JP2007068345A - Control unit of synchronous machine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a control unit of a synchronous machine that can prevent the occurrence of an overcurrent, even when the control of the synchronous machine is started from a state that a rotor is rotating. <P>SOLUTION: A dead time dw of an inverter circuit 22 during a rough estimation period, an initial dead time d0 in particular, is set wider than the dead time dn of the inverter circuit 22 during a normal estimation period. Thereby, a current that flows from the inverter circuit 22 to a synchronous motor 21 can be reduced during the initial period at the start of the control. Therefore, becoming an overcurrent can be controlled even if the estimation during the initial period is not accurate. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a control device capable of starting control of a synchronous machine without an angle sensor from a state in which a rotor rotates.

  As a prior art, a synchronous motor may be used for a ventilation device for a train car. This ventilator controls the rotation of the synchronous motor with the electric power supplied to the vehicle and ventilates the interior of the vehicle. When the vehicle passes through a section that is a non-electric section, power is not supplied to the vehicle, and the ventilator is in an instantaneous power failure state.

  After passing through the section, the ventilator is supplied with power again and resumes rotation control of the synchronous motor. At this time, the ventilator starts rotation control with the rotor of the synchronous motor rotating. The ventilation device needs to estimate the angle and the angular velocity of the rotor of the synchronous motor when starting the rotation control. If the rotor angle and angular velocity are not accurately estimated, the magnitude and phase of the inverter voltage will not correspond to the magnitude and phase of the induced voltage of the motor, and an overcurrent will flow through the synchronous motor and the inverter circuit. May occur. In this case, a great load is applied to the ventilation device, and the device may be damaged.

  Conventionally, the following measures have been taken to solve the above problems. (1) A voltage sensor that detects an induced voltage of the synchronous motor is used. (2) An initial current is passed, so that estimation is performed using an observer (see Patent Document 1). (3) Use parts with large current and voltage capacities. (4) Control is started after the rotation of the rotor stops.

JP 2004-40837 A

  Each method described above has the following problems. (1) Use of a voltage sensor causes an increase in cost and a decrease in maintainability. (2) The overcurrent generated in the initial stage until the estimation by the observer is completed cannot be suppressed. (3) The use of parts having a large capacity leads to an increase in size, weight and cost of the apparatus. (4) When the control is started after the rotation of the rotor stops, it takes time to rotate to a predetermined rotational speed.

  Such a problem occurs in all synchronous machine control devices that start the synchronous machine in a rotated state. For example, the same problem occurs in converter control in a wind power generator using a synchronous generator. Accordingly, an object of the present invention is to provide a control device for a synchronous machine that can prevent an overcurrent from occurring when starting control is started from a state in which a rotor rotates.

The present invention is a control apparatus for a synchronous motor that rotationally drives a synchronous motor by applying AC power from an inverter circuit,
An estimation unit that estimates the angular velocity and angular position of the magnetic poles of the rotor of the synchronous motor based on the current flowing through the synchronous motor;
A control unit that provides a gate signal to the inverter circuit so that the angular velocity of the magnetic pole becomes a predetermined angular velocity based on the angular velocity and the angular position of the magnetic pole estimated by the estimation unit;
The control unit is configured such that the dead time in the inverter circuit when the estimation unit starts estimation is wider than the dead time in the inverter circuit during a normal estimation period after a predetermined condition is satisfied after the estimation unit starts estimation. This is a control device for a synchronous motor.

  Further, according to the present invention, the estimation unit performs an estimation method between a rough estimation period from when the estimation is started until a predetermined condition is satisfied, and a normal estimation period after the predetermined condition is satisfied after the estimation is started. It is characterized by making it different.

Further, the present invention provides a rough estimation period,
The control unit adjusts the dead time so that the current flowing through the synchronous motor becomes a predetermined set current value from a state in which the dead time is extended,
The estimation unit is characterized in that the angular velocity and the angular position of the magnetic pole are estimated based on the dead time.

Further, the present invention provides a rough estimation period,
The control unit is characterized in that the dead time is changed from a state in which the dead time is extended so that the dead time is reduced with time.

