JP2003009578A - Controller for motor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To positively cause a motor to produce maximum torque even at start or very slow speed without complicating the constitution of a rotation sensor. SOLUTION: As one control mode for the motor, a correction mode is provided, in which mode, a control current is determined for each set phase which is varied between the minimum value and the maximum value in the phase judgment section of the rotation sensor, instead of actually detected phase. By supplying a control current based on the dummy detected phase variably set, the motor can be driven in such a current vector phase that the motor produces the maximum torque in the process of variably setting phases. Thus, torque shortage and torque shock which may occur at start or very low rpm wherein phases are difficult to estimate or compute with a rotation sensor with a low resolution are eliminated, and the motor and the drivability of vehicles using the motor can be improved.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a starting control device for an electric motor.

[0002]

2. Description of the Related Art As a starting control method for an electric motor, as in Japanese Patent Laid-Open No. 9-74790, time intervals of angle signals for every 60 electrical degrees are measured based on a signal indicating the rotation angle of the electric motor. A method is known in which the rotation angle is interpolated based on this time interval to create an estimated rotation angle, and the current flowing through the stator winding or the voltage applied to the stator winding is calculated based on the estimated rotation angle. According to this method, even if the angular interval of the rotational position signal is large, the interpolating angle information can be supplemented by interpolation calculation, so the desired detection accuracy can be maintained while simplifying the configuration of the sensors that detect the rotational position. Is possible.

However, according to such a conventional starting control method, only the angle of every 60 degrees can be detected at the time of starting. Therefore, if the actual phase of the rotor deviates from the rotation sensor, a control error due to this phase deviation will occur. This causes a problem that the starting torque and the maximum torque at an extremely low speed are significantly reduced.

The present invention has been made by paying attention to such a conventional problem, and makes the electric motor surely exert the maximum torque at the time of starting or extremely low speed without complicating the structure of the rotation sensor. It is an object of the present invention to provide a control device for an electric motor capable of performing the above.

[0005]

According to a first aspect of the present invention, there is provided a control device for controlling a current supplied to an electric motor based on a signal from a rotation sensor for detecting a rotation phase of the electric motor. Based on the state, a control means for selecting a plurality of preset control modes to control the supply current is provided. Further, as one of the control modes, the control means determines the control current for each set phase changed between the minimum value and the maximum value of the phase determination section of the rotation sensor, instead of the actual detection phase. It has a correction mode.

A second aspect of the present invention is a control device for controlling a current supplied to an electric motor based on a signal from a rotation sensor for detecting a rotation phase of the electric motor, based on a state of a phase signal from the rotation sensor in advance. A control means for selecting a plurality of set control modes and controlling the supply current is provided. Further, the control means, as one of the control modes, a correction mode that determines the phase of the control current supplied to the electric motor for each set phase that is changed between a minimum value and a maximum value in a predetermined range. Equipped with.

According to a third aspect of the present invention, the control means according to the first or second aspect of the present invention has one of the control modes in which a current phase is a predetermined phase with respect to a detection phase (including an estimated value by interpolation processing). It has a normal mode in which a control current is supplied so as to proceed.

According to a fourth aspect of the present invention, in the control means of the first or second aspect of the invention, the rotation sensor is used even when the torque command to the electric motor or its representative parameter is equal to or greater than a reference value for a set time. The correction mode is selected when the phase signal from is not detected.

According to a fifth aspect of the invention, the setting unit of the set phase of the first or second aspect of the invention is set to be equal to or less than the speed calculated from the maximum load and the maximum torque assumed in design.

A sixth aspect of the present invention is the control means according to the first or second aspect of the present invention, wherein when the phase signal from the rotation sensor is not detected even when the set phase in the correction mode reaches the maximum value. , The correction mode is repeated.

According to a seventh invention, when the phase signal from the rotation sensor is detected during the current control in the correction mode, the control means of the first or second invention uses the detected phase as an initial value. It is configured to continue the phase setting.

According to an eighth aspect of the invention, the control means of the sixth aspect of the invention is configured to gradually return the set phase to the minimum value when the set phase reaches the maximum value in the correction mode. .

In a ninth aspect of the invention, the control means of the sixth aspect of the invention is configured to decrease the set unit of the set phase as the number of repetitions of the correction mode increases.

