.,The deviation between the speed command calculated by the position controller 41 and the detected speed is multiplied by a speed integral gain Kvi, and the sum of the calculation result and the speed deviation is multiplied by a speed proportional gain Kvp. Note that the speed controller 42 has a speed integral gain Kvi and a speed proportional gain Kvp as control parameters in advance. Furthermore, the speed controller 42 can perform P control instead of PI control. In this case, the proportional gain Kvp is specified as the control parameter for the speed controller 42. Next, the current controller 43 generates a command voltage to drive an amplifier 44 based on the torque command calculated by the speed controller 42.The amplifier 44 outputs a drive current for driving the motor 2 according to the generated command voltage, thus driving and controlling the motor 2. The current controller 43 includes a filter (first-order low-pass filter) for a torque command and one or more notch filters, and has cutoff frequencies for the energy or power of these filters and the like as control parameters.

[0045] The control structure of the servo driver 4 includes a speed feedback system including the speed controller 42, the current controller 43, the motor 2 to be controlled, and the like as feedforward elements, and further includes a position feedback system including the speed feedback system and the position controller 41 as feedforward elements. With the control structure configured as described above, the servo driver 4 can control the motor 2 so that the motor 2 follows the position command provided by the standard PLC 5.

[0046] If an excessive load (e.g., a load exceeding the rated load of the motor 2) is applied to the motor 2 for a relatively long time when the motor 2 is servo-controlled in this way, an excessive current flows through the winding of the motor 2 for a long time. Therefore, the winding temperature may rise excessively and lead to burnout. To prevent the motor 2 from operating in the overload condition described above, the servo driver 4 includes the electronic thermal unit 100. The electronic thermal unit 100 specifically includes the Fig. 1 and an overload judgment unit 110. As described above, the calculation model 10 includes the winding temperature characteristic model 11 and the predetermined temperature characteristic model 12. When the power applied to the motor 2 is input to each model and the detection value of the temperature sensor 3 is input, the calculation model 10 outputs the winding temperature of the motor 2 as a result. Then, the overload judgment unit 110 judges whether there is a possibility that the motor 2 will reach the overload state, in other words, whether there is a possibility that the temperature of the winding of the motor 2 will rise excessively, based on the winding temperature, which is an output of the calculation model 10.It should be noted that in a case where the overload judgment unit 110 determines that the motor 2 is in an overload state, the servo driver 4 may stop driving the motor 2 to protect the motor 2.

[0047] Now, the control structure for adapting the calculation model 10 contained in the electronic thermal unit 100 to the motor 2 to be controlled by the servo driver 4 will be described with reference to Fig. 6. The servo driver 4 includes a model adaptation unit 200 for adapting the calculation model 10 to the motor 2. The model adaptation unit 200 calculates the model parameters of the calculation model 10 corresponding to the motor 2, ie, the thermal resistance Ra and the thermal time constant Ta of the winding temperature characteristic model 11 and the thermal resistance Rs and the thermal time constant Ts of the predetermined temperature characteristic model 12, the winding temperature characteristic model 11 and the predetermined temperature characteristic model 12 corresponding to the motor 2, and adapts the calculation model 10 to the motor 2 using these parameters. It should be noted that when adapting the calculation model 10, the Fig. 5, the current controller 43 and the amplifier 44 are used; however, the position controller 41 and the speed controller 42 are not used. Therefore, the position controller 41 and the speed controller 42 are Fig. 6 not shown.

[0048] Here, the model adaptation unit 200 comprises an application control unit 210, a temperature transition detection unit 220 and a determination unit 230. The application control unit 210 outputs to the current controller 43 a command for applying voltage for calculating the model parameters of the calculation model 10, ie, the values ​​shown in the upper part of Fig. 3. It should be noted that the voltage application is controlled by the application control unit 210 so that it is suitable for calculating the model parameters of the calculation model 10.