Further, the present invention is a control device for a synchronous generator that rectifies AC power supplied from a synchronous generator by a converter circuit,
An estimation unit that estimates the angular velocity and the angular position of the magnetic poles of the rotor of the synchronous generator based on the current flowing through the synchronous generator;
A control unit that applies a gate signal to the converter circuit so that the operation of the converter circuit follows the angular velocity and angular position of the magnetic pole estimated by the estimation unit;
The control unit is configured such that the dead time in the converter circuit when the estimation unit starts estimation is wider than the dead time of the converter circuit in a normal estimation period after a predetermined condition is satisfied after the estimation unit starts estimation. It is the control apparatus of the synchronous generator characterized by doing.

  Further, according to the present invention, the estimation unit includes a rough estimation period from when the estimation is started until a predetermined condition is satisfied, and a normal estimation period after the predetermined condition is satisfied after the estimation is started. It is characterized by making different.

Further, the present invention provides a rough estimation period,
The control unit adjusts the dead time so that the current flowing from the synchronous generator becomes a predetermined set current value from the state in which the dead time is extended,
The estimation unit is characterized in that the angular velocity and the angular position of the magnetic pole are estimated based on the dead time.

Further, the present invention provides a rough estimation period,
The control unit is characterized in that the dead time is changed from a state in which the dead time is extended so that the dead time is reduced with time.

Further, the present invention is a method for controlling a synchronous motor that rotationally drives a synchronous motor by applying AC power from an inverter circuit,
Estimate the angular velocity and angular position of the magnetic poles of the rotor of the synchronous motor based on the current flowing through the synchronous motor, and establish a predetermined condition after starting the estimation of the dead time in the inverter circuit when the estimation starts After the dead time of the inverter circuit in the normal estimation period after
A control method for a synchronous motor, wherein a gate signal is given to an inverter circuit based on an estimation result.

Further, the present invention is a method for controlling a synchronous generator that rectifies AC power supplied from a synchronous generator by a converter circuit,
Estimate the angular velocity and angular position of the magnetic pole of the rotor of the synchronous generator based on the measured current value flowing through the synchronous generator, and estimate the dead time in the converter circuit when the estimation is started. It is wider than the dead time of the converter circuit in the normal estimation period after the specified condition is satisfied,
A control method for a synchronous generator, wherein a gate signal is given to a converter circuit based on an estimation result.

  According to the first aspect of the present invention, the gate signal is supplied to the inverter circuit so that the magnetic pole has a predetermined angular velocity based on the estimation result of the angular velocity and the angular position of the magnetic pole while the magnetic pole of the rotor is rotated. give. Note that the estimation accuracy of the angular velocity and the angular position of the magnetic pole increases with time. In other words, the estimation accuracy is low in the initial period when the estimation starts.

  In the present invention, when estimation is started by the estimation unit, the dead time in the inverter circuit is set widely. Thereby, in the initial period of the estimation start, the current flowing from the inverter circuit to the synchronous motor is reduced. Here, the dead time is a period during which both of the switching elements included in the inverter circuit are turned off in series. When the initial period ends and the normal estimation period starts, the dead time is narrowed.

  By reducing the current flowing through the synchronous motor during the initial period in this way, it is possible to prevent the current flowing through the inverter circuit and the synchronous motor from becoming an overcurrent even if the estimation of the angular velocity and the angular position of the magnetic pole is not accurate. Can do. In the normal estimation period, since the estimation accuracy has increased since time has elapsed from the start of estimation, the current flowing to the inverter circuit and the synchronous motor can be increased even if the dead time is reduced and the current flowing from the inverter circuit to the synchronous motor is increased. Is prevented from becoming overcurrent. Also, by using the estimation result from the initial period, the estimation accuracy in the normal estimation period can be increased, and the torque that can be generated by the synchronous motor by the control unit is increased even if the current applied to the synchronous motor is small. Can do. That is, the torque of the synchronous motor can be generated efficiently.

  Further, according to the present invention, by estimating the angular velocity and angular position of the magnetic pole, a sensor for actually measuring the angular velocity and angular position of the magnetic pole is not required, the structure can be simplified, and the cost can be greatly reduced. It can be expected and maintainability can be improved. Further, even if the estimation accuracy in the initial period when the estimation is started is low, it is possible to prevent an overcurrent, so that an excessive load is not applied to the inverter circuit and the synchronous motor, and they are configured. Each element can be prevented from being damaged. Further, since the control of the inverter circuit can be started without waiting for the rotor to stop, time loss can be eliminated.