[0014]

According to the first or second aspect of the invention, the pseudo detection phase set in place of the actual rotation phase of the electric motor or the phase of the output current to the electric motor is variably set in predetermined phase units. Since the correction mode is provided, the electric motor can be driven with the current vector phase that surely exhibits the maximum torque in the process of this variable setting. Therefore, for example, when a rotation sensor with a low resolution of a phase angle of 60 degrees is used, it is possible to eliminate torque shortages and torque shocks that occur at the time of starting or in an extremely low speed rotation range where it is difficult to estimate and calculate the phase. The drivability of the used vehicle can be improved significantly.

In the low load operation range where the correction mode does not need to be selected or in the operating range where the rotation speed is relatively high, the current phase is a predetermined phase β (generally β) with respect to the detection phase as shown in the third invention. The control may be performed in the normal mode in which the control current is supplied so as to advance only by π / 2 (rad). In this case, the current phase can obtain sufficient control accuracy even by a method of estimating and outputting the phase of the electric motor by interpolation processing.

According to the fourth aspect of the present invention, when the electric motor is actually judged to have insufficient torque, the correction mode in which the maximum torque is obtained is selected, so that the engine can be efficiently started with a large load. Is possible.

According to the fifth aspect of the present invention, the unit of the phase setting is set to be equal to or less than the speed calculated from the maximum load and the maximum torque assumed in the design, so that the high torque output state is maintained, The rotation angle can be advanced, and the rotor of the electric motor is less likely to get out of step with respect to the control current vector, whereby the starting performance is more reliably improved. Also,
In the vehicle to which the electric motor according to the present invention is applied, the unit of phase setting is determined from the vehicle weight and the maximum torque, so that, for example, when a wheel rides on a step, it is extremely large in a stopped state or an extremely low speed traveling state. Step-out of the electric motor when a load is applied can be suppressed.

According to the sixth aspect of the invention, even if the set phase reaches the maximum value within the set range, if the phase signal from the rotation sensor is not detected, the control is started again from the minimum value in the correction mode. Since it is repeated, it is possible to prevent the abnormal operation after that when the electric motor is out of step with respect to the rotor magnetic field, to reliably perform the start control again, or to repeat the start even if the start fails. .

According to the seventh aspect, when the phase signal from the rotation sensor is detected during the control in the correction mode, the detected phase is set as the initial value and the correction mode is continued. High torque is continuously obtained during the high load state, so that the start under the high load can be smoothly and reliably performed.

According to the eighth invention, in the sixth invention, since the speed at which the phase is returned to the minimum value within the phase setting range is reduced, the torque shock due to the sudden change of the current vector phase is suppressed. You can

According to the ninth aspect of the invention, when the control of the correction mode is repeated in the sixth aspect of the invention, the setting unit of the phase is decreased as the number of repetitions is increased. Accordingly, it becomes possible to reduce the angular velocity of the current vector phase to the extent that step out is not achieved, and the load range that can be started can be expanded.

[0022]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows an outline of a mechanical structure of an embodiment of the present invention. In the figure, 1 is an internal magnet type synchronous motor, and 2 is a phase detector for detecting the rotor phase. The phase detector includes a first hall element 2a,
The second hall element 2b, the third hall element 2c, and a disk 2d attached to the rotor of the electric motor 1 and magnetized so as to have magnetic poles similar to the rotor magnetic poles, and constitute the rotation sensor of the present invention. 3 is a motor control unit, 4 is an inverter driver, 5 is an inverter, and 6 is a battery. The control unit 3 is composed of a microcomputer and its peripheral devices, and functions as the control means of the present invention.

The first to third Hall elements 2a of the phase detector 2
2 to 2c output on / off signals as shown in FIG. From this combination of signals, the rotor phase can be detected with a resolution of 60 electrical degrees. For example, 2a =
When on, 2b = on, 2c = off, rotor phase is 12
It can be seen that the range is from 0 to 180 degrees.

FIG. 3 shows torque characteristics of the electric motor 1. When the rotor phase and the phase of the rotating magnetic field generated by the stator match, the torque generated in the rotor becomes zero (point A),
When the phase of the rotating magnetic field is advanced from this point A, the torque increases, and after reaching the maximum at the point B, the torque decreases when the phase of the rotating magnetic field is further advanced. Such torque characteristics
The (torque curve) is basically determined according to the specifications of the electric motor and changes depending on the operating conditions (rotational speed etc.) of the electric motor.