[0049] Based on the winding resistance value of the motor 2, the temperature transition detection unit 220 detects the first rising transition L1 and the second rising transition L2 of the winding temperature when the calculation model 10 is adjusted (when a voltage is applied). The detection of the winding temperature is performed according to the following formula 4. Winding temperature θ2=R2 / R1−(234.5+θ1)−234.5

[0050] R1 is the value of the winding resistance at the beginning of the voltage application (time T1 in Fig. 3).

[0051] θ1 is the winding temperature at the beginning of voltage application. For example, θ1 can be the ambient temperature around motor 2 (in a case where the ambient temperature can be detected by servo driver 4) or the detection value of the temperature sensor in the encoder mounted on motor 2.

[0052] R2 is the winding resistance value at the time the voltage is applied. The measurement of the winding resistance value R2 is described below.

[0053] When the application control unit 210 applies a voltage, the temperature transition detection unit 220 detects the winding temperature of the motor 2 at that time according to Formula 4.

[0054] The determination unit 230 calculates the thermal resistance Ra and the thermal time constant Ta of the winding temperature characteristic model 11 and the thermal resistance Rs and the thermal time constant Ts of the predetermined temperature characteristic model 12, wherein the winding temperature characteristic model 11 and the predetermined temperature characteristic model 12 correspond to the motor 2, based on the first rising transition L1 and the second rising transition L2 detected by the temperature transition detection unit 220. The calculation of these model parameters is performed as described above. Further, the determination unit 230 applies the calculated model parameters to the winding temperature characteristic model 11 and the predetermined temperature characteristic model 12 of the calculation model 10 to determine the respective models. As a result, the calculation model 10 for the Fig. 1 is adapted to the motor 2 itself, which is controlled by the servo driver 4.

[0055] Now, a method for adapting the calculation model 10 by the model adaptation unit 200 will be described with reference to Fig. 7 described. Fig. 7 is a flowchart showing the flow of the method for adapting the calculation model 10 by the model adaptation unit 200. First, in S101, immediately before the application control unit 210 starts applying the voltage, an initialization process is performed to acquire the winding resistance value (R1 in Formula 4) of the motor 2 and the winding temperature (θ1 in Formula 4). The winding resistance value is calculated based on a current value obtained when a measurement voltage is applied between the terminals of the motor. The winding temperature in this initialization process can be considered to be approximately equal to the outside air temperature because the motor 2 is placed in the environment for a sufficiently long time. Therefore, the outside air temperature or the temperature detected by the temperature sensor in the encoder installed in the motor 2 is acquired as the winding temperature in the initialization process.

[0056] Next, in S102, while the voltage is being applied by the application control unit 210, the temperature transition detection unit 220 detects the first rising transition L1 of the winding temperature of the motor 2 and the second rising transition L2 of the temperature detected by the temperature sensor 3. These rising transitions have a common voltage application period. If the voltage for calculating the resistance value is applied separately when the winding temperature rises due to the applied voltage, the control of the temperature rise will be hindered by the original voltage application. When applying the voltage to calculate the model parameters, the winding temperature of the motor 2 must be increased to t1. Therefore, it is difficult to adequately calculate the model parameters (thermal resistance and thermal time constant) if the temperature rise is disturbed each time the resistance value is calculated.Therefore, in this embodiment, when applying voltage to calculate the model parameters, the voltage is applied periodically to increase the winding temperature due to the voltage application. At the same time, the value of the winding resistance of motor 2 is calculated using the frequency response (i.e., the frequency characteristic) of the current to the applied voltage when the voltage application is used as the input to motor 2 and the current flowing through the winding is used as the output.

[0057] As especially in Fig. As shown in Figure 8A, a periodic sinusoidal voltage is applied during the application period (T1 to T2). At this time, the effective value (root mean square) of the sinusoidal voltage is the value shown in Fig. 3. By applying the periodic sinusoidal voltage in this way, the winding temperature of motor 2 can be raised to t1. When the periodic voltage is applied, the temperature transient detection unit 220 detects the applied voltage value and the current value flowing through the winding of motor 2 as input and output values, respectively. The frequency response of the output value to the input value reflects the electrical characteristics of motor 2, which is expressed by the following formula 5. Electrical characteristics of motor 2: (1 / R)-(1 / (Ts+1)). where R is a winding resistance of motor 2 and T is an electrical time constant of motor 2.