  Further, according to the present invention, the current flowing from the inverter circuit to the synchronous motor is different even if the induced voltage is constant due to the different dead times. In the present invention, the angular velocity and the angular position of the magnetic pole can be accurately estimated from the current values in the rough estimation period and the normal estimation period by using different estimation methods for the normal estimation period and the rough estimation period, respectively. And the estimation accuracy can be improved. For example, the estimation accuracy can be improved by changing the control model used for estimation between the rough estimation period and the normal estimation period.

  According to the third aspect of the present invention, the current value at which the angular velocity and the angular position of the magnetic pole can be estimated is set as the set current value, so that the dead time and the angular velocity and the angular position of the magnetic pole are one-to-one. Can be supported. As a result, the estimation unit can easily estimate the angular velocity and the angular position of the magnetic pole only by obtaining the dead time. Also, the current value flowing through the synchronous motor is set as a command value, and the current flowing through the synchronous motor from the inverter circuit is set to the command value by estimating the angular velocity and angular position of the magnetic pole based on the dead time when the command value is reached. It is possible to prevent the overcurrent from being exceeded.

  According to the fourth aspect of the present invention, the dead time is changed so that the dead time is narrowed as time elapses from the widened dead time. This eliminates the need to set feedback control for changing the dead time based on the current value. For example, by setting the dead time at the end of the rough estimation period to be substantially the same as the dead time set in the normal estimation period, the rough estimation period can be smoothly shifted to the normal estimation period.

  According to the fifth aspect of the invention, the angular velocity and the angular position of the magnetic pole are estimated in a state where the magnetic pole of the rotor is rotated, and the gate signal is supplied to the converter circuit so that the operation of the converter circuit follows the estimation result. give. Note that the estimation accuracy of the angular velocity and the angular position of the magnetic pole increases with time. In other words, the estimation accuracy is low in the initial period when the estimation starts.

  In the present invention, when estimation is started by the estimation unit, the dead time in the converter circuit is set widely. As a result, in the initial period of estimation start, the current flowing from the converter circuit to the synchronous generator is reduced. Here, the dead time is a period during which both of the switching elements included in the converter circuit are switched off in series. When the initial period ends and the normal estimation period starts, the dead time is narrowed.

  By reducing the current flowing to the synchronous generator during the initial period in this way, the current flowing to the converter circuit and the synchronous generator is prevented from being overcurrent even if the estimation of the angular velocity and the angular position of the magnetic pole is not accurate. can do. In the normal estimation period, since the estimation accuracy has increased since time has elapsed from the start of estimation, even if the dead time is reduced and the current flowing from the converter circuit to the synchronous generator is increased, the converter circuit and the synchronous generator It is prevented that the flowing current becomes an overcurrent. In addition, by using the estimation results from the initial period, the estimation accuracy in the normal estimation period can be increased, and even if the current supplied from the synchronous generator is small, the control unit efficiently converts AC power into DC power. can do.

  Further, according to the present invention, by estimating the angular velocity and angular position of the magnetic pole, a sensor for actually measuring the angular velocity and angular position of the magnetic pole is not required, the structure can be simplified, and the cost can be greatly reduced. It can be expected and maintainability can be improved. Moreover, even if the estimation accuracy in the initial period from the start of estimation is low, it is possible to prevent an overcurrent, so that an excessive load is not applied to the converter circuit and the synchronous generator. It is possible to prevent each element to be damaged. In addition, since the control of the converter circuit can be started without waiting for the rotor to stop, time loss can be eliminated.

  According to the sixth aspect of the present invention, the current flowing from the synchronous generator to the converter circuit differs even when the induced voltage is constant due to the different dead times. In the present invention, the angular velocity and the angular position of the magnetic pole can be accurately estimated from the current values in the rough estimation period and the normal estimation period by using different estimation methods for the rough estimation period and the normal estimation period, respectively. And the estimation accuracy can be improved. For example, the estimation accuracy can be improved by changing the control model used for estimation between the rough estimation period and the normal estimation period.

