Simulation device and program

Abstract

To provide a simulation device which can effectively realize a configuration of virtually generating an abnormal state of a motor.SOLUTION: A simulation device (1) includes: a model storage unit (2) in which a motor physical model (2111) and an abnormal state model (211A) are stored, the motor physical model being a model modelled by a coil circuit unit (2111B) and a motion equation part (2111A) of a rotation and the abnormal state model being a model of an abnormal state of the motor; and a model operation unit (3) configured to perform operation processing by using the motor physical model. An abnormal parameter, which represents the deviation amount from a normal state calculated by the abnormal state model, is an input of the motor physical model.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present disclosure relates to a simulation device. [Background technology]

[0002] Conventionally, in the field of industrial machinery, the application of AI (artificial intelligence) to condition-based maintenance of mechanical systems in factory equipment maintenance has been progressing (for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2019 / 035279

[0004] [overview] When introducing AI to motor anomaly detection, it would be useful, for example, to be able to confirm the effectiveness of AI-based anomaly detection without using actual equipment such as motors.

[0005] In view of the above circumstances, an object of the present disclosure is to provide a simulation device that can effectively realize a configuration for virtually generating an abnormal state of a motor.

[0006] A simulation device according to an embodiment of the present disclosure includes: A motor physical model modeled by a winding circuit section and a rotational motion equation section; An abnormal state model that models an abnormal state of the motor; A model storage unit in which the following is stored: a model calculation unit configured to perform calculation processing using the motor physical model; Equipped with An abnormality parameter that indicates the amount of deviation from a normal state calculated by the abnormal state model is configured to be input to the motor physical model. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a diagram illustrating a configuration of a computer according to an exemplary embodiment of the present disclosure. [Diagram 2] FIG. 2 is a diagram illustrating a configuration of a simulation device according to an exemplary embodiment of the present disclosure. [Diagram 3] FIG. 3 is a side view (left) and a front view (right) showing a schematic configuration example of a motor. [Figure 4] FIG. 4 is a diagram showing the configuration of the motor model. [Diagram 5] FIG. 5 is a diagram showing the input / output relationship between the equation of motion part and the winding circuit part. [Figure 6] FIG. 6 is a diagram showing an example of a plot of the back electromotive force constant versus the rotation speed. [Figure 7] FIG. 7 is a diagram showing an example of a plot of the motor terminal current, the back electromotive force constant, and the motor torque with respect to the rotation speed. [Figure 8] FIG. 8 is a schematic plan sectional view showing a configuration example of a brushed DC motor. [Figure 9] FIG. 9 is a development view of the configuration shown in FIG. 8 developed in the circumferential direction. [Figure 10] FIG. 10 is a diagram showing model parameters in the motor physical model. [Figure 11] FIG. 11 is a diagram showing the gaps between the commutator segments. [Figure 12] FIG. 12 is a diagram for explaining the contact resistance between the commutator segments and the brushes. [Figure 13] FIG. 13 is a diagram showing the rotational movement of the winding sides. [Figure 14] FIG. 14 is a diagram showing an example of a magnetic flux density distribution. [Figure 15] FIG. 15 is a diagram showing the positional deviation between the permanent magnet and the brush. [Figure 16] FIG. 16 is a diagram showing the lengths of the sides of the windings. [Figure 17] FIG. 17 is a diagram showing the distance of the winding sides from the axis of rotation. [Figure 18]FIG. 18 is a diagram showing a filter configuration near the motor terminal. [Figure 19] FIG. 19 is a schematic plan view, a perspective view, and a development view showing the rotational movement of the windings. [Figure 20] FIG. 20 is a schematic development showing the rotational movement of the windings. [Figure 21] FIG. 21 is a diagram showing an overall view of the winding circuit section. [Figure 22] FIG. 22 is a table showing a method for calculating the contact resistance. [Figure 23] FIG. 23 is a diagram showing the configuration of a motor physical model when modeling is performed using Simscape (registered trademark) / Simulink (registered trademark). [Figure 24] FIG. 24 is a diagram showing a partial configuration of the winding circuit model unit. [Diagram 25] FIG. 25 is a diagram showing an example of a simulation result. [Figure 26] FIG. 26 is a diagram showing an example of a comparison between a simulation and an actual device. [Figure 27] FIG. 27 is a diagram showing the relationship between the equation of motion part and the bearing lubrication failure model. [Figure 28] FIG. 28 is a diagram showing the relationship between the equation of motion part and the bearing damage model. [Figure 29] FIG. 29 shows a front view and a cross-sectional side view of the bearing and shaft. [Diagram 30] FIG. 30 is a diagram showing bearing parameters. [Diagram 31] FIG. 31 is a diagram for explaining the revolution angle and rotation angle of the rolling elements. [Diagram 32] FIG. 32 is a diagram showing an example of a vibration model of a support system. [Diagram 33] FIG. 33 is a diagram illustrating an example of the configuration of a sensor model. [Diagram 34] FIG. 34 is a diagram illustrating an example of the configuration of a machine learning model. [Diagram 35] FIG. 35 is a diagram showing the configuration of a three-layer neural network. [Diagram 36]FIG. 36 is a diagram showing an example of a setting screen for the motor type and abnormal state. [Figure 37] FIG. 37 is a diagram showing an example of the motor basic setting screen. [Figure 38] FIG. 38 is a diagram showing an example of the support system setting screen. [Figure 39] FIG. 39 is a diagram showing an example of the first abnormal condition setting screen. [Diagram 40] FIG. 40 is a diagram showing an example of the second abnormal condition setting screen. [Diagram 41] FIG. 41 is a diagram showing an example of a time / deterioration setting screen. [Diagram 42] FIG. 42 is a diagram showing an example of the sensor setting screen. [Diagram 43] FIG. 43 is a diagram showing an example of an AI setting screen. [Diagram 44] FIG. 44 is a diagram showing an example of the abnormality determination setting screen. [Diagram 45] FIG. 45 is a diagram showing an example of a simulation result screen.

[0008] [Detailed Description] Hereinafter, exemplary embodiments of the present disclosure will be described with reference to the drawings.

[0009] <Computer configuration> 1 is a diagram showing a configuration of a computer 100 according to an exemplary embodiment of the present disclosure. The computer 100 functions as a simulation device according to the present disclosure, which will be described later. The computer 100 is, for example, a PC (personal computer). When the computer 100 is a PC, it does not matter whether it is a desktop type or a notebook type.

[0010] The computer 100 includes a CPU (Central Processing Unit) 100A, a memory 100B, an auxiliary storage device 100C, an operation input unit 100D, and a display unit 100E.

[0011] The CPU 100A has a control device and an arithmetic device (neither shown). The control device interprets program instructions and controls each part of the computer 100. The arithmetic device is a device that performs arithmetic processing.

[0012] The memory 100B is a semiconductor storage device that temporarily stores programs or data. The information stored in the memory 100B is erased when the computer 100 is powered off.

[0013] The auxiliary storage device 100C is configured with a hard disk drive (HDD) or a solid state drive (SSD) and stores programs or data. The programs stored in the auxiliary storage device 100C are loaded into the memory 100B. The CPU 100A executes the programs loaded into the memory 100B.

[0014] The operation input unit 100D is configured with a keyboard, a mouse, or the like, and is a device for providing operation input to the computer 100. Information input from the operation input unit 100D is sent to the memory 100B.

[0015] The display unit 100E is configured with, for example, a liquid crystal display, and converts the information acquired from the memory 100B into an image and outputs it.

[0016] <Configuration of simulation device> 2 is a diagram showing a configuration of a simulation device 1 according to an exemplary embodiment of the present disclosure. The simulation device 1 is a device capable of simulating anomaly detection by machine learning (AI) for a motor system. However, a physical system to which the simulation device of the present disclosure can be applied is not limited to a motor system.

[0017] The simulation device 1 includes a model storage unit 2, a model calculation unit 3, a model setting unit 4, a display control unit 5, an operation input unit 6, a display unit 7, and a data storage unit 8. A program P (FIG. 1) stored in an auxiliary storage device 100C of the computer 100 is a program for causing the computer 100 to function as the simulation device 1.

[0018] The model storage unit 2 stores a system model 21, a sensor model 22, a machine learning model 23, and an abnormality determination model 24, and is configured by an auxiliary storage device 100C of the computer 100. The system model 21, the sensor model 22, the machine learning model 23, and the abnormality determination model 24 are configured as a program P using, for example, MATLAB (registered trademark) / Simulink (registered trademark). Details of each model will be described later.

[0019] The functions of the model calculation unit 3, the model setting unit 4, and the display control unit 5 are realized by the CPU 100A executing the program P. The operation input unit 6 and the display unit 7 correspond to the operation input unit 100D and the display unit 100E in the computer 100, respectively.

[0020] The model calculation unit 3 executes a simulation by performing calculation processing for each model stored in the model storage unit 2. The model setting unit 4 performs settings (parameter settings, model selection settings, etc.) for each model stored in the model storage unit 2 in response to an input from the operation input unit 6. The simulation by the model calculation unit 3 is performed in accordance with the settings made by the model setting unit 4. The display control unit 5 controls the display unit 7 to display a model setting screen (described later) in response to an input from the operation input unit 6.

[0021] The data storage unit 8 also stores state monitoring data DT held by the user who uses the simulation device 1. The data storage unit 8 corresponds to the auxiliary storage device 100C in the computer 100. For example, if the user is considering introducing anomaly detection using machine learning (AI) to equipment owned by the user, the state monitoring data DT obtained by monitoring a motor in the equipment is stored in the data storage unit 8. The state monitoring data DT is stored in the data storage unit 8 by acquiring data from outside the simulation device 1, for example, via a network or a Universal Serial Bus (USB).

[0022] When a simulation using the condition monitoring data DT is executed, a predetermined input is made by the operation input unit 6. Then, the model calculation unit 3 inputs the condition monitoring data DT to the machine learning model 23, thereby performing learning and inference using the machine learning model 23. After learning is performed using normal data included in the condition monitoring data DT, inference can be performed using abnormal data. As described below, the machine learning model 23 outputs an abnormality degree, and the abnormality determination model 24 performs an abnormality determination based on the abnormality degree. In this way, the effect of anomaly detection by machine learning can be confirmed using the condition monitoring data DT held by the user, and a simulation that matches the equipment environment held by the user can be performed.

[0023] <System model> Next, the system model 21 will be described. The system model 21 is a model that represents a physical model of the motor system. The system model 21 makes it possible to virtually generate physical signal waveforms in normal and abnormal states of the motor system.