[0058] Therefore, the temperature transition detection unit 220 calculates the frequency response of the output value and further calculates the winding resistance R of the motor 2 according to the following formula 6 using the gain G (ω) and the phase P (ω) obtained based on the frequency response. [Mathematical Formula 1] R=1G(ω)1+(tanP(ω))2

[0059] Further, the temperature transition detection unit 220 replaces the winding resistance R calculated by Formula 6 with R2 in Formula 4 to calculate the winding temperature (θ2 in Formula 4) at the time of detecting the frequency response.

[0060] As described above, by using the frequency characteristic of the current flowing through the winding of the motor 2 at the time of voltage application, the temperature transition detection unit 220 can detect the first rising transition L1 and the second rising transition L2 using the winding resistance value without hindering the temperature increase process (process of increasing the winding temperature to t1 and process of increasing the temperature detected by the sensor 3 to t2) of the motor 2. Note that the detection timing of each rising transition by the temperature transition detection unit 220, that is, the detection timing of the frequency characteristic, can be appropriately set within a range in which each rising transition can be detected to such an extent that the model parameters can be calculated.

[0061] It should be noted that in the Fig. 8A, a sinusoidal voltage is applied continuously during the application period. Alternatively, as in Fig. 8B, a sinusoidal voltage can be applied intermittently as long as the winding temperature of the motor 2 can approach the equilibrium state of t1. At this time, the root mean square value of the intermittent sinusoidal voltage in the application period is the voltage V1. In addition, the cycle of the applied voltage at the time of voltage application can be appropriately determined as long as a suitable frequency characteristic is acquired for calculating the winding resistance value. If the cycle of the applied voltage becomes too long, the winding temperature tends to change rapidly due to the application of the voltage. Conversely, if the cycle of the applied voltage becomes too short, it becomes difficult to adequately reflect the electrical characteristics of the motor 2 in the frequency characteristic. Therefore, the frequency of the applied sinusoidal voltage is determined, for example, at 100 MHz.to 1 / 3 to 3 times, preferably 1 / 2 to 2 times, and more preferably to the frequency corresponding to the inverse of the electrical time constant of the motor 2. This enables balanced adjustment and detection of the temperature of the motor 2.

[0062] Next, in S103, it is judged whether a predetermined voltage application time suitable for calculating the model parameters has elapsed. As an example, the predetermined voltage application time may be an application time until the winding temperature of the motor 2 converges to t1. When the rate of increase of the winding temperature of the motor 2 is equal to or less than a predetermined threshold, it can be considered that the increase has converged. Note that the rate of increase is defined as the increase in the winding temperature per unit time. Furthermore, the threshold may be a predetermined fixed value. Alternatively, the threshold may be determined based on the rate of increase of the winding temperature immediately after the start of voltage application, that is, the rate of increase that is considered to be the highest in the application period. For example, the threshold may be1 / 10 of the highest assumed rate of increase or decrease should be used. If the judgment in S103 is positive, proceed to S104, and if the judgment is negative, repeat the process in S102 and the subsequent process to continue applying the voltage.

[0063] Next, in S104, as described with reference to the Fig. 1 to 3, the thermal resistance Rb and the thermal time constant Tb, which are model parameters of the stator temperature characteristic models 22 and 32, and the thermal resistance Rs and the thermal time constant Ts, which are model parameters of the predetermined temperature characteristic models 12 and 31, are calculated based on the second rising transition L2 detected in S102, and the respective models are determined. Furthermore, in S105, the thermal resistance Ra and the thermal time constant Ta, which are model parameters of the winding temperature characteristic models 11 and 21, are calculated based on the first rising transition L1 detected in S102, and the winding temperature characteristic models 11 and 21 are determined.