Generally, the maximum torque is obtained by advancing the rotating magnetic field phase by about π / 2 (rad) from the rotor phase. Therefore, the rotating magnetic field phase is advanced by about π / 2 from the rotor phase. , It is common to express the rotating magnetic field phase by the angle of deviation from this basic state, and this is also followed in this description. Hereinafter, this deviation angle is referred to as a beta angle β.

This beta angle is an angle formed by the current vector and the q-axis when the current supplied to the stator coil is divided into a field component current (d-axis current) and a torque component current (q-axis current) and displayed as a vector. (See FIG. 4). When performing motor control, the beta angle β and the length I of the current vector (corresponding to the amplitude of the alternating current supplied to the stator coil) are determined according to the rotation speed of the rotor and the target torque, and the target is determined from these β and I. The d-axis current tId and the target q-axis current tIq are obtained, the inverter command value is obtained from these tId, tIq and the rotor phase θ, and the inverter 5 is driven based on these command values to control the current flowing through the stator coil.

The phase detector 2 has an angular resolution of only 60 degrees as described above, but when the rotor is rotating at a certain speed, it is possible to obtain a relatively accurate rotor phase θ by a known estimation calculation. Is. Also in this embodiment, when the rotor rotation speed is equal to or higher than the predetermined speed VO, the estimation calculation of the rotor phase θ is performed (normal mode).

The rotor phase θ cannot be accurately estimated when the rotor is not rotating or is rotating at an extremely low speed. Therefore, when the rotor rotation speed is lower than the predetermined speed VO, a fixed value within the determined section is set as the rotor phase θ (fixed mode). For example, when it is known that the rotor phase is in the section from 120 degrees to 180 degrees, the rotor phase θ is fixed to an intermediate value of 150 degrees and the subsequent current control is performed. In this case, since the true phase of the rotor and the overnight phase θ used for current control are deviated by a maximum of 60 degrees, a state occurs in which the actual torque generated by the electric motor 1 becomes smaller than the target torque. However, if a certain amount of torque is obtained and the rotor rotation speed increases, the normal mode can be entered, so there is no problem.

There is a possibility that the rotor does not rotate and the rotor does not rotate even if the electric motor 1 generates a certain amount of torque. For example, the above-described state occurs when the electric motor 1 is an electric motor driving the drive wheels of the vehicle and the drive wheels try to get over the step. Even if the electric motor 1 has a step over which the maximum torque is generated, it may not be possible to overcome the step by the current control in the fixed mode due to the above reason. Therefore, when such a state occurs, the rotor phase change control for gradually changing the rotor phase θ within the determined section is executed (correction mode). In the process of gradually changing the rotor phase θ, there always occurs a time when the rotor phase θ coincides with the rotor phase that produces the maximum torque, so that the maximum torque can be reliably obtained even temporarily.

Next, details of the control will be described with reference to the flow charts shown in FIG. The flowcharts shown in FIG. 5 and subsequent figures show a control routine that is executed by the control unit 3 periodically or at a predetermined timing, and the symbol S in the figure and the following description represents each processing step.

FIG. 5 is a rotor rotation speed calculation routine in which the output of any one of the three Hall elements 2a to 2c changes.
It is executed every time (OFF → ON, ON → OFF), that is, every time the rotor rotates by 60 electrical degrees. S1: The current count value is acquired from the timer counter that is incremented at fixed time intervals, and that value is t0.
And S2: The value of t1 is set to t2, and the value of t0 acquired in S1 is set to t
Substituting 1 in order. t1 represents the current time, and t2 represents the time when this routine was last executed. S3: Time required for the rotor to rotate 60 degrees (t1-
The rotor rotation speed V is calculated by dividing 60 degrees at t2).