[0024] The system model 21 includes a motor model 211 , a driver model 212 , and a load model 213 .

[0025] The driver model 212 is a model of a driver for driving a motor. For example, when the motor is a brushed DC motor (hereinafter, a BDC motor), the driver can be a circuit that applies a DC voltage to the motor using one switch, or an H-bridge circuit. The H-bridge circuit is configured using two half bridges. The half bridge is configured from two switching elements connected in series between an application end of a DC voltage and a ground end (an application end of a ground potential). Also, for example, when the motor is a brushless DC motor (hereinafter, a BLDC motor), the driver can be a circuit configured from three half bridges corresponding to a three-phase motor.

[0026] The driver in the driver model 212 may be selectable and set by the model setting unit 4. For example, when the motor in the motor model 211 is the same BDC motor, the circuit using one switch as described above and the H-bridge circuit may be selectable. Also, for example, when the motor in the motor model 211 is a BLDC motor, the driver may be selectable according to the number of phases of the motor.

[0027] The load model 213 is a model of an object driven by the motor in the motor model 211. The driven object is, for example, a fan in a fan device, an arm in an industrial robot, etc. The load model 213 provides the motor model 211 with information on an external torque.

[0028] <Motor model> Here, the motor model 211 will be described. The motor model 211 is a model in which a motor (an example of a physical system) is physically modeled. The motor is, for example, a BDC motor or a BLDC motor. In this embodiment, the type of motor for the motor model 211 can be selected by the model setting unit 4. At this time, for example, multiple types of BDC motors and BLDC motors may be selectable. Different types of BDC motors include, for example, different numbers of pole pairs, different numbers of windings, and different ways of winding the windings. Different types of BLDC motors include, for example, different numbers of phases, different numbers of poles, and different numbers of slots.

[0029] FIG. 3 is a side view (left) and a front view (right) showing a schematic configuration example of a motor. Note that FIG. 3 is illustrated as a common configuration diagram regardless of the type of motor (e.g., BDC motor, BLDC motor). Note that the Cartesian coordinate system shown in FIG. 3 is a stationary coordinate system fixed to the base 201. The X-axis, the Y-axis, and the Z-axis are orthogonal to each other. The Z-axis is taken to extend in the direction in which the shaft 20C extends and pass through the center of the shaft 20C. The Y-axis is taken to extend perpendicular to the plane of the base 201. The X-axis extends parallel to the above plane and extends in the horizontal direction. In FIG. 3, as an example, the origin O of the Cartesian coordinate system is included in the rotor 20B.

[0030] As shown in FIG. 3, the motor 20 is fixed on a base 201. The motor 20 has a housing 20A, a rotor 20B, a shaft 20C, a stator 20D, and a bearing 20E. The rotor 20B, the shaft 20C, the stator 20D, and the bearing 20E are accommodated inside the housing 20A. The stator 20D is fixed to the housing 20A. The rotor 20B is disposed on the inner periphery side of the stator 20D. The shaft 20C protrudes from the rotor 20B on both sides in the rotation axis direction (Z axis direction). The shaft 20C is rotatably supported by the bearings 20E on both sides in the rotation axis direction. The bearings 20E are fixed to the housing 20A. Note that, in FIG. 3, the rotor 20B is an inner rotor type located on the inner periphery side of the stator 20D, but an outer rotor type in which the rotor is located on the outer periphery side of the stator may be used.

[0031] For example, when the motor 20 is a BDC motor, the stator 20D includes a magnet, and the rotor 20B includes an iron core, a winding, and a commutator. The brushes included in the BDC motor can come into contact with the commutator. When a current flows from the brushes through the commutator to the winding, the rotor 20B rotates due to the interaction between the magnetic field lines generated in the winding and the magnetic field lines of the magnet.

[0032] For example, when motor 20 is a BLDC motor, stator 20D includes an iron core and windings, rotor 20B includes a magnet, and rotor 20B rotates when a current is passed through the windings.

[0033] The rotor 20B and the shaft 20C rotate about the rotation axis, thereby driving a load connected to the shaft 20C.

[0034] 4 is a diagram showing the configuration of the motor model 211. The motor model 211 has a motor physical model 2111. The motor physical model 2111 has an equation of motion part 2111A and a winding circuit part 2111B.

[0035] FIG. 5 is a diagram showing the input / output relationship between the equation of motion unit 2111A and the winding circuit unit 2111B. An input voltage Vin is input to the winding circuit unit 2111B, and a motor terminal current im is calculated and output. The input voltage Vin is applied between the motor positive terminal Tp and the motor negative terminal Tn as shown in FIG. 9 described later, and the motor terminal current im is a current flowing through the motor terminal. The motor terminal current im is input to the equation of motion unit 2111A, and a mechanical angle velocity ωm and a mechanical angle θm of the shaft 20C are calculated and output. The mechanical angle velocity ωm and the mechanical angle θm are fed back to the winding circuit unit 2111B.

[0036] The equation of motion unit 2111A has the following equation (1) as the equation of motion.

number

[0037] The motor torque Tm is expressed by the following equation (2).

number

[0038] In a motor, torque is generated by the interaction of the magnetic flux distribution of the permanent magnets and the magnetic flux distribution due to the winding current. The contribution of the magnetic flux distribution due to the permanent magnets to the torque is determined by the shape and arrangement of the magnetic poles, as well as the geometric arrangement of the windings, and is a constant gain-like contribution that is unrelated to the rotation speed and the value of the motor terminal current. Therefore, as expressed in the above equation (2), the motor torque Tm is the product of the torque constant Kt, which is a constant coefficient, and the motor terminal current im.

[0039] In addition, the torque constant Kt has a relationship with the back electromotive force constant Ke, Kt=Ke, and the back electromotive force Vbemf and the mechanical angular velocity ωm have a relationship of Vbemf=Ke·ωm. The back electromotive force is the voltage generated at both ends of the motor terminals of the motor to be modeled when the shaft of the motor to be modeled is connected to another motor and the shaft is rotated at a constant speed by the other motor. The back electromotive force Vbemf was measured while changing the rotation speed, and Ke=Vbemf / ωm was calculated. As a result, as shown in Figure 6, the back electromotive force constant was almost constant regardless of the rotation speed. Therefore, the back electromotive force constant can be considered as a constant, and the average value of the plotted back electromotive force constant was set as the back electromotive force constant Ke. This value can be set as the value of the torque constant Kt. Since the back electromotive force constant Ke is a constant in this way, it is the basis for the above formula (2).

[0040] In the equation of motion (1) above, the left side represents the product of the inertia Jm and the mechanical angle acceleration, and the right side represents the combined torque of the motor torque Tm generated by application of the input voltage Vin to the motor terminals, the loss torque Tloss which collectively represents various losses, and the external torque Tex. The external torque Tex corresponds to the torque output from the load model 213 and the torque output from a person or the environment.

[0041] This is a component of the loss torque Tloss.

number

[0042] As mentioned above, the loss torque is assumed to be a quadratic expression of the mechanical angular velocity ωm. The assumption of the loss torque is that in a normal state, at a constant rotation speed, and without any external torque applied, the motor torque Tm - loss torque = 0, i.e.

number

[0043] First, the average value of the motor terminal current was measured while changing the rotation speed and changing the input voltage Vin at each rotation speed. The measurement results are shown in the upper left of Figure 7. Using the average value of the motor terminal current and the back electromotive voltage constant Ke (= torque constant Kt) shown in Figure 6 described above, the motor torque Tm was calculated from the above formula (2). As shown in Figure 7, the motor torque Tm is obtained for each rotation speed. Since the motor torque Tm and the loss torque are balanced during constant speed rotation, the plot of the motor torque Tm shown in Figure 7 can be considered as a plot of the loss torque. Since the coefficient of determination is good if the plot shown in Figure 7 is regressed with a quadratic equation, the loss torque was made a quadratic equation of the mechanical angle angular velocity ωm as described above.

[0044] <BDCモータ> Before describing the winding circuit section 2111B, a BDC motor, which is an example of a modeling target of the winding circuit section 2111B, will be described in more detail. FIG. 8 is a schematic planar cross-sectional view showing an example of the configuration of motor 20 when it is a BDC motor. FIG. 8 is a view seen in the direction of rotation axis J. Rotation axis J coincides with the Z-axis described above. In the following, the direction in which rotation axis J extends will be referred to as the axial direction, the direction around rotation axis J as the circumferential direction, and the direction perpendicular to rotation axis J as the radial direction.

[0045] Stator 20D includes permanent magnets Mg and brushes BR. In the configuration of Fig. 8, permanent magnets Mg include magnets MgS1, MgN1, MgS2, and MgN2. Magnets MgS1, MgN1, MgS2, and MgN2 respectively arrange S poles, N poles, S poles, and N poles in the circumferential direction on the radially inner side. In other words, magnetic poles are arranged alternately in the circumferential direction as S poles and N poles on the radially inner side.

[0046] The brushes BR include positive and negative brushes, as described below. The brushes have polarities that alternate in the circumferential direction.

[0047] The rotor 20B includes an iron core 202, a winding WR, and a commutator bar CM. The iron core 202 is formed, for example, by stacking electromagnetic steel sheets in the axial direction. The iron core 202 is disposed radially inside the permanent magnets Mg. The iron core 202 has an annular portion 202A and teeth 202B. The annular portion 202A extends in the axial direction and is formed in an annular shape in the circumferential direction. The teeth 202B protrude radially outward from the outer circumferential surface of the annular portion 202A. A plurality of teeth 202B are arranged in the circumferential direction.

[0048] In the configuration of FIG. 8, the windings WR include windings WR1 to WR16, i.e., 16 windings. Note that FIG. 8 representatively illustrates only windings WR1 to WR4. Each winding WR is wound around a tooth 202B so as to pass through one circumferential side of a certain tooth 202B (illustrated as θ1) and the other circumferential side of a second tooth 202B from the tooth 202B. The teeth 202B around which each winding WR is wound are shifted by one in the circumferential direction. In this way, the configuration of FIG. 8 is a concentrated winding configuration.

[0049] The commutator segments CM are arranged radially inward of the iron core 202 and radially outward of the brushes BR. In the configuration of Fig. 8, the commutator segments CM include commutator segments CM1 to CM16 (reference numbers are not shown in Fig. 8), i.e., 16 commutator segments. The commutator segments CM1 to CM16 are arranged in a circular ring shape in the circumferential direction. One side and the other side of the lead wires drawn from each winding WR are connected to the commutator segments CM adjacent in the circumferential direction. The commutator segments CM connected to each winding WR are shifted one from the other in the circumferential direction.