  According to the seventh aspect of the present invention, the current value at which the angular velocity and the angular position of the magnetic pole can be estimated is set as the set current value, so that the dead time and the angular velocity and the angular position of the magnetic pole are one-to-one. Can be supported. As a result, the estimation unit can easily estimate the angular velocity and the angular position of the magnetic pole only by obtaining the dead time. In addition, the value of the current flowing through the synchronous generator is set as a command value, and the current flowing from the synchronous generator to the converter circuit is estimated by estimating the angular velocity and angular position of the magnetic pole based on the dead time when the command value is reached. The overcurrent can be reliably suppressed without exceeding the value.

  According to the eighth aspect of the present invention, the dead time is changed so that the dead time is narrowed as time elapses from a state where the dead time is widened. This eliminates the need to set feedback control for changing the dead time based on the current value. For example, by setting the dead time at the end of the rough estimation period to be substantially the same as the dead time set in the normal estimation period, the rough estimation period can be smoothly shifted to the normal estimation period.

  According to the present invention described in claim 9, in the initial period when the estimation is started, the dead time of the inverter circuit is widened so that the current flowing from the inverter circuit to the synchronous motor is kept small, and the magnetic pole angular velocity and The angular position is estimated, and a gate signal is applied to the inverter circuit so that the angular velocity of the magnetic pole becomes a predetermined angular velocity based on the estimated angular velocity and the angular position. As a result, even if the estimation of the angular velocity and the angular position of the magnetic pole in the initial period is not accurate, it is possible to prevent the current flowing through the inverter circuit and the synchronous motor from becoming an overcurrent.

  According to the tenth aspect of the present invention, in the initial period when the estimation is started, the dead time of the converter circuit is widened so that the current flowing from the synchronous generator to the converter circuit is suppressed, and the magnetic pole angular velocity is reduced. Further, the angle position can be estimated, and the operation of the converter circuit can be made to follow the estimation result. As a result, even if the estimation of the angular velocity and the angular position of the magnetic pole in the rough estimation period is not accurate, it is possible to prevent the current flowing through the converter circuit and the synchronous generator from becoming an overcurrent.

  FIG. 1 is a block diagram showing an electric motor drive control system 20 according to the first embodiment of the present invention. In the present embodiment, the electric motor drive control system 20 is used in a ventilation device that ventilates a train vehicle using the synchronous motor 21. In this case, when the vehicle passes through the section, power is not supplied to the vehicle, and the control system 20 enters an instantaneous power failure state. Then, after passing through the section, the control system 20 is supplied with power again, and resumes rotation control of the synchronous motor. At this time, the control system 20 starts rotation control of the synchronous motor with the rotor of the synchronous motor rotating.

  The control system 20 estimates the angular velocity and angular position of the magnetic pole of the rotor of the synchronous motor 21, and the synchronous motor 21 rotates so that the rotor rotates at a desired angular velocity based on the estimated angular velocity and angular position of the magnetic pole. Adjust the AC voltage applied to the.

  The motor drive control system 20 includes a synchronous motor 21, a PWM inverter circuit 22, a DC circuit unit 23, and a control device 24. The synchronous motor 21 is a three-phase synchronous motor, and is configured sensorless without an angle sensor for detecting the angular velocity of the rotor. For example, a cylindrical brushless DC motor is used as the synchronous motor 21. The DC circuit unit 23 is a circuit that generates predetermined DC power, and is provided with a voltage sensor 29. The voltage sensor 29 supplies a DC bus voltage signal indicating a voltage applied from the DC circuit unit 23 to the inverter circuit 22 to the control device 24.

FIG. 2 is a diagram illustrating an example of a circuit configuration of the PWM inverter circuit 22. PWM (Pulse
The inverter circuit 22 is realized by, for example, a three-phase bridge circuit. The inverter circuit 22 converts the DC power supplied from the DC circuit unit 23 into three-phase AC power and supplies it to the synchronous motor 21. The inverter circuit 22 has three output terminals 25a, 25b, and 25c, which are connected to the three terminals 26a, 26b, and 26c of the synchronous motor 21 through connection paths 27a, 27b, and 27c, respectively.

  The inverter circuit 22 is provided with a plurality of switching elements Tr for flowing an alternating current through the connection paths 27a to 27c. The switching element Tr is turned on and off when a gate signal is given from the control device 24. The switching element Tr is in a conductive state between terminals at both ends of the switching element Tr in an on state, and is in a disconnected state between terminals at both ends of the switching element Tr in an off state. In the present embodiment, each switching element Tr is realized by a transistor. A diode D is connected to each switching element Tr in antiparallel.