[0064] As described above, according to the Fig. 7, the appropriate calculation model 10 adapted to the motor 2 driven by the servo driver 4 can be created. When the motor 2 is driven, the winding temperature can be accurately estimated by the electronic thermal unit 100 while appropriately suppressing the detection delay of the temperature sensor 3, and the motor 2 can be appropriately protected from overload. <Anderes Beispiel>

[0065] In the example described above, the model adaptation unit 200 is housed in the servo driver 4. Alternatively, the model adaptation unit 200 may be formed in a processing device (e.g., a personal computer (PC) or the like) that can be electrically connected to the servo driver 4. The processing device is a device for adapting the calculation model to the motor 2 and includes adaptation software (a program). Specifically, the processing device is a computer having a computing device, a memory, and the like, and the Fig. The method for adapting the calculation model shown in Figure 7 is realized by installing a program that can be executed in it and executing the program.

[0066] The dimensions, materials, shapes and relative arrangement of the configurations described in the above-described embodiment, the order of the respective processes included in the method in the above-described embodiment, and the like are not intended to limit the technical scope of the invention only to these unless otherwise specified. <nachtrag>

[0067] A processing device (4) that determines a calculation model (10) for estimating a temperature of a winding, wherein the calculation model (10) is contained in an electronic thermal unit (100) of a motor (2) that includes a stator around which the winding is wound and a rotor, wherein the calculation model (10) includes a winding temperature characteristic model (11) that includes a winding-related parameter related to a temperature characteristic of the winding, and a predetermined temperature characteristic model (12) that includes a predetermined parameter related to a characteristic of a temperature in the vicinity of the winding, which is detected by a temperature sensor (3) arranged in the vicinity of the winding, wherein the processing device (4) includes: a temperature transition detection unit (220) configured to detect a first rising transition (L1), which is a rising transition of a temperature of the coil, and a second rising transition (L2), which is a rising transition of a temperature detected by the temperature sensor in a state in which application of a voltage for increasing the temperature of the coil to a predetermined temperature is performed; and a determination unit (230) configured to determine the predetermined temperature characteristic model (12) by calculating the predetermined parameter based on the second rising transition (L2), and further configured to determine the winding temperature characteristic model (11) by calculating the winding-related parameter based on the first rising transition (L1).

[0068] A method for determining a winding temperature calculation model, which is a method for determining a calculation model (10) for estimating a temperature of a winding, wherein the calculation model (10) is included in an electronic thermal unit (100) of a motor (2) having a stator around which the winding is wound and a rotor, the calculation model (10) including a winding temperature characteristic model (11) including a winding-related parameter relating to a temperature characteristic of the winding, and a predetermined temperature characteristic model (12) including a predetermined parameter relating to a characteristic of a temperature in the vicinity of the winding, which is detected by a temperature sensor (3) arranged in the vicinity of the winding, the determination method comprising: a step (S102) for detecting a first rising transition (L1), which is a rising transition of a temperature of the coil, and a second rising transition (L2), which is a rising transition of a temperature detected by the temperature sensor in a state in which the application of a voltage for raising the temperature of the coil to a predetermined temperature is performed; and a step (S104, S105) for determining the predetermined temperature characteristic model (12) by calculating the predetermined parameter based on the second rising transition (L2) and for further determining the winding temperature characteristic model (11) by calculating the winding-related parameter based on the first rising transition (L1). LIST OF REFERENCE SYMBOLS 2 engines 3 Temperature sensor 4 servo drivers 10 Calculation model 11 Winding temperature characteristic model 12 predetermined temperature characteristic model 100 electronic thermal unit 200 model adaptation unit 210 Application control unit 220 Temperature transition detection unit 230 Determination unit< / nachtrag> < / anwendungsbeispiel>