[0050] The commutator segments CM are capable of coming into contact with the brushes BR. As the rotor 20B rotates, the commutator segments CM rotate, and which commutator segments CM come into contact with the brushes BR and the contact resistance thereof change over time.

[0051] Fig. 9 is a development view of the configuration shown in Fig. 8 developed in the circumferential direction. Fig. 9 shows the rotation direction θrt of the rotor 20B. The rotation direction θrt is the same direction as θ1 on one side in the circumferential direction shown in Fig. 8. The S pole and the N pole are arranged along the rotation direction θrt. In addition, the commutator segments CM1 to CM16 are arranged along the rotation direction θrt. Specifically, the commutator segments CM16 → CM15 → ... are lined up in the order of the commutator segments CM16 → CM15 → ... along the rotation direction θrt, and when the commutator segment CM1 is lined up, the commutator segment CM returns to the commutator segment CM16. In other words, the commutator segments CM are arranged in a loop along the rotation direction θrt.

[0052] As described above, one end and the other end of the lead wire drawn from the winding WR are connected to the commutator segments CM adjacent in the circumferential direction. Specifically, as shown in FIG. 9, one end of the lead wire drawn from the winding WR1 is connected to the commutator segment CM16, and the other end is connected to the commutator segment CM1. One end of the lead wire drawn from the winding WR2 is connected to the commutator segment CM1, and the other end is connected to the commutator segment CM2. One end of the lead wire drawn from the winding WR3 is connected to the commutator segment CM2, and the other end is connected to the commutator segment CM3. One end of the lead wire drawn from the winding WR4 is connected to the commutator segment CM3, and the other end is connected to the commutator segment CM4. Similarly, the windings WR16 and WR17 are connected to the commutator segments CM. Note that FIG. 9 shows only the windings WR1 to WR4 as representative examples. As a result, the windings WR are connected in series via the commutator segments CM to form a loop circuit.

[0053] The brushes BR include positive brushes BR_P1 and BR_P2, and negative brushes BR_N1 and BR_N2. The negative brush BR_N1, the positive brush BR_P1, the negative brush BR_N2, and the positive brush BR_P2 are arranged in this order along the rotation direction θrt.

[0054] As the rotor 20B rotates, the commutator segments CM1 to CM16 move in the rotation direction θrt, and the commutator segments CM that contact the positive brushes BR_P1 and BR_P2 and the negative brushes BR_N1 and BR_N2 are switched in sequence. Fig. 9 shows, as an example, a state in which the commutator segments CM4 and CM3 are in contact with the negative brush BR_N1, the commutator segments CM16 and CM15 are in contact with the positive brush BR_P1, the commutator segments CM12 and CM11 are in contact with the negative brush BR_N2, and the commutator segments CM8 and CM7 are in contact with the positive brush BR_P2. The winding WR moves in the rotation direction θrt together with the commutator segments CM, thereby crossing the magnetic flux generated by the magnetic poles.

[0055] <About the winding circuit> The winding circuit section 2111B is a model that directly models the geometric arrangement of the windings WR, permanent magnets Mg, brushes BR, and commutator segments CM in the motor 20 (BDC motor). Fig. 10 shows model parameters in the motor physical model 2111. Details of the winding circuit section 2111B will be described below using the parameters in the winding circuit section 2111B shown in Fig. 10. Note that Fig. 10 also shows the above-mentioned parameters in the equation of motion section 2111A, and input variables and internal variables in the motor physical model 2111. Although the external torque Tex is treated as a parameter, it may be treated as an input variable.

[0056] The number of pole pairs p is the number of pairs of magnetic poles formed by the permanent magnets Mg. In the configuration of Figure 9 (Figure 8), there are two pairs of N and S poles, so the number of pole pairs p = 2. The total number of windings Ncoil is the total number of windings WR. In the configuration of Figure 9, the total number of windings Ncoil = 16.

[0057] The inductance L_1 per winding WR and the resistance R_1 per winding WR can be set by measuring, for example, two parallel circuits of eight windings in series at the midpoints of the loop circuit of the winding WR, which are 180 degrees symmetrical, in the configuration of FIG. 9, and calculating the value per winding as an average value.

[0058] In reality, there is a gap between adjacent commutator bars CM. As shown in Fig. 11, the gap between the commutator bars CM is highlighted, and the commutator bar gap GAP is set as the distance of the gap. The commutator bar gap GAP is set in units of the ripple angle θr [rad], which will be described later. The commutator bar gap GAP is set based on the observation results of an actual machine.

[0059] Brush resistance R_B is the contact resistance value when the brush BR and commutator segment CM come into contact so that their widths exactly overlap, as shown in the upper part of Figure 12. Figure 12 shows, as an example, a case in which the brush BR and commutator segment CM7 come into contact. In this case, the contact resistance between the commutator segment CM and the brush BR is at its smallest. Brush resistance R_B is adjusted so that the waveform (ripple waveform) of the motor terminal current im matches between the actual machine and the model.

[0060] When the contact width between the brush BR and commutator segment CM is very small, the ultimate contact resistance becomes infinite. The lower part of Figure 12 shows an example of a state in which the contact width between the brush BR and commutator segment CM7 is very small. However, it is difficult to perform calculations that include infinity in a simulation, and it is believed that such an extreme state does not actually occur in reality. Therefore, a saturation value Sat is set as the upper limit of the contact resistance to prevent division by zero. The saturation value Sat is adjusted so that the waveform (ripple waveform) of the motor terminal current im matches that of the actual device and the model.

[0061] 13 is a plan cross-sectional view that typically illustrates the movement of side WR1_H of the winding WR1 around the rotation axis J and the magnetic flux density B due to the permanent magnet MgN1 (with north pole on the radially inner side) when focusing on side WR1_H. Side WR1_H is a side of the winding WR1 that extends in the axial direction at the end of the winding WR1 on the opposite side of the rotation direction θrt. As shown in FIG. 13, the state in which side WR1_H is located on the line segment connecting the end of the permanent magnet MgN1 on the opposite side of the rotation direction θrt to the rotation axis J is defined as mechanical angle θm=0, and the magnetic flux density B due to the magnetic flux passing through side WR1_H is determined according to the mechanical angle θm.

[0062] Therefore, as shown in FIG. 14, the magnetic flux density distribution is set with the magnetic flux density B as a function B(θm) of the mechanical angle θm. At this time, the maximum magnetic flux density Bm is set as a parameter, and B=+Bm in the range where the N pole is arranged on the radially inner side, and B=-Bm in the range where the S pole is arranged on the radially inner side. Note that the polarity of the magnetic flux density B is positive in the direction toward the rotation axis J as shown in FIG. 13. In the range between the N pole and the S pole, the magnetic flux density B is set to change between +Bm and -Bm. In FIG. 14, B is set to change linearly between +Bm and -Bm, but the method of change is not limited to this. The magnetic flux density distribution can be set by the maximum magnetic flux density Bm and a formula block. Note that the relative positions of the windings WR2 to WR16 with respect to the winding WR1 are determined, and therefore the magnetic flux density distribution for the side corresponding to each side WR1_H is determined according to the mechanical angle θm.

[0063] The positional deviation Sgap (unit: [rad]) is a parameter that represents the positional relationship between the magnetic poles of the permanent magnet Mg and the brush BR. In the example shown in FIG. 15, the positional deviation Sgap is defined when the brush BR deviates from the reference state (similar to FIG. 14) in the positional relationship between the magnetic poles of the permanent magnet Mg and the brush BR shown at the top to the state shown at the bottom. The positional deviation Sgap is handled in the model so that the reference position (origin) of the magnetic flux density distribution can be changed according to the positional deviation Sgap. This is because even in the same mechanical angle θm state, i.e., in the same position state of the commutator segments CM relative to the brush BR, if a positional deviation Sgap occurs, the magnetic flux density relative to the winding WR will change. The positional deviation Sgap is set based on the observation results of an actual machine.

[0064] The winding side length l is a parameter that represents the length of a side of the winding WR that extends in the axial direction at the end of the winding in the direction of rotation θrt, as shown in Fig. 16. Note that the winding side length l of the winding WR1 corresponds to the length of the side WR1_H described above. The length l is set based on the observation results of the actual device.

[0065] The distance r from the axis of rotation to the side of the winding is a parameter that represents the radial distance from the axis of rotation J to the side WR_H. Figure 17 (similar to Figure 13) shows the distance r from the side WR1_H of the winding WR1 as an example. The distance r is set based on the observation results of the actual device.

[0066] FIG. 18 is a diagram showing the circuit configuration around the motor terminals (also shown in FIG. 9, etc.). A capacitor C0 is connected between the motor positive terminal Tp and the motor negative terminal Tn. An inductor L0 is connected to the motor positive terminal Tp and the motor negative terminal Tn, respectively. One end of one inductor L0 is connected to the motor positive terminal Tp, and the positive brushes BR_P1 and BR_P2 are connected to the other end. One end of the other inductor L0 is connected to the motor negative terminal Tn, and the negative brushes BR_N1 and BR_N2 are connected to the other end. An LC filter is made up of the capacitor C0 and the inductor L0. The LC filter is used to suppress pulse-like noise that occurs when the commutator segments CM and the brushes BR are switched. The capacitance value of the capacitor C0 and the inductance value of the inductor L0 are each set as a parameter.

[0067] Next, the induced voltage generated in the winding WR will be described. When each of the windings WR1 to WR16 crosses the magnetic flux generated by the permanent magnet Mg, an induced voltage is generated in each of the windings WR1 to WR16. As described above, the magnetic flux density distribution B is set as a function of the mechanical angle θm. Using this set magnetic flux density distribution, the induced voltage is calculated from the time change in the magnetic flux linking each of the windings WR1 to WR16.

[0068] 19 is a schematic plan view, a perspective view, and a development view showing a state in which one winding WR is moved by the rotation of the rotor 20B. It is assumed that the winding WR moves from a state represented by ABCD to a state represented by A' B' C' D'.