  The control device 24 can supply a sinusoidal alternating current whose phase is shifted by 120 degrees to each of the connection paths 27a to 27c by applying a pulse width modulated gate signal to each switching element Tr. By flowing such an alternating current through the connection paths 27a to 27c, the rotor of the synchronous motor 21 is rotationally driven. The control device 24 can change the speed of the rotor of the synchronous motor 21 by adjusting the gate signal.

  Current sensors 28 a and 28 c are interposed in two connection paths 27 a and 27 c among the three connection paths 27 a to 27 c that connect the inverter circuit 22 and the synchronous motor 21. The first current sensor 28a measures the current flowing through one of the three connection paths 27a, and gives the measured value to the control device 24 as a first current sensor signal. The second current sensor 28c measures the current flowing through the other connection path 27c among the three, and gives the measured value to the control device 24 as the second current sensor signal. In this way, each of the current sensors 28a and 28c measures the current flowing from the inverter circuit 22 to the synchronous motor 21, and gives the actually measured value to the control device 24 as a current sensor signal.

  The control device 24 acquires a DC bus voltage sensor signal given from the voltage sensor 29 and each current sensor signal given from the current sensor 28. The control device 24 determines an actual measurement value of the current flowing through the synchronous motor 21 based on the acquired current sensor signal. When the power supply is started after the power supply from the power supply device for driving the control device 24 is stopped, the control device 24 determines that the section has passed and estimates the angular velocity and angular position of the magnetic poles of the rotor. Determine the start. In addition, the estimation start timing may be determined based on a control start command from another device.

  The control device 24 estimates the angular velocity and angular position of the magnetic poles of the rotor of the synchronous motor 21 based on the actual measured value of the current flowing through the synchronous motor 21, and the angular velocity and angular position of the magnetic poles estimated by the estimating unit. And a control unit that applies a gate signal to each switching element Tr of the inverter circuit 22 so that the angular velocity of the magnetic pole becomes a predetermined angular velocity.

  The control unit starts the estimation of the angular velocity and the angular position of the magnetic pole by the estimation unit and then the rough estimation period until the predetermined condition is satisfied, and after the predetermined condition is satisfied after the estimation by the estimation unit is started. The normal estimation period is determined. In the present embodiment, the control unit determines the rough estimation period from the start of estimation by the estimation unit until a predetermined time elapses, and determines the start of the normal estimation period after the predetermined time elapses. Moreover, in this Embodiment, a control part controls the synchronous motor 21 via the inverter circuit 22 based on the estimation result of an estimation part in a normal estimation period.

  The rough estimation period includes an initial period immediately after the estimation is started. The control unit gives a gate signal to each switching element Tr of the inverter circuit 22 so that the dead time dn in the rough estimation period is wider, that is, longer than the dead time dw in the normal estimation period. Here, the dead time d is a period during which all the switching elements Tr contributing to AC conversion among the switching elements Tr included in the inverter circuit 22 are in an off state, and both the switching elements Tr connected in series are off. It is a period.

Hereinafter, in order to facilitate understanding, the inverter circuit 22 will be described as a single-phase full-bridge circuit 22A. FIG. 3 is a circuit diagram showing a part of the inverter circuit 22 as a single-phase full bridge circuit 22A. The full bridge circuit 22A includes a first input path 43 and a second input path 44 connected in parallel to the two input terminals 41 and 42 to which the DC voltage from the DC circuit unit 23 is applied. Two switching elements Tr _1U , Tr _1L ; Tr _2U , Tr _2L are interposed in series in the first input path 43 and the second input path 44, respectively. A path portion 90 between the two switching elements Tr _1U and Tr _1L of the first input path 43 is connected to the first output terminal 25a via the first output path 92, and the two switching elements Tr of the second input path 44 are connected. A path portion 91 between _2U and Tr _2L is connected to the second output terminal 25b through the second output path 93.

  The first output terminal 25 a and the second output terminal 25 b are connected by a connection path 46. In the connection path 46, the resistor R and the inductor L are connected in series. The connection path 46 corresponds to a path connecting two terminals of the three terminals 26 a to 26 c of the synchronous motor 21. Therefore, when the rotor of the synchronous motor 21 rotates, an induced voltage Em is generated in the connection path 46.