FIG. 6 is a main control routine. The command value calculated in this routine is sent to the inverter driver 4, and the inverter driver 4 outputs P based on this command value.
The WM signal is generated, and the switching element of the inverter 5 is turned on / off according to the PWM signal, whereby the electric motor 1
The current flowing through the stainer coil is controlled. S10: Calculate the rotor phase θ. S20: β is calculated based on the rotor rotation speed V and the target torque tT given from the outside. For this, for example, a value is looked up from a control map in which β is stored in advance in association with V and tT. S30: The current amplitude I is calculated based on the rotor rotation speed V and the target torque tT given from the outside. For this, for example, a value is looked up from a control map in which I is stored in advance in association with V and tT. S40: A target d-axis current tId and a target q-axis current tIq are calculated based on β and I. S50: A command value is calculated based on the target d-axis current tId, the target q-axis current tIq, and the rotor phase θ.

FIG. 7 is a rotor phase calculation subroutine, and shows the details of the processing performed in S10 of the main control routine. S100: The rotor phase section is discriminated from the combination of the ON / OFF signals of the first to third Hall elements 2a to 2c of the rotor phase detector 2, and the rotor phase at the head of the discriminated section is set to the section head phase θ1. To do. For example, when 2a = on, 2b = on, 2c = off, the rotor phase is 12
It can be determined that the angle is between 0 and 180 degrees, and in this case, θ1 is 120 degrees. S110-150: The rotor phase θ is calculated according to the mode indicated by the mode selection flag fm. The mode selection flag fm is a mode selection routine (see FIG. 1) described later.
The flag is set in 2). When 3 is selected, the estimation mode is selected, when it is 2, the fixed mode is selected, and when 1 and 0, the correction mode is selected.

FIG. 8 shows an estimation mode rotor phase calculation subroutine, which is S12 of the rotor phase calculation subroutine.
Details of the processing performed in 0 are shown. S121: The current count value is acquired from the timer counter (the same as that of S1 in FIG. 5), and the value is set as ti. S122: The rotor phase θ is calculated by the phase estimation calculation. Specifically, the rotor phase is just the section start phase θ
The difference between the time t1 when it becomes 1 (same as t1 in S2 in FIG. 5) and the current time ti is multiplied by the rotor rotation speed V, and θ
The value obtained by adding 1 is taken as the rotor phase θ.

FIG. 9 shows a fixed mode rotor phase calculation subroutine, which is S14 of the rotor phase calculation subroutine.
Details of the processing performed in 0 are shown. S141: A fixed value (here, 30 degrees) is added to the section head phase θ1 to calculate the rotor phase θ.

FIG. 10 shows a correction mode rotor phase calculation subroutine, which is S1 of the rotor phase calculation subroutine.
Details of the processing performed at 50 are shown. S151: It is determined whether or not the mode selection flag fm is 1. When the correction mode is selected, fm is 1 or 0, and particularly when fm = 1, it indicates that the mode has just been changed from another mode to the correction mode. When the mode selection flag fm is 1, the process proceeds to S152, and the initial process immediately after the shift to the correction mode is performed. Specifically, the value of the counter n is reset to 0, and the initial value of the rotor phase θ is set to the section head phase θ1. The fixed mode should have been selected before shifting to the correction mode, and the rotor phase at this time is θ1 + 30 degrees. If this rotor phase θ is changed stepwise to θ1, some torque shock may occur. If this torque shock becomes a problem, it is advisable to gradually return the rotor phase θ from θ1 + 30 degrees to θ1. Counter n
Is a counter that counts up the number of execution sets of rotor phase change control (the number of times θ is changed from the minimum value to the maximum value of the section). S153: The mode selection flag fm is not 1 (fm =
0), the process proceeds to S153 and S10 of FIG.
It is determined whether or not the section head phase θ1 set at 0 and the section head phase θ1z when the present routine is executed last time are different, that is, whether or not the section is shifted during the correction mode. If a section transition has occurred, proceed to S154,
As the processing when the section transition occurs, the rotor phase θ is set to the section start phase θ1, and the processing proceeds to S155 and thereafter. It should be noted that the fact that the transition of the section has occurred indicates that the rotor has rotated to some extent, and thus the correction mode may be terminated once. If it is determined in S153 that the section transition has not occurred, the process directly proceeds to S155 and thereafter. S155: A phase change width dθ for rotor phase change control is calculated based on the counter n. Specifically, FIG.
As illustrated in, the value is looked up from a control table preset to give dθ according to n. In this control table, dθ is set smaller as n is larger, as shown in the figure. If dθ is large, one set of control can be completed in a short time, but the time for obtaining the maximum torque also becomes short. Therefore, first, a large dθ is used to perform the overnight phase change control,
When the transition to the estimation mode cannot be achieved by the control (corresponding to the state in which the step up cannot be achieved in the example of the vehicle described above), the rotor phase changing control is repeatedly executed while gradually decreasing dθ. The initial value (maximum value) of dθ is the rotor when the maximum torque is generated during the maximum load operation that is assumed in the design (in the example of the vehicle described above, during the riding operation to the maximum step that guarantees the riding). It is set to a value that does not increase the rotor phase change speed faster than the rotation speed. S156: The phase change width dθ is added to the current θ to calculate a new overnight phase θ. S157: It is determined whether or not the rotor phase θ calculated in S156 exceeds the section maximum phase (θ1 + 60 degrees).
If θ> θ1 + 60 degrees, the process proceeds to S158, where the counter n
And the rotor phase θ is returned to the section head phase θ1. The rotor phase θ is θ1 + 60
If a stepwise return from θ to θ1 occurs, some torque shock may occur. If this torque shock poses a problem, it is advisable to gradually return the rotor phase θ to θ1. If θ ≦ θ1 + 60 degrees, S1 as it is
The process proceeds to 59 and below. S159: The mode selection flag fm is set to 0 (that is, the correction mode and the rotor phase changing control is being executed), and the current section head phase θ1 is stored as θ1z.