[0069] Referring to the development diagram shown in FIG. 20, the area DCC'D' scanned by the front side CD in the rotation direction θrt is the area that has increased due to the rotational movement of the winding WR, and is expressed as a positive area. On the other hand, the area ABB'A' scanned by the rear side BA in the rotation direction θrt is the area that has decreased due to the rotational movement of the winding WR, and is expressed as a negative area. Moreover, the area A'B'CD is the area that does not change before and after the rotational movement, and the magnetic flux that passes through this area also does not change, so it does not contribute to the magnetic flux change Δφ. From the above, the magnetic flux change Δφ is calculated as the product of the signed area and the magnetic flux density that passes through that area. Therefore, the induced electromotive force e coil,No. is expressed by the following equation. e coil,No. =Δφ / Δt=((magnetic flux density at the front edge)×(scan area of ​​the front edge)+(magnetic flux density at the rear edge)×(-(scan area of ​​the rear edge))) / Δt

[0070] In addition, e coil,No. is the induced voltage generated in the winding represented by No. (winding number) (for example, WR1 if No.=1).

[0071] Additionally, using the length of the winding side, l, and the distance of the winding side from the axis of rotation, r, the above positive area = l rωm, and the negative area = -l rωm.

[0072] FIG. 21 is a diagram showing an overall image of the winding circuit section 2111B. The induced voltage e coil,No. was modeled so that each of the windings WR1 to WR16 was output from a voltage source inserted in series with the inductance L_1 and the resistance R_1.

[0073] Next, a ripple angle θr shown in Fig. 21 will be described. If the brush BR of interest is referred to as a specified brush (positive brush BR_P1 in Fig. 21) and the commutator segment CM of interest is referred to as a specified commutator segment (commutator segment CM2 in Fig. 21), the contact resistance Rc between the specified brush and the specified commutator segment is found as follows.

[0074] If the commutator segment CM adjacent to the specified commutator segment on the side of the rotational direction θrt is called the front commutator segment (commutator segment CM1 in Figure 21), and the adjacent commutator segment CM on the opposite side of the rotational direction θrt of the specified commutator segment is called the rear commutator segment (commutator segment CM3 in Figure 21), then the state in which the rear end of the specified brush coincides with the rear end of the front commutator segment (the state in which the width of the specified brush and the width of the front commutator segment exactly coincide) is defined as the ripple angle θr = 0, and the distance from the rear end of the specified brush to the rear end of the front commutator segment is expressed as the ripple angle θr [rad]. When the rear end of the specified brush coincides with the rear end of the specified commutator segment (when the width of the specified brush and the width of the specified commutator segment exactly match), the ripple angle θr = 2π, and when the rear end of the specified brush coincides with the rear end of the rear commutator segment (when the width of the specified brush and the width of the rear commutator segment exactly match), the ripple angle θr = 4π. The ripple angle θr can be converted from the mechanical angle θm.

[0075] Based on the ripple angle θr, the contact resistance Rc between a given brush and a given commutator segment can be calculated as shown in the table in FIG. B corresponds to the brush resistance R_B. In addition, the contact resistance value is ∞ outside the range of the ripple angle θr shown in Fig. 22. However, as described above, when the calculated contact resistance Rc exceeds the saturation value Sat, the contact resistance Rc = Sat.

[0076] FIG. 23 is a diagram showing the configuration of a motor physical model 211 when modeling is performed using Simscape (registered trademark) / Simulink (registered trademark).

[0077] The equation of motion section 2111A receives the motor terminal current im output from the winding circuit section 2111B, and outputs the mechanical angle θm and the mechanical angle angular velocity ωm to feed back to the winding circuit section 2111B.

[0078] The winding circuit section 2111B has an induced voltage generating section 2A, a winding circuit model section 2B, a ripple angle conversion section 2C, a contact resistance value generating section 2D, and a switching signal generating section 2E.

[0079] The induced voltage generating unit 2A generates an induced voltage in each of the windings WR based on the mechanical angle θm. FIG. 24 shows a partial configuration of an example of the winding circuit model unit 2B. Here, considering the number of pole pairs p=2, modeling is performed for eight windings WR, which is half of the total of 16 windings WR in reality. Note that modeling may be performed for the entire number of windings WR in reality.

[0080] Therefore, as shown in FIG. 24, the winding WR is modeled by WR1 to WR8, and the brush BR is modeled by a pair of a positive brush and a negative brush. coil,No. is output from a voltage source E inserted in the winding WR of the corresponding winding number. In FIG. 24, the induced voltage e coil,8 is shown output from voltage source E8 at winding WR8.

[0081] As shown in Fig. 24, the winding circuit model unit 2B is modeled so that switches SW1, SW2 and variable resistors VR1, VR2 are provided for each winding WR. Specifically, in the winding WR, an inductor L_1 and a resistor R_1 are connected in series, and one end of the resistor R_1 (opposite the end connected to the inductor L_1) is connected to one end of the variable resistor VR1, and a switch SW1 is connected between the other end of the variable resistor VR1 and a positive line LP. The positive line LP is connected to a positive brush. In addition, one end of the resistor R_1 is also connected to one end of the variable resistor VR2, and a switch SW2 is connected between the other end of the variable resistor VR2 and a negative line LN. The negative line LN is connected to a negative brush.

[0082] When the commutator segment CM to which the lead wire of the winding WR is connected comes into contact with the positive brush, the switch SW1 is turned on and the resistance value of the variable resistor VR1 is set to the contact resistance value. When the commutator segment CM to which the lead wire of the winding WR is connected comes into contact with the negative brush, the switch SW2 is turned on and the resistance value of the variable resistor VR2 is set to the contact resistance value. When the commutator segment CM is not in contact with the positive or negative brush, the switch SW1 or SW2 is turned off. There are cases when both switches SW1 and SW2 are turned off.

[0083] A switching signal generating unit 2E shown in FIG. 23 determines whether to turn on or off the switches SW1 and SW2 for each winding WR based on the mechanical angle θm, and generates a switching signal. The switches SW1 and SW2 are turned on or off based on the generated switching signal. A ripple angle conversion unit 2C converts the mechanical angle θm into a ripple angle θr. A contact resistance value generating unit 2D generates a contact resistance value between the commutator segments CM and the brush BR based on the ripple angle θr. The generated contact resistance value is set as the resistance value of the variable resistors VR1 and VR2.

[0084] The winding circuit model unit 2B calculates and outputs the motor terminal current im when an input voltage Vin is input, with the induced voltage by the voltage source E, the on / off states of the switches SW1 and SW2, and the resistance values ​​of the variable resistors VR1 and VR2 being determined. Note that when modeling with eight windings WR taking into account the number of pole pairs = 2 as described above, the motor terminal current im is reduced by half, so the calculated motor terminal current im is input to an amplifier with a gain = 2 and output to the equation of motion unit 2111A.

[0085] <Simulation results> FIG. 25 is a diagram showing an example of the results when a simulation according to the present disclosure is performed. In response to application of the input voltage Vin, a ripple waveform occurs in the motor terminal current im, similar to that of an actual machine. FIG. 26 is a diagram showing an example of a comparison between the relationship between the input voltage Vin and the mechanical angle rotation speed in the actual machine and the simulation. In this way, the simulation reproduces a phenomenon similar to that of an actual machine, in which the mechanical angle rotation speed increases with an increase in the input voltage Vin. The simulation according to the present disclosure can significantly reduce the amount of calculation and shorten the simulation time.

[0086] <Abnormal state model> As shown in FIG. 4, the motor model 211 has an abnormal state model 211A. The abnormal state model 211A has a bearing lubrication failure model 2112 and a bearing damage model 2113. The reason for modeling an abnormality in the bearing 20E as an abnormal state is as follows. The bearing 20E is a mechanical element that plays a role of supporting the shaft 20C, and is independent of other mechanical elements inside the motor and the load. In other words, the abnormal state of the bearing 20E can be handled independently of other abnormal states. Furthermore, the bearing E is an element that is provided regardless of the type of motor and the type of load, and the abnormal state model of the bearing can be used regardless of the type of motor and the type of load.

[0087] Bearing abnormalities can be roughly divided into two types: poor lubrication and damage, so a poor bearing lubrication model 2112 and a damaged bearing model 2113 are modeled. Bearing 20E is mechanically connected to shaft 20C via friction and normal force. Therefore, an abnormal state model 211A is modeled as a model that outputs friction torque as the difference (deviation amount) between the normal state and the abnormal state, and normal force caused by the abnormality.

[0088] <Bearing lubrication failure model> FIG. 29 shows a front view and a side cross-sectional view of a bearing 20E and a shaft 20C. The bearing 20E has an outer ring OR and an inner ring IR. The shaft 20C is fixed to the inside of the inner ring IR. A lubricant LUB is provided between the outer ring OR and the inner ring IR. This allows the inner ring IR and the shaft 20C to rotate relative to the stationary outer ring OR. Note that FIG. 29 shows an example in which the bearing is a rolling bearing, but the bearing may also be a sliding bearing.

[0089] The lubrication failure of the bearing 20E depends on whether the lubricant failure is global or local, and on the degree of the lubricant viscosity and fluid contribution. Based on the above, the lubrication failure can be expressed by a formula, which is a superposition of a mode that depends on the mechanical angle and rotation speed of the rotating shaft and a mode that depends only on the rotation speed. The mode that depends on the mechanical angle and rotation speed of the rotating shaft is a mode related to mechanical variations or fluctuations within one rotation of the shaft. For example, it relates to the surface condition, dimensional variations, deviation from a perfect circle, fit, foreign matter adhesion locations, damaged locations, radial load, etc. The mode that depends only on the rotation speed is related to the dynamic viscosity when the fluid (air, lubricating oil, grease, etc.) slides against the surface. As a result, the friction torque caused by the lubrication failure (deviation from the normal state) is expressed by the following formula (3).

number

[0090] As shown in FIG. 27, the mechanical angle θm and the mechanical angle angular velocity ωm output from the motion equation unit 2111A are input to the above equation (3), and the friction torque T lubrication is calculated. The calculated friction torque T lubrication is input to the equation of motion part 2111A as the torque loss component ΔT due to the abnormal state in the above equation (1). The progress of the abnormal state (i.e., deterioration) is expressed by increasing or decreasing the parameter in the above equation (3) as a function of time.

[0091] In addition, in the bearing lubrication failure model 2112, the normal force caused by the abnormality is calculated by the following formula (4).

number

[0092] That is, the normal force is calculated using the friction torque calculated by the above formula (3). The normal force calculated in this way is input as an excitation force to a vibration model of the support system, which will be described later.