  Of the two input terminals 41 and 42, the input terminal having a higher potential due to the voltage applied from the DC circuit unit 23 is the first input terminal 41, and the input terminal having the lower potential is the second input. Terminal 42 is assumed. Each switching element Tr allows current to flow from the first input terminal 41 side to the second input terminal 42 side through the input paths 43 and 44 in the ON state. Each diode D connected in antiparallel to each switching element Tr allows current to flow from the second input terminal 42 side to the first input terminal 41 side.

Of the two switching elements Tr _1U and Tr _1L of the first input path 43, the switching element closer to the first input terminal 41 is referred to as Tr _1U, and the switching element closer to the second input terminal 42 is referred to as Tr _1L . Of the two switching elements Tr _2U and Tr _2L of the second input path 44, the switching element closer to the first input terminal 41 is referred to as Tr _2U, and the switching element closer to the second input terminal 42 is referred to as Tr _2L .

Controller 24, the _{Tr 1U} and _{Tr 2L} is turned on by the off state and _{Tr 2U} and _{Tr 1L,} current connection path 46 flows in a first direction indicated by the arrow marks in FIG. Further, when the control device 24 turns off Tr _1U and Tr _2L and turns on Tr _2U and Tr _1L , the current flows through the connection path 46 in the second direction opposite to the first direction. . In this way, the control device 24 switches the on / off state of each switching element Tr and alternately switches the direction of the current flowing through the connection path 46, thereby allowing an alternating current to flow through the connection path 46. In this case, the dead time d means a period in which Tr _1U , Tr _1L , Tr _2U , and Tr _2L are all off.

  FIG. 4 is a timing chart showing the on / off state of each switching element Tr during the normal estimation period.

If the two switching elements Tr connected in series are simultaneously turned on, the DC power supply unit 23 is short-circuited and a very large current flows, which may cause the switching element Tr to be destroyed. In order to avoid such a state, in the normal estimation period, the two switching elements Tr _1U , Tr _1L ; Tr _2U , Tr _2L connected in series are not simultaneously turned on, and as much as possible. The dead time dn is set to be narrow. Such dead time dn is determined due to the performance of the switching element Tr.

Specifically, as shown in FIG. 4, each switching element Tr repeats the switching operation every PWM carrier cycle W0 in order to flow an alternating current. In the PWM carrier cycle W0, a first on period W1 in which both Tr _1U and Tr _2L are in an on state and a second on period W2 in which both Tr _2U and Tr _1L are in an on state are set. A period dn1 from the end of the first on-period W1 to the start of the second on-period W2 and a period dn2 from the end of the second on-period W2 to the start of the first on-period W1 are respectively Dead time dn is reached. For example, when the PWM carrier frequency is 8 kHz, that is, the PWM carrier period W0 is 125 μsec, the dead times dn1 and dn2 in the normal estimation period are set to 5 μsec, for example.

Such an on / off state of each switching element Tr in the normal estimation period is an example. Therefore, an on / off state different from the on / off state of each switching element Tr described above may be performed in the normal estimation period. For example, corresponding to setting a period W11 in which Tr _1U and Tr _2U are in an on state and a period W12 in which Tr _1L and Tr _2L are in an on state, such as a rough estimation period to be described later. , Tr _1U and Tr _1L may be paired and Tr _2U and Tr _2L may be paired to switch the on / off state.

  FIG. 5 is a timing chart showing an on / off state of each switching element Tr in the rough estimation period. In the present embodiment, the dead time dw in the rough estimation period is set wider than the dead time dn in the normal estimation period.

Specifically, as shown in FIG. 5, each switching element Tr repeats the switching operation every PWM carrier cycle W0. In the PWM carrier cycle W0, a first on period W11 in which both Tr _1U and Tr _2U are in an on state and a second on period W12 in which both Tr _1L and Tr _2L are in an on state are set. A period dw1 from the end of the first on period W11 to the start of the second on period W12 and a period dw2 from the end of the second on period W12 to the start of the first on period W11 are: Each becomes the dead time dw. For example, the dead time dw in the rough estimation period is set to be less than ½ of the PWM cycle. For example, when the PWM carrier frequency is 8 kHz, that is, the PWM carrier cycle W0 is 125 μsec, the dead times dw1 and dw2 in the rough estimation period are For example, 57 μs is set.