FIG. 12 shows a mode selection routine, which is executed at predetermined intervals independently of the main control routine. S181: It is determined whether the rotor rotation speed V is higher than a predetermined speed VO. If V> VO, the process proceeds to S182, the counter C is reset to 0, and then the mode selection flag fm is set to 3 (estimation mode) in S183. S181
When it is determined that V≤VO, the process proceeds to S184, and it is determined whether the target torque tT is smaller than the predetermined torque TO. Even if V ≦ VO, it is not necessary to select the special mode when the driver of the electric motor does not demand a large torque, and therefore such a determination is made. S184: If tT <TO, the process proceeds to S185, the counter C is reset to 0, and then the mode selection flag fm is set to 2 (fixed mode) in S186. When it is determined that tT ≧ TO or more, the process proceeds to S187, and the counter C
Is counted up by a predetermined step value dC, and then S1
At 88, it is determined whether the counter C is smaller than the predetermined value ts. S189: If C <ts, the process proceeds to S189, and the mode selection flag fm is set to 2 (fixed mode). That is, even if a state occurs in which the rotor hardly rotates despite the driver requesting a large torque, the fixed mode is selected until the state continues for a predetermined time. If it is determined that C ≧ ts, the correction mode is selected by the processing from S190. S190: The current mode selection flag fm in order to distinguish whether it is a selection to shift from another mode to the correction mode or a reselection in a state where the correction mode has already been selected and the rotor phase change control is being executed. Is greater than 0. If fm> 0, it is determined that the mode is to be shifted to another mode, the process proceeds to S191, and the mode selection flag fm is set to 1 (special mode and immediately after mode transition). If fm = 0, the process proceeds to S192 and f is maintained.
Keep the value of m at 0.

FIG. 13 shows a second embodiment of the motor control according to the present invention. In the first embodiment, the control for generating the maximum torque is performed by variably setting the detection phase in the correction mode, whereas in this embodiment, the beta angle β of the control current is variably set. The points are different. That is, in this embodiment, when the rotor phase is calculated in S10 of the main control routine of FIG. 6, when the mode selection flag fm is 3, the rotor phase is calculated in the estimation mode (FIG. 8) and the mode selection flag gm is 2. , 1,
When 0, the rotor phase is calculated in the fixed mode (FIG. 9).
The rotor phase is not manipulated in the correction mode. On the other hand, when the beta angle is calculated in S20 of the main control routine (FIG. 6), if the mode selection flag fm is 3 or 2, the value βm looked up from the control map is set as it is to the beta angle β, while the mode selection flag is set. fm is 1,0
In that case, the beta angle is calculated using the correction mode beta angle calculation subroutine described in FIG. 13 and below. S201: It is determined whether or not the mode selection flag fm is 1. When fm = 1, the process proceeds to S202, and the initial process immediately after the shift to the correction mode is performed. Specifically, the value of the counter n is reset to 0, and the initial value of the beta angle β is set to a value obtained by subtracting 30 degrees from the map value βm. S203: The angle change width dβ for the beta angle change control is calculated based on the counter n. This is obtained, for example, by referring to a control table preset to give dβ according to n. This control table is configured such that dβ is set smaller as n is larger. S204: A new beta angle β is calculated by adding the angle change width dβ to the current β. S205: The beta angle β calculated in S204 is βm + 30
Judge whether or not the degree is exceeded. Beta angle β is βm +
If it exceeds 30 degrees, the process proceeds to S206 and the counter n
And the beta angle β is βm−
Return to 30 degrees. S207: The mode selection flag fm is set to 0.