[0093] <Bearing damage model> As shown in Fig. 29, in a bearing 20E, a rolling element RE is provided between an outer ring OR and an inner ring IR. A plurality of rolling elements RE are provided in the circumferential direction. Note that, for convenience, in Fig. 29, the lubricant LUB and the rolling elements RE are illustrated separately.

[0094] Bearing damage is classified into modes depending on which mechanical element of the bearing the damage occurred in. Specifically, there are three modes: outer ring damage, inner ring damage, and rolling element damage.

[0095] The rolling element RE moves in the circumferential direction by revolution while rotating on its axis. In the case of damage to the outer ring, an impact force (hereinafter, an impact pulse) is generated every time the rolling element RE slides against the damaged area. The impact pulse acts on the shaft 20C as a normal force and a friction torque. As a result, in the case of damage to the outer ring, the revolution angle θ revolution Each time equation (A) below is satisfied, a normal force expressed by equation (5) below is generated.

number

number

[0096] In addition, in the case of inner ring damage, a shock pulse is generated every time the rolling element RE slides against the damaged area. revolution Each time the following equation (B) is satisfied, a normal force expressed by the above equation (5) is generated.

number

[0097] In addition, in the case of damage to the rolling element, an impact pulse is generated every time the outer or inner ring slides against the damaged part due to the rotation of the rolling element RE. rotation Each time equation (C) below is satisfied, a normal force expressed by equation (5) above is generated.

number

[0098] The normal force N caused by the above damage bearing_damage Based on this, the friction torque caused by the damage is expressed as the following equation (6).

number

[0099] Revolution angle θ revolution and rotation angle θ rotation is calculated based on the mechanical angle θm as follows:

number

[0100] Here, the derivation of the above formula will be described with reference to FIG. inner (=θm) revolution angle θ of rolling element RE revolution and rotation angle θ rotation Points Ar and Br indicate the contact points between the outer ring OR and the inner ring IR on the rolling element RE side. Points Ar' and Br' indicate the points to which points Ar and Br have moved due to the revolution of the rolling element RE. At that time, the contact points on the outer ring OR side and the inner ring IR side change from Ao to Ao'', Bi to Bi'', and contact point Bi has moved to Bi'.

[0101] Since the length of Ar'Ao'' = the length of AoAo'', the following equation (D) holds. r r θ rotation =r o θ revolution (D)

[0102] On the other hand, since the length of Br'Bi''=the length of Bi'Bi'', the following formula (E) holds. r r θ rotation =r i (θm-θ revolution ) (E)

[0103] From the above formulas (D) and (E), r o θ revolution =r i θm-r i θ revolution And then, θ revolution =(r i / (r o +r i ))θm. Where: r i =(D pitch -D rolling· cosα) / 2 r o =(D pitch +D rolling· cosα) / 2, so θ revolution =1 / 2·(1-D rolling / D pitch· cosα) θm.

[0104] On the other hand, from the above formula (D), θ rotation =r o / r r· θ revolution It becomes. Here, the above θ revolution =1 / 2·(1-D rolling / D pitch· cosα) θm and r r =D rolling From / 2, θ rotation =1 / 2 D pitch / D rolling· (1-(D rolling / D pitch· cosα) 2 ) θm.

[0105] As shown in Fig. 28, the mechanical angle θm output from the equation of motion part 2111A is input to the bearing damage model 2113. The friction torque T bearing_damageis input to the equation of motion part 2111A as the torque loss component ΔT due to the abnormal state in the above equation (1). The progress of the abnormal state (i.e., deterioration) increases with the height P hight and width P width It is expressed by increasing or decreasing as a function of time. bearing_damage is input to the vibration model of the support system described later.

[0106] <Relationship between abnormal state model and motor physical model> As shown in Fig. 4, signals are exchanged between the motor physical model 2111 (motion equation part) and the bearing poor lubrication model 2112, and between the motor physical model 2111 and the bearing damage model 2113. However, there is no signal exchange between the bearing poor lubrication model 2112 and the bearing damage model 2113. This is against the backdrop of a transition from a normal state to an abnormal state, based on various findings such as statistics, whereby a transition to one abnormal state is first made, and then as the abnormality progresses, a transition to a composite abnormal state occurs. Since this is a model for studying early detection of abnormal symptoms, it is sufficient to be able to express the initial stage of a transition to one abnormal state.

[0107] <Vibration model of support system> As shown in FIG. 4, the motor model 211 has a vibration model 2114 of a support system. First, the support system is defined as all the mechanical elements constituting the motor 20, including the housing 20A. The base 201 (FIG. 3) is not the support system and is stationary. The coordinate system for handling the vibration of the support system is a stationary coordinate system as shown in FIG. 3. The coordinate axes in the stationary coordinate system are as described above. The reason for taking the coordinate axes in this way is that the rigidity of the motor 20 is generally anisotropic in the vertical and horizontal directions of the installation surface.

[0108] Vibration of the support system occurs when a force (exciting force) caused by an abnormality is applied to the support system. In the case of poor lubrication and damage of the bearing 20E described above, the normal force caused by the abnormality becomes the exciting force. In the case of a bearing abnormality, the exciting force is a vector in the XY plane.

[0109] The vibration of the support system is determined by the combined stiffness (spring constant) and damping of the individual mechanical elements that make up the support system. The vibration model of the support system is assumed to be a system consisting of springs Kx, Ky, dampers Cx, Cy, and a mass point of mass M, as shown in Figure 32. The entire motor 20 including the housing 20A and all of the loads connected to the shaft 20C are considered to be a mass point of mass M. On both sides of the above mass point in the X direction, the spring Kx and the damper Cx are connected in parallel. On both sides of the above mass point in the Y direction, the spring Ky and the damper Cy are connected in parallel.

[0110] The translational equation of motion in such a support system is expressed by the following equation (7). Note that the translational equation of motion must be established for each vibration measurement point (position of the vibration sensor) by the vibration sensor. This is because the parameters M, k, and c may change depending on the vibration measurement point.

number

[0111] Since it is not practical to combine the stiffness and damping of each mechanical element that constitutes the support system, the stiffness and damping of the entire support system are estimated, and each estimated value can be set as a representative value to set the parameters. Note that equations of motion can be established for the X-axis, Y-axis, and Z-axis separately. For example, when measuring vibration on an axis tilted at 45 degrees on the XY plane, the X-axis vibration and the Y-axis vibration can be combined.

[0112] As mentioned above, the normal force N output from the bearing lubrication failure model 2112 lubrication (Equation (4) above), or the normal force N obtained from the bearing damage model 2113 bearing_damage(Equation (5) above) is input to vibration model 2114 of the support system as an external force F. Time series data of X-direction displacement and Y-direction displacement is output according to the equation of motion (Equation (7) above) in vibration model 2114 of the support system. Note that velocity data can be obtained by first-order differentiation of the displacement data, and acceleration data can be obtained by further first-order differentiation.

[0113] There is no signal input from the vibration model 2114 of the support system to the abnormal state model 211A (FIG. 4). This is because vibration is treated as a result of the application of force.

[0114] <Motor model calculation> When the simulation is executed, the model calculation unit 3 performs calculation processing of the motor model 211. At this time, while signals are exchanged between the equation of motion unit 2111A and the abnormal state model 211A, numerical calculations are performed by the motor physical model 2111 (the equation of motion unit 2111A and the winding circuit unit 2111B), and time series data of the motor terminal current im is output from the winding circuit unit 2111B. Meanwhile, in the vibration model 2114 of the support system, numerical calculations are performed while receiving an input from the abnormal state model 211A, and time series data of the X-direction displacement, Y-direction displacement, X-direction acceleration, and Y-direction acceleration are output.

[0115] In this way, the motor terminal current im, and the time series data of the X-direction displacement, the Y-direction displacement, the X-direction acceleration, and the Y-direction acceleration are output as physical signal waveform data from the motor model 211. By setting the abnormal state model 211A to an abnormal state, it is possible to virtually generate a physical signal waveform in an abnormal state.

[0116] Even if signal waveform data of an abnormal motor state is generated, the simulation speed is very slow in a model based on the finite element method. Also, in a model in which frequency component noise is added based on experimental data, contradictions occur in the relationship between each physical signal. Therefore, the motor model 211 according to this embodiment can generate physical signal waveform data of an abnormal state with the required accuracy and at a moderate simulation speed. At this time, no contradictions occur between each physical signal.

[0117] In addition, by setting the abnormal state model 211A to the normal state, it is possible to virtually generate a physical signal waveform in the normal state. In this case, in the bearing lubrication failure model 2112, the coefficient B lub_cof1 and B. lub_cof2 In the bearing damage model 2113, the height of the shock pulse P in the above equation (5) is set to 0. hight Let be 0.

[0118] <Sensor model> Next, the sensor model 22 will be described. Fig. 33 is a diagram showing a configuration example of the sensor model 22. The sensor model 22 has a transfer function model 221 and an AD conversion model 222. The sensor model 22 has the transfer function model 221 and the AD conversion model 222 for each type of sensor. The types of sensors are, for example, a current sensor and a vibration sensor. This makes it possible to select the type of sensor, as will be described later.

[0119] The transfer function model 221 receives an input signal Sin1 and outputs a sense signal SS as an analog signal. The sense signal SS is AD converted by an AD conversion model 222 and is output as an output signal (sense signal) Sout1 as a digital signal. The transfer function model 221 includes a filter. As will be described later, the type of the filter and the characteristics of the filter (such as a cutoff frequency) can be set. In addition, the characteristics of the AD conversion model 222 (such as a sampling rate) can also be set.

[0120] The physical signal waveform data (or data based on the physical signal waveform data) output from the motor model 211 is input as an input signal Sin1 to the sensor model 22. If the physical signal waveform data is a motor terminal current im, it is input to the sensor model 22 as a current sensor. If the physical signal waveform data (or data based on the physical signal waveform data) is an X-direction displacement, a Y-direction displacement, an X-direction acceleration, and a Y-direction acceleration, it is input to the sensor model 22 as a vibration sensor.

[0121] 33, noise Ns1, Ns2, and Ns3 can be input before the transfer function model 221, between the transfer function model 221 and the AD conversion model 222, and after the AD conversion model, respectively. The type of noise can be specified, such as normal distribution noise or uniform distribution noise, and parameters such as the average value or variance (standard deviation) can be set according to the type.

[0122] <Machine learning model> Next, a description will be given of the machine learning model 23. Fig. 34 is a diagram showing an example of the configuration of the machine learning model 23. The machine learning model 23 has a preprocessing unit 231 and a machine learning unit 232.