  As described above, the control device 24 adjusts the gate signal supplied to the inverter circuit 22A so that the dead time dw in the rough estimation period is wider than the dead time dn in the normal estimation period. In other words, the control device 24 makes the total period (W11 + W12) of the on periods W11 and W12 in the rough estimation period narrower than the total period (W1 + W2) of the on periods W1 and W2 in the normal estimation period. . Thereby, in the rough estimation period, the power source impedance viewed from the synchronous motor 21 becomes equivalently high, and the current flowing through the connection path 46 can be reduced. This is a three-phase bridge circuit 22 and can also be applied to the inverter circuit 22 of the synchronous motor 21.

  In this way, by reducing the current flowing through the synchronous motor 21 during the rough estimation period, the overcurrent generated in the inverter circuit 22 is suppressed even if the estimation of the angular velocity and the angular position of the magnetic poles during the rough estimation period is not accurate. be able to. In the normal estimation period, the angular velocity and the angular position of the magnetic pole are estimated using the estimated value in the rough estimation period. As a result, the estimation accuracy is increased. Therefore, even if the current that can flow from the inverter circuit 22 to the synchronous motor 21 is increased by narrowing the dead time dn in the normal estimation period, an overcurrent is prevented from occurring in the inverter circuit 22. Can do. Further, by using the estimation result from the initial period to improve the estimation accuracy in the normal estimation period and controlling the inverter circuit 22 based on the estimation result, the synchronous motor can be obtained even if the current supplied to the synchronous motor 21 is small. The torque that can be generated 21 can be increased. That is, torque can be generated efficiently.

  FIG. 6 is a timing chart showing temporal changes in the on / off state of the switching element Tr, the induced voltage Em, the line voltage Ea, and the line current ia during the rough estimation period. The induced voltage Em is a voltage generated by the angular displacement of the rotor of the synchronous motor 21. The line voltage Ea is a voltage between the two output terminals 25a and 25b in the inverter circuit 22A. The line current ia is a current flowing through the connection path 46 and can be detected by the current sensor 28a. The line current ia is a positive value when the current flows in the first direction indicated by the arrow in FIG. The induced voltage Em is a positive value when the current flows through the connection path 46 in the second direction opposite to the first direction indicated by the arrow in FIG. 3 at the dead time dw.

  In FIG. 6, it is assumed that the voltage effect by the switching element Tr and the diode D is zero. Since the PWM carrier cycle W0 is extremely short, it is assumed that the induced voltage Em is constant during one PWM carrier cycle W0. FIG. 6 shows a case where the induced voltage Em is a negative value. In this case, the line current ia flows in the first direction indicated by the arrow in FIG. 3 through the connection path 46 at the dead time dw.

In the present embodiment, the control device 24, in the rough estimation period, the first on period W11 in which Tr _1U and Tr _2U are turned on and Tr _1L and Tr _2L are turned off per PWM carrier cycle W0. Then, the gate signal is adjusted so that the magnitudes of the second ON period W12 in which Tr _1U and Tr _2U are turned off and Tr _1L and Tr _2L are turned on coincide with each other. In other words, the PWM modulation rate of the inverter circuit 22 is adjusted to zero. Also, a first dead time dw1 that is a period from the end of the first on-period W11 to the start of the second on-period W12, and a period from the end of the second on-period W12 to the start of the first on-period W11. The gate signal is adjusted so that the magnitude of the second dead time dw2 that is the period of time coincides.

The value of the line voltage Ea in each of the ON periods W11 and W12 is equal to the value obtained by inverting the sign of the induced voltage Em. Therefore, when the value of the induced voltage Em and _{-V Em,} the value of the line voltage Ea of the on-period W11, W12 _{becomes V EM.} Since the inductor L is provided in the connection path 46 and the diode D is provided in the inverter circuit 22, the value of the line voltage Ea at each dead time dw1 and dw2 is between the two input terminals 41 and 42 of the inverter circuit 22. A value (−V _PN ) obtained by adding a value (−V _PN ) obtained by reversing the value V _PN of the power supply voltage E, which is a voltage, and a value (V _Em ) obtained by reversing the positive and negative values of the induced voltage Em (−V _PN + V _Em ). .

  Thus, the voltage value of the line voltage Ea is determined by the induced voltage Em, the power supply voltage E, and the on / off state of the switching element Tr. The current value of the line current ia is determined by the line voltage Ea.