[Brief description of drawings]

FIG. 1 is a schematic diagram showing a mechanical configuration of an embodiment of the present invention.

FIG. 2 is an output timing chart of the phase detector.

FIG. 3 is a characteristic diagram showing a relationship between a phase angle (β) of a control current and a generated torque.

FIG. 4 is a flowchart showing a first embodiment of electric motor control according to the present invention.

FIG. 5 is a flowchart showing a first embodiment of the electric motor control according to the present invention.

FIG. 6 is a flowchart showing a first embodiment of electric motor control according to the present invention.

FIG. 7 is a flowchart showing a first embodiment of the electric motor control according to the present invention.

FIG. 8 is a flowchart showing a first embodiment of electric motor control according to the present invention.

FIG. 9 is a flowchart showing a first embodiment of electric motor control according to the present invention.

FIG. 10 is a flowchart showing a first embodiment of electric motor control according to the present invention.

11 is an explanatory diagram of a control table used in the processing of FIG.

FIG. 12 is a flowchart showing a first embodiment of electric motor control according to the present invention.

FIG. 13 is a flowchart showing a first embodiment of the electric motor control according to the present invention.

[Explanation of symbols] 1 electric motor 2 Phase detector 2a, 2b, 2c Hall element 2d disc 3 control unit 4 Inverter driver 5 inverter 6 battery

Claims (9)

[Claims] 1. A control device for controlling a current supplied to an electric motor based on a signal from a rotation sensor for detecting a rotation phase of the electric motor, wherein a plurality of preset values are set based on a state of a phase signal from the rotation sensor. It comprises a control means for selecting a control mode to control the supply current, the control means, as one of the control mode, instead of the actual detection phase, from the minimum value to the maximum value of the phase determination section of the rotation sensor A starting control device for an electric motor, comprising: a correction mode for determining a control current for each set phase changed between the two. 2. A control device for controlling a current supplied to an electric motor based on a signal from a rotation sensor for detecting a rotation phase of the electric motor, wherein a plurality of preset values are set based on a state of a phase signal from the rotation sensor. The control means is provided with control means for controlling the supply current by selecting a control mode, the control means, as one of the control modes, the phase of the control current supplied to the electric motor, from the minimum value to the maximum value of a predetermined range. A starting control device for an electric motor having a correction mode that is determined for each set phase that has been changed up to. 3. The control means has, as one of the control modes, a normal mode for supplying a control current so that the current phase advances by a predetermined phase β with respect to the detection phase.
Alternatively, the starting control device for the electric motor according to claim 2. 4. The control means, when the torque command to the electric motor or the representative parameter thereof is equal to or more than a reference value, when the phase signal from the rotation sensor is not detected even when the set time is continued, The starting control device for an electric motor according to claim 1 or 2, which is configured to select the correction mode. 5. The starting control device for the electric motor according to claim 1, wherein the setting unit of the set phase is set to be equal to or less than a speed calculated from the maximum load and the maximum torque assumed in design. 6. The control unit repeats the correction mode when the phase signal from the rotation sensor is not detected even when the set phase in the correction mode reaches the maximum value. The starting control device for the electric motor according to. 7. The control means, when detecting a phase signal from a rotation sensor during current control in the correction mode, continues the phase setting with the detected phase as an initial value. The starting control device for the electric motor according to. 8. The starting control device for the electric motor according to claim 6, wherein the control means gradually returns the set phase to the minimum value when the set phase reaches the maximum value in the correction mode. 9. The starting control device for an electric motor according to claim 6, wherein the control means decreases the set unit of the set phase as the number of repetitions of the correction mode increases.