[0123] The preprocessing unit 231 performs preprocessing on the input signal Sin2. The preprocessing includes envelope processing, window function processing, and FFT (fast Fourier transform) processing. As described later, it is possible to select whether or not to perform envelope processing, window function processing, and FFT processing. As patterns, it is possible to select the execution of only window function processing, only envelope processing, only FFT processing, window function processing and FFT processing, envelope processing and FFT processing, or envelope processing, window function processing, and FFT processing. Note that the present invention is not limited to FFT processing, and frequency analysis processing such as wavelet transform may also be used.

[0124] Furthermore, in the preprocessing unit 231, normalization processing is also performed for appropriate learning in the machine learning unit 232. The normalization processing is a process in which data is multiplied by a normalization coefficient to place the data within a range of approximately 0 to 1 (or -1 to +1). If the input data has a value outside a predetermined range, learning is not performed or learning is performed assuming the data to be a saturated value, so normalization processing is required as preprocessing in order to learn all data.

[0125] When "with preprocessing" is selected, at least one of window function processing and FFT processing is performed on the input signal Sin2, and then normalization processing is performed to obtain input data Din as an input to the machine learning unit 232. When "without preprocessing" is selected, normalization processing is performed on the input signal Sin2 to obtain input data Din.

[0126] The machine learning unit 232 performs learning and inference on the input data Din.

[0127] As an AI model used in the machine learning unit 232, for example, a three-layer neural network 30 as shown in FIG.

[0128] As shown in Fig. 35, the three-layer neural network 30 is an AI model having an input layer 30A, a hidden layer 30B, and an output layer 30C. In general, in the three-layer neural network 30, n-dimensional input data x ∈ R k×n For n'-dimensional inference result y∈R k×n’ is given as y = G(x α + b)β, where α∈R n×m are the weights connecting the input layer 30A and the hidden layer 30B, and β∈R m×n’ is the weight connecting the hidden layer 30B and the output layer 30C. Also, b∈R m is the bias of hidden layer 30B, and G is the activation function of hidden layer 30B.

[0129] In this embodiment, an algorithm is used that can sequentially train the three-layer neural network 30 with any batch size.i The i-th training data {x i ∈R ki×n , t i ∈R ki×n’} is obtained, β that minimizes the error shown in the following equation (8) i It is necessary to ask for.

number

[0130] Optimized weight β i is calculated using the following equation (9). P i =P i-1 -P i-1 H i T (I+H i P i-1 H i T ) -1 H i P i-1 β i =β i-1 +P i H i T (t i -H i β i-1 ) (9)

[0131] Here, P0 and β0 are obtained by the following equation (10). P0=(H0 T H0) -1 β0=P0H0 T t0(10)

[0132] The learning algorithm is as follows: (1) The values ​​of the weight α and bias b are initialized by random numbers. (2) Calculate H0 for x0, and calculate P0 and β0. (3) Batch size ki Each time the i-th training data of P is obtained, i and β i It should be noted that instead of using the formula for calculating β0 in equation (10), a value initialized by a random number may be used as β0.

[0133] In this embodiment, learning is performed using an autoencoder. The autoencoder uses input data as it is as training data, and learns so that the input data can be reconstructed as an inference result. That is, in the above example, learning is performed as t=x. Since the autoencoder does not need to create training data separately, it is a type of unsupervised learning algorithm.

[0134] According to such an AI model in the machine learning unit 232, learning is possible with a calculation device of a microcomputer level in the edge device. In particular, the bottleneck in terms of the amount of calculation in the above formula (9) is (I + H i P i-1 H i T ) -1 However, (I+H i P i-1 H i T Since the matrix size of x is k×k, when k=1, the inverse matrix operation can be replaced with the reciprocal operation. Therefore, by fixing the batch size to k=1, the operation becomes easy even for a calculation device such as a microcomputer. In other words, when such on-device learning is introduced into motor anomaly detection, the effect of anomaly detection can be confirmed by simulation. Note that the input data x is time-series data when no FFT processing is performed in the pre-processing unit 231, and is frequency domain data when FFT processing is performed.

[0135] In the machine learning unit 232, the degree of anomaly is calculated by a loss function L(y,t) that represents the error between the inference result y and the training data t. For example, the mean absolute error (MAE) or the mean squared error (MSE) is used as the loss function. When the loss function is the MAE, the loss function L is expressed as in the following formula (11).

number

number

[0136] Since learning is performed using an autoencoder, the error is calculated using a loss function L(y,t)=L(y,x), and the calculated error is taken as the degree of anomaly. The calculated degree of anomaly is output from the machine learning unit 232 as the degree of anomaly data Dab.

[0137] In this embodiment, a forgetting rate can be set. The forgetting rate is a parameter that indicates the degree to which the learning result is forgotten. Examples of methods for not reflecting the learning result include using past learning results and initializing the learning result.

[0138] <Abnormality Judgment Model> Next, the anomaly determination model 24 will be described. Anomaly determination is performed based on the anomaly level data Dab output from the machine learning model 23. The anomaly determination model 24, for example, compares the anomaly level with one threshold value and determines whether the anomaly level is abnormal or normal. The anomaly determination model 24 may also compare the anomaly level with multiple threshold values ​​and determine the anomaly level in stages. As will be described later, such an anomaly determination method can be selected. In addition, the anomaly determination model 24 may perform processing such as integration or averaging of the anomaly level before comparing with the threshold value.

[0139] <Simulation setting GUI> Next, a description will be given of a GUI (Graphical User Interface) that enables the setting of simulation conditions in the simulation device 1 according to this embodiment. Examples of various setting screens described below are displayed on the display unit 7 by the display control unit 5 (FIG. 2). Selection and setting on the setting screen, or switching of the screen, is performed based on input from the operation input unit 6. The contents to be set on the various setting screens are set by the model setting unit 4.

[0140] FIG. 36 is a diagram showing a setting screen for a motor type and an abnormal state (hereinafter, the first setting screen). On the first setting screen, tabs TB are displayed side by side at the top. The setting screen can be switched by switching the tabs TB. Also, on the first setting screen, a motor type selection section SG1 and an abnormal state selection section SG2 are displayed.

[0141] In the motor type selection section SG1, motors as selection candidates are displayed, and the motor can be selected. In Fig. 36, a BDC motor and a BLDC motor are displayed as examples of motor types, and the BDC motor is selected as an example.

[0142] In the abnormal condition selection section SG2, the abnormal condition of the motor can be selected. Specifically, poor bearing lubrication, damage to the outer ring of the bearing, damage to the inner ring of the bearing, and damage to the rolling elements of the bearing are displayed as selection candidates, and any one of the abnormal conditions can be selected. In FIG. 36, damage to the inner ring is selected as an example.

[0143] Furthermore, in the first setting screen, a lamp LP0 is displayed which indicates a mechanical element corresponding to a selectable abnormal state. The lamp LP0 is displayed in a display showing the overall configuration of the motor, and can be turned on and off. The lamp LP0 of the mechanical element corresponding to the selected abnormal state is turned on. In FIG. 36, inner race damage is selected as an example, so the lamp LP0 of the bearing, which is the corresponding mechanical element, is turned on. Such a lamp LP0 makes it easier to check the mechanical element corresponding to the abnormal state of the simulation target.

[0144] Furthermore, in the first setting screen, lamps LP1 to LP4 are displayed to indicate the location of occurrence of a selectable abnormal state. In FIG. 36, as an example, a lamp LP1 indicating the lubricant of the bearing, a lamp LP2 indicating the outer ring, a lamp LP3 indicating the inner ring, and a lamp LP4 indicating the rolling element are displayed. LP1 corresponds to poor bearing lubrication, LP2 corresponds to outer ring damage, LP3 corresponds to inner ring damage, and LP4 corresponds to rolling element damage. LP1 to LP4 can be turned on and off. The lamps LP1 to LP4 corresponding to the selected abnormal state are turned on. In the example of FIG. 36, the inner ring damage is selected, so the lamp LP3 is turned on. Such lamps LP1 to LP4 make it easier to confirm the location of occurrence of the abnormal state of the simulation target.

[0145] Also, at the bottom of the first setting screen, a simulation time setting section ST, a simulation start button SB, and an abnormal state lamp FLP are displayed. These displays are displayed in common even when the setting screen is switched. In the simulation time setting section ST, the simulation time can be set. When the time of the value inputted into the simulation time setting section ST has elapsed from the start of the simulation, the simulation is stopped. In addition, if a time is set such that the simulation is stopped before the inference start timing described later, the simulation is stopped before the inference starts.

[0146] The simulation start button SB can be pressed to start the simulation. The abnormal condition lamp FLP is a lamp that corresponds to a selectable abnormal condition, and can be turned on and off. In FIG. 36, as an example, the abnormal condition lamps FLP are displayed side by side for bearing lubrication failure, outer ring damage, inner ring damage, and rolling element damage. In the example of FIG. 36, inner ring damage has been selected, so the abnormal condition lamp FLP for inner ring damage is turned on.

[0147] When switching from the first setting screen to the adjacent tab TB, a basic motor setting screen is displayed as shown in Fig. 37. In the basic motor setting screen, a basic parameter setting section SG3 and a drive setting section SG4 are displayed.

[0148] In the basic parameter setting section SG3, the basic parameters of the motor can be set for the motor type selected on the first setting screen. In Fig. 37, as an example, the winding resistance (R_Wire), winding inductance (L_Wire), and inertia (Jm) can be set.

[0149] In the drive setting section SG4, the drive method and drive voltage can be set for the motor type selected on the first setting screen. The drive method can be set to, for example, constant voltage drive, PWM drive, three-phase drive (rectangular wave drive, sine wave drive), etc. The drive voltage can be set to, for example, the voltage applied to the motor terminal in the case of constant voltage drive, or the high-level voltage in the case of applying a voltage to the motor terminal by switching between high and low levels.

[0150] In addition, the items that can be set in basic parameter setting section SG3 and drive setting section SG4 may change depending on the selected motor type.

[0151] When switching from the basic motor setting screen to the adjacent tab TB, the support system setting screen as shown in Fig. 38 is displayed. In the support system setting screen, a stiffness / damping setting section SG5 is displayed. In the stiffness / damping setting section SG5, the spring constants (stiffness) kx, ky and the damping coefficients cx, cy in the vibration model 2114 of the support system can be set.