Claims

[1] A processing device (4) that determines a calculation model (10, 30) for estimating a temperature of a winding, the calculation model (10, 30) being included in an electronic thermal unit (100) of a motor (2) that includes a stator around which the winding is wound and a rotor, the calculation model (10, 30) including a winding temperature characteristic model (11, 21) including a winding-related parameter related to a temperature characteristic of the winding, and a predetermined temperature characteristic model (12, 31) including a predetermined parameter related to a characteristic of a temperature in the vicinity of the winding, which is detected by a temperature sensor (3) arranged in the vicinity of the winding, the processing device (4) comprising: a temperature transition detection unit (220) configured to detect a first rising transition, which is a rising transition of a temperature of the winding, and a second rising transition, which is a rising transition of a temperature detected by the temperature sensor (3) in a state in which a voltage application for increasing the temperature of the winding to a predetermined temperature is performed; and a determination unit (230) configured to determine the predetermined temperature characteristic model (12, 31) by calculating the predetermined parameter based on the second rising transition, and further configured to determine the winding temperature characteristic model (11, 21) by calculating the winding-related parameter based on the first rising transition, wherein the determination unit (230) determines the predetermined temperature characteristic model (12, 31) by calculating the predetermined parameter based on the second rising transition via a stator temperature characteristic model (22, 32) that includes a stator-related parameter related to a temperature characteristic of the stator and has a correlation with the winding temperature characteristic model (11, 21) and the predetermined temperature characteristic model (12, 31),and determines the winding temperature characteristic model (11, 21) by calculating the winding-related parameter based on the first rising transition via the stator temperature characteristic model (22, 32). [2] The processing device (4) according to claim 1, wherein the temperature transition detection unit (220) detects the first rising transition based on a resistance value of the winding. [3] Processing device (4) according to claim 2 further comprises: a frequency response detection unit configured to detect a frequency response in the motor (2) obtained when a voltage applied to the winding is an input and a current flowing through the winding is an output; and a resistance calculation unit configured to calculate a resistance value of the winding based on the frequency response. [4] Processing device (4) according to claim 3, wherein the voltage application is carried out in a first cycle when the voltage is applied, the resistance calculation unit calculates a resistance value of the winding at the time of voltage application based on the frequency response detected by the frequency response detection unit corresponding to a current output by the motor (2) when the voltage application is input in the first cycle, and the temperature transition detection unit (220) detects the first rising transition based on the resistance value of the winding calculated by the resistance calculation unit. [5] Processing device (4) according to one of claims 1 to 4, wherein the rotor of the motor (2) is driven at a predetermined constant speed when the voltage is applied. [6] A method of determining a winding temperature calculation model for estimating a temperature of a winding, the winding temperature calculation model being included in an electronic thermal unit (100) of a motor (2) including a stator around which the winding is wound and a rotor, the winding temperature calculation model including a winding temperature characteristic model (11, 21) including a winding-related parameter relating to a temperature characteristic of the winding, and a predetermined temperature characteristic model (12, 31) including a predetermined parameter relating to a characteristic of a temperature in the vicinity of the winding, which is detected by a temperature sensor (3) arranged in the vicinity of the winding, the method comprising: a step of detecting a first rising transition, which is a rising transition of a temperature of the coil, and a second rising transition, which is a rising transition of a temperature detected by the temperature sensor (3) in a state in which a voltage application for increasing the temperature of the coil to a predetermined temperature is performed; and a step of determining the predetermined temperature characteristic model (12, 31) by calculating the predetermined parameter on the basis of the second rising transition and of further determining the winding temperature characteristic model (11, 21) by calculating the winding-related parameter on the basis of the first rising transition, wherein in the determining step, the predetermined temperature characteristic model (12, 31) is determined by calculating the predetermined parameter on the basis of the second rising transition via a stator temperature characteristic model (22, 32) which includes a stator-related parameter relating to a temperature characteristic of the stator and having a correlation with the winding temperature characteristic model (11, 21) and the predetermined temperature characteristic model (12, 31), and the winding temperature characteristic model (11,21) is determined by calculating the winding-related parameter based on the first rising transition via the stator temperature characteristic model (22, 32).,

Images