[0152] When switching from the support system setting screen to the adjacent tab TB, a first abnormal state setting screen is displayed as shown in FIG. 39. In the first abnormal state setting screen, an abnormality parameter setting section SG6 is displayed. In the abnormality parameter setting section SG6, lub_cof1 ,B lub_index1 ,B lub_cof2 ,SW_B lub_cof2 ,B lub_index2,μ lub can be set respectively.

[0153] When switching from the first abnormality setting screen to the adjacent tab TB, a second abnormality condition setting screen is displayed as shown in Fig. 40. In the second abnormality condition setting screen, a poor lubrication position setting section SG7 and a bearing damage position setting section SG8 are displayed.

[0154] In poor lubrication position setting unit SG7, the position where poor lubrication occurs in the bearing can be set by angle. In bearing damage position setting unit SG8, ​​the position where outer ring damage occurs can be set by angle. The angular positions set in poor lubrication position setting unit SG7 and bearing damage position setting unit SG8 correspond to the angular position θf indicating the direction in which external force F is applied as a normal force caused by an abnormality in vibration model 2114 of the support system (FIG. 32).

[0155] When switching from the second abnormal condition setting screen to the adjacent tab TB, a time / deterioration setting screen is displayed as shown in Fig. 41. The time / deterioration setting screen displays a time setting section SG9, a lubrication failure deterioration setting section SG10, and a bearing damage deterioration setting section SG11. The time / deterioration setting screen also displays an explanation display ED for explanatory purposes.

[0156] In the time setting section SG9, the learning start time, inference start time, and deterioration step time can be set. The learning time and inference start time are expressed as the elapsed time from the start of the simulation. The learning start time and inference start time are displayed in the explanation display ED. Depending on the setting of the learning start time, it is also possible not to use the data at the time of motor start for learning.

[0157] As shown in the explanation display ED, the progression of deterioration is expressed by gain. A state where gain = 0 is the normal state, and deterioration progresses stepwise from the normal state to the first deterioration state Gain 0, the second deterioration state Gain 1, and the third deterioration state Gain 2. The deterioration step time is the time during which the normal state and each deterioration state are maintained to their maximum. If the simulation ends before the deterioration step time is completed, the normal state or any of the deterioration states will be maintained for a time shorter than the deterioration step time.

[0158] In the lubrication failure / deterioration setting section SG10, the constant coefficient B in the above formula (3) is used to calculate the friction torque caused by poor bearing lubrication. lub_cof1 and B. lub_cof2 A gain value can be set for each of the first to third deterioration states. For example, a gain value of 0 indicates the value of the coefficient in a normal state, a gain value of 1 indicates the value of the coefficient in a predefined abnormal state (set by the abnormality parameter setting unit SG6), a gain value of 0.5 indicates 0.5 times the value of the coefficient in the predefined abnormal state, a gain value of 2 indicates 2 times the value of the coefficient in the predefined abnormal state, and a gain value of 4 indicates 4 times the value of the coefficient in the predefined abnormal state. The larger the gain value, the larger the coefficient, indicating that deterioration due to poor lubrication is progressing.

[0159] The bearing damage / deterioration setting unit SG11 can set gain values ​​for the height Phight and width Pwidth of the shock pulse caused by bearing damage (damage to the outer ring, inner ring, or rolling element) in each of the first to third deterioration states. Specific examples of the gain values ​​are the same as those in the case of the coefficients described above. The higher the gain value, the larger the height and width of the shock pulse, indicating that deterioration is progressing.

[0160] When switching from the time / deterioration setting screen to the adjacent tab TB, a sensor setting screen is displayed as shown in Fig. 42. The sensor setting screen displays a sensor type setting section SG12, a filter setting section SG13, a sensor characteristic setting section SG14, a sampling speed setting section SG15, and a noise setting section SG16.

[0161] In the sensor type setting unit SG12, the sensor type in the sensor model 22 can be selected. In FIG. 42, as an example, at least one of a current sensor, a first vibration sensor, and a second vibration sensor can be selected. A transfer function model 221 and an AD conversion model 222 shown in FIG. 33 are provided for each sensor type. The current sensor receives the motor terminal current as a physical signal waveform from the motor model 211 and outputs a current sense signal. The first vibration sensor and the second vibration sensor receive the displacement and acceleration as physical signal waveforms from the motor model 211, respectively, and output a displacement sense signal and an acceleration sense signal, respectively.

[0162] In the filter setting unit SG13, it is possible to set the filter (included in the transfer function model 221) in the sensor model 22. The filter is added taking into consideration noise removal or the frequency range and Nyquist frequency of the sensor. The filter can be set for each sensor type. For example, it is possible to set the type of filter (low pass filter (LPF), band pass filter (BPF), etc.) and parameters related to the filter (cutoff frequency, etc.).

[0163] In the sensor characteristic setting unit SG14, the characteristics of the sensor itself can be set for each sensor type. For example, the frequency characteristics of the sensor itself can be expressed by a transfer function (LPF, BPF, etc.). G(s) shown in FIG. 42 represents the transfer function, and s is the Laplace operator.

[0164] The sampling rate setting unit SG15 can set the sampling rate in the AD conversion model 222. The sampling rate can be set for each sensor type.

[0165] In the noise setting section SG16, the type and parameters of each of the noises Ns1, Ns2, and Ns3 can be set.

[0166] In addition, a sensor position setting section SG17 is also displayed on the sensor setting screen. In the sensor position setting section SG17, the position of the sensor can be set. The sensor position can be set by selecting a sensor with the operation input section 6 and moving it on the screen. In FIG. 42, as an example, a first vibration sensor VS1 and a second vibration sensor VS2 are displayed, and the positions of the first vibration sensor VS1 and the second vibration sensor VS2 can be set. In FIG. 42, the first vibration sensor VS1 is placed on the X-axis, and the second vibration sensor VS2 is placed on the Y-axis.

[0167] When switching from the sensor setting screen to the adjacent tab TB, an AI setting screen is displayed as shown in Fig. 43. In the AI ​​setting screen, a normalization coefficient setting section SG18 and a machine learning setting section SG19 are displayed.

[0168] A preprocessing unit 231 and a machine learning unit 232 in the machine learning model 23 are provided for each sense signal output from the sensor model 22. As an example, a preprocessing unit 231 and a machine learning unit 232 are provided for each of the current sense signal output from the current sensor, the displacement sense signal and the acceleration sense signal output from the first vibration sensor, and the displacement sense signal and the acceleration sense signal output from the second vibration sensor (five sense signals).

[0169] The normalization coefficient setting unit SG18 can set a normalization coefficient for normalization for each sense signal in the pre-processing unit 231. In Fig. 43, as an example, a normalization coefficient for each of the above five sense signals can be set.

[0170] In the machine learning setting unit SG19, various items related to the machine learning unit 232 can be set. In Fig. 43, as an example, for the three-layer neural network 30 (Fig. 35), the number of input data n, the number of hidden layer nodes m, the activation function, the loss function, and the forgetting rate can be set. For example, the activation function can be Sigmoid or ReLU. For example, the loss function can be MAE or MSE.

[0171] The AI ​​setting screen also displays a pre-processing setting section SG20. In the pre-processing setting section SG20, it is possible to set whether or not to perform envelope processing, FFT processing (frequency analysis processing), and window function processing.

[0172] When switching from the AI ​​setting screen to the adjacent tab TB, an abnormality determination setting screen is displayed as shown in Fig. 44. In the abnormality determination setting screen, an abnormality determination method setting section SG21 and an abnormality determination threshold setting section SG22 are displayed.

[0173] In the abnormality determination method setting section SG21, it is possible to select an abnormality determination method based on the degree of abnormality in the abnormality determination model 24. In Fig. 44, as an example, a first determination method (Method1) and a second determination method (Method2) can be selected. The first determination method is a method in which the degree of abnormality is compared with one threshold value and abnormality or normality is determined. The second determination method is a method in which the degree of abnormality is compared with multiple threshold values ​​and the abnormality level is determined in stages.

[0174] In the abnormality determination threshold setting section SG22, a threshold used in the selected abnormality determination method can be set.

[0175] When all the setting items have been set using the above setting screens and the simulation start button SB is pressed by the operation input unit 6, the model calculation unit 3 executes a simulation according to the contents set by the model setting unit 4. After the simulation is completed, by pressing the tab TB of the simulation result screen as shown in Fig. 45 by the operation input unit 6, the display control unit 5 displays the simulation results on the display unit 7.

[0176] In the simulation result screen, time series data of the sense signal by the sensor selected above and time series data of the degree of abnormality corresponding to the sense signal can be displayed. In Fig. 45, as an example, time series data of the motor terminal current, X-direction displacement, Y-direction displacement, X-direction acceleration, and Y-direction acceleration, and time series data of the degree of abnormality of the motor terminal current, X-direction displacement, Y-direction displacement, X-direction acceleration, and Y-direction acceleration are displayed. Note that, since the first vibration sensor VS1 is arranged on the X-axis and the second vibration sensor VS2 is arranged on the Y-axis as shown in Fig. 42, the X-direction displacement and X-direction acceleration are detected by the first vibration sensor VS1, and the Y-direction displacement and Y-direction acceleration are detected by the second vibration sensor.

[0177] It is also possible to display time-series data of the abnormality determination result in the abnormality determination model 24, or the FFT processing result (power spectrum) in the preprocessing unit 231 (FIG. 34).

[0178] In this way, the GUI makes it possible to intuitively set simulation conditions, and by running a simulation and checking the simulation results, it is possible to confirm the effectiveness of machine learning in detecting signs of abnormalities in a motor.

[0179] <Other> In addition to the above-mentioned embodiment, various technical features disclosed in this specification can be modified in various ways without departing from the spirit of the technical creation. In other words, the above-mentioned embodiment should be considered to be illustrative and not restrictive in all respects, and the technical scope of the present disclosure should not be limited to the above-mentioned embodiment, but should be understood to include all modifications that fall within the meaning and scope equivalent to the scope of the claims.

[0180] <Additional Notes> As described above, the simulation device (1) according to one embodiment of the present disclosure has: A motor physical model (2111) modeled by a winding circuit section (2111B) and a rotational motion equation section (2111A), An abnormal condition model (211A) that models the abnormal condition of the motor, A model storage unit (2) in which the above is stored; A model calculation unit (3) configured to perform calculation processing using the motor physical model; Equipped with An abnormality parameter that indicates the amount of deviation from a normal state calculated by the abnormal state model is configured to be input to the motor physical model (first configuration).

[0181] In the first configuration, the abnormal state model includes two or more motor abnormality models (2112, 2113), The two or more motor abnormality models may be configured so that there is no signal exchange between the models (second configuration).

[0182] In the first or second configuration, the abnormal state model includes a bearing lubrication failure model (2112) that models a lubrication failure in a bearing of the motor, The bearing lubrication failure model is configured to receive a mechanical angular velocity from at least the motion equation unit and calculate a friction torque caused by poor lubrication based on the mechanical angular velocity, The calculated friction torque may be input to the equation of motion unit as the abnormality parameter (third configuration).

[0183] In the third configuration, the bearing lubrication failure model is configured to calculate the mechanical angular velocity ω m and mechanical angle θ m and the friction torque T lubrication It is also possible to use a configuration in which the following is calculated (fourth configuration).

number

[0184] In the fourth configuration, the bearing lubrication failure model calculates a normal force N due to poor lubrication based on the following formula: lubrication may be calculated (fifth configuration).

number

[0185] In any one of the first to fifth configurations, the abnormal state model includes a bearing damage model (2113) that models damage in a bearing of the motor, The bearing damage model is configured to calculate a friction torque caused by a bearing damage; The calculated friction torque may be input to the equation of motion unit as the abnormality parameter (sixth configuration).

[0186] In the sixth configuration, the bearing damage model is a normal force N bearing_damage Based on the following formula, the friction torque T bearing_damage The seventh aspect may be configured to calculate the following.

number

[0187] In the seventh configuration, the bearing damage model is such that, in the case of outer ring damage, the revolution angle θ of the rolling element is revolution Each time the following formula is satisfied, the shock pulse height P hight The normal force N bearing_damage may be modeled as occurring (eighth configuration).

number

[0188] In the seventh configuration, the bearing damage model is such that, in the case of inner ring damage, the revolution angle θ of the rolling element is revolution Each time the following formula is satisfied, the shock pulse height P hight The normal force N bearing_damage may be modeled as occurring (ninth configuration).

number

[0189] In the seventh configuration, the bearing damage model is configured such that, in the case of damage to a rolling element, the rotation angle θ rotation Each time the following formula is satisfied, the shock pulse height P hight The normal force N bearing_damage may be modeled as occurring (tenth configuration).

number

[0190] In any one of the fifth, seventh to tenth configurations, the model storage unit stores a vibration model (2114) of the support system when all mechanical elements constituting the motor are defined as a support system, The normal force may be input to a vibration model of the support system (eleventh configuration).

[0191] In the above-mentioned eleventh configuration, the vibration model of the support system may be modeled by an equation of motion of a configuration in which a parallel connection configuration of a spring (Kx, Ky) and a damper (Cx, Cy) is connected to a mass point including the support system (twelfth configuration).

[0192] In the eleventh or twelfth configuration, no signal may be input from the vibration model of the support system to the abnormal state model (thirteenth configuration).

[0193] In any one of the first to thirteenth configurations, the model storage unit a motor model (211) including the motor physical model and the abnormal state model; a sensor model (22) configured to receive physical signal waveform data output from the motor model and output a sense signal; A machine learning model (23) configured to receive the sense signal and perform learning and inference; an anomaly determination model (24) configured to receive an anomaly degree as error data output from the machine learning model and perform an anomaly determination; may be stored (fourteenth configuration).

[0194] Moreover, a program (P) according to an embodiment of the present disclosure is a program for causing a computer (100) to function as a simulation device of any one of the first to fourteenth configurations (fifteenth configuration). [Industrial Applicability]

[0195] The present disclosure can be used, for example, in simulations when considering the introduction of a system for detecting abnormalities in motors. [Explanation of symbols]

[0196] 1 Simulation device 2 Model storage section 3. Model calculation section 4. Model setting section 5. Display control section 6 Operation input section 7 Display section 8 Data storage unit 20 Motor 20A housing 20B rotor 20C axis 20D Stator 20E Bearing 21 System Model 22 Sensor Model 23 Machine Learning Models 24 Anomaly Judgment Model 30 Three-layer neural network 30A Input Layer 30B Hidden layer 30C output layer 100 Computers 100A CPU 100B Memory 100C auxiliary storage 100D Operation input section 100E Display section 201 units 211 Motor Model 211A Abnormal Condition Model 212 Driver Model 213 Load Model 221 Transfer Function Model 222 AD conversion model 231 Pretreatment section 232 Machine Learning Department 2111 Motor Physical Model 2111A Motion equation section 2111B Winding circuit section 2112 Bearing lubrication failure model 2113 Bearing Damage Model 2114 Vibration model of support system Cx, Cy damper ED Explanation Display F external force FLP Abnormal Status Lamp IR Insider Kx,Ky spring LP1~LP4 Lamps LUB Lubricant OR Outer ring P Program RE rolling element SB Simulation start button SG1 Motor type selection section SG2 Abnormal state selection section SG3 Basic parameter setting section SG4 Drive setting section SG5 Stiffness / damping setting section SG6 Abnormal parameter setting section SG7 Poor lubrication position setting section SG8 Bearing damage position setting section SG9 Time setting section SG10 Lubrication failure / deterioration setting section SG11 Bearing damage and deterioration setting section SG12 Sensor type setting section SG13 Filter setting section SG14 Sensor characteristic setting section SG15 Sampling speed setting section SG16 Noise setting section SG17 Sensor position setting unit SG18 Normalization coefficient setting section SG19 Machine Learning Settings SG20 Pre-processing setting section SG21 Abnormality judgment method setting section SG22 Abnormality judgment threshold setting unit ST Simulation time setting section TB Tab VS1 1st vibration sensor VS2 Second vibration sensor

Claims

1. A motor physical model modeled by a winding circuit section and a rotational motion equation section; An abnormal state model that models an abnormal state of the motor; A model storage unit in which the following is stored: a model calculation unit configured to perform calculation processing using the motor physical model; Equipped with An abnormality parameter that indicates the amount of deviation from a normal state calculated by the abnormal state model is input to the motor physical model.

2. The abnormal state model includes two or more motor abnormality models, The simulation device according to claim 1 , wherein the two or more motor abnormality models do not exchange signals between the models.

3. The abnormal state model includes a bearing lubrication failure model that models a lubrication failure in a bearing of the motor, The bearing lubrication failure model is configured to receive a mechanical angular velocity from at least the motion equation unit and calculate a friction torque caused by poor lubrication based on the mechanical angular velocity, The simulation device according to claim 1 , wherein the calculated friction torque is input to the motion equation unit as the abnormality parameter.

4. The bearing lubrication failure model is a model in which the mechanical angular velocity ω ｍ and mechanical angle θ ｍ and the friction torque T ｌｕｂｒｉｃａｔｉｏｎ The simulation device according to claim 3 , wherein the simulation device calculates: [0023] However, B ｌｕｂ_ｃｏｆ１ and B ｌｕｂ_ｃｏｆ２ is a coefficient, and B ｌｕｂ_ｉｎｄｅｘ１ and B ｌｕｂ_ｉｎｄｅｘ２ is the exponent of the mechanical angle velocity, and f(θm) is a function of the mechanical angle θm

5. The bearing lubrication failure model is based on the following formula: ｌｕｂｒｉｃａｔｉｏｎ The simulation device according to claim 4 , which calculates: [0024] However, μ ｌｕｂ is the friction coefficient, and R is the radius of the motor's rotating shaft.

6. The abnormal state model includes a bearing damage model that models damage in a bearing of the motor, The bearing damage model is configured to calculate a friction torque caused by a bearing damage; The simulation device according to claim 1 , wherein the calculated friction torque is input to the motion equation unit as the abnormality parameter.

7. The bearing damage model is a normal force N ｂｅａｒｉｎｇ_ｄａｍａｇｅ and the friction torque T ｂｅａｒｉｎｇ_ｄａｍａｇｅ The simulation apparatus of claim 6 , configured to calculate: [0025] However, μ ｂｅａｒｉｎｇ is the friction coefficient, and R is the radius of the motor's rotating shaft.

8. In the case of outer ring damage, the bearing damage model is ｒｅｖｏｌｕｔｉｏｎ Each time the following formula is satisfied, the height of the shock pulse P ｈｉｇｈｔ The normal force N ｂｅａｒｉｎｇ_ｄａｍａｇｅ The simulation apparatus according to claim 7, wherein the simulation is modeled as occurring. [0026] where Z is the number of rolling elements, P ｗｉｄｔｈ is the width of the pulse impulse, and n is an integer starting from 0

9. In the case of inner ring damage, the bearing damage model is ｒｅｖｏｌｕｔｉｏｎ Each time the following formula is satisfied, the height of the shock pulse P ｈｉｇｈｔ The normal force N ｂｅａｒｉｎｇ_ｄａｍａｇｅ The simulation apparatus according to claim 7, wherein the simulation is modeled as occurring. [0027] Where θ ｍ is the mechanical angle input from the motion equation unit, Z is the number of the rolling elements, P ｗｉｄｔｈ is the width of the pulse impulse, and n is an integer starting from 0

10. In the case of damage to the rolling elements, the bearing damage model is ｒｏｔａｔｉｏｎ Each time the following formula is satisfied, the height of the shock pulse P ｈｉｇｈｔ The normal force N ｂｅａｒｉｎｇ_ｄａｍａｇｅ The simulation apparatus according to claim 7, wherein the simulation is modeled as occurring. [0028] However, P ｗｉｄｔｈ is the width of the pulse impulse, and n is an integer starting from 0

11. the model storage unit stores a vibration model of a support system when all mechanical elements constituting the motor are defined as the support system, 8. The simulation device according to claim 5, wherein the normal force is input to a vibration model of the support system.

12. 12. The simulation device according to claim 11, wherein the vibration model of the support system is modeled by an equation of motion of a configuration in which a spring and a damper are connected in parallel to a mass point including the support system.

13. The simulation device according to claim 11 , wherein no signal is input from the vibration model of the support system to the abnormal state model.

14. The model storage unit includes: a motor model including the motor physical model and the abnormal state model; a sensor model configured to receive physical signal waveform data output from the motor model and output a sense signal; a machine learning model configured to receive the sense signal and perform learning and inference; an anomaly determination model configured to receive an anomaly degree as error data output from the machine learning model and perform an anomaly determination; The simulation device according to claim 1 , wherein:

15. A program for causing a computer to function as the simulation device according to claim 1.

Images