  Since the resistance R and the inductor L exist in the line path 46, the line current ia repeats four states X1 to X4. First, in the first ON period W11, the current flowing in the first direction through the connection path 46 in the first state X1 increases, and the line current ia increases with time. Next, at the first dead time dw1, the current flowing through the connection path 46 in the first direction as the second state X2 decreases, and the line current ia decreases with time.

  Next, in the second ON period W12, the current flowing in the first direction through the connection path 46 in the third state X3 increases, and the line current ia increases with time. Next, at the second dead time dw2, the line current ia flowing in the first direction through the connection path 46 as the fourth state X4 decreases, and the line current ia decreases with time. Table 1 shows the state of the switching element Tr and the line current ia in each of the states X1 to X4. In Table 1, an increase in the line current ia means an increase in the current flowing in the first direction, and a decrease in the current means a decrease in the current flowing in the first direction.

  The line-to-line current ia is small if the ON periods W11 and W12 of the switching element Tr are short. In other words, the line current ia becomes smaller if the dead times dw1 and dw2 of the switching element Tr are longer. Further, since the line voltage Ea changes corresponding to the induced voltage Em, the line current ia that changes corresponding to the line voltage Ea also changes corresponding to the induced voltage Em.

In the present embodiment, when the set value that can be set such as the dead time dw is appropriately set, and the value of the induced voltage Em (−V _Em ) is a negative value, the average current of the line current ia The value is adjusted to maintain a positive value corresponding to the magnitude of the value of the induced voltage Em. In the case the value of the induced voltage Em (V _Em) is a positive value, the average current value of the line current ia is adjusted to maintain a negative value corresponding to the magnitude of the value of the induced voltage Em .

Thus, since the line current ia and the induced voltage Em have a one-to-one relationship, the induced voltage Em can be obtained by actually measuring the line current ia. Specifically, when the value of the line current ia is Ia and the value of the induced voltage Em is Em, the value of the induced voltage Em satisfies the following relational expression.
Em = f (Ia) (1)

  In the equation (1), f (Ia) is a function of the value of the line current ia. This function can be obtained in advance by an approximate expression or a theoretical expression obtained through experiments. By following the equation (1), the induced voltage Em can be obtained from the line current ia.

Furthermore, since the induced voltage Em and the magnetic pole angular velocity ω satisfy the following relational expression, the magnetic pole angular velocity ω can be estimated from the induced voltage Em.
Em = Ke · ω (2)

  In equation (2), Em represents the induced voltage, ω represents the angular velocity of the magnetic pole, and Ke represents a constant. Further, the angular position of the magnetic pole can be estimated by integrating the angular velocity ω of the magnetic pole obtained by the equation (2).

FIG. 7 is a timing chart showing temporal changes in the on / off state of the switching element Tr, the induced voltage Em, the line voltage Ea, and the line current ia during the rough estimation period. 7, the induced voltage Em is a positive value V _Em as compared with the case shown in FIG. In this case, the line current ia becomes a negative value. That is, at the dead time dw, current flows in the second direction through the connection path 46.

In this case, the value of the line voltage Ea in each of the on periods W11 and W12 is equal to a value obtained by inverting the sign of the induced voltage Em. Therefore, when the value of the induced voltage Em is V _Em , the value of the line voltage Ea in the on periods W11 and W12 is −V _EM . The value of the line voltage Ea of each dead time dw1, dw2 includes a value _{V PN} of the power supply voltage E, the value sign has been inverted induced voltage Em _{(-V Em)} added value and _(V PN _{-V Em} ) Table 2 shows the switching element Tr and the state of the line current ia in this case. In Table 2, an increase in current means an increase in current flowing in the second direction, and a decrease in current means a decrease in current flowing in the second direction.

  Thus, regardless of whether the induced voltage Em is positive or negative, the angular velocity ω of the magnetic pole can be obtained from the line current ia by using the above-described equations (1) and (2). When the line current ia is zero, the line voltage Ea is also zero.

In the present embodiment, when the induced voltage Em is a positive value V _Em , the line current ia is maintained at a negative value, and when the induced voltage Em is a negative value (−V _Em ) By setting the dead time dw so that the current ia is maintained at a positive value, the induced voltage Em can be obtained easily and accurately.

  FIG. 8 is a timing chart showing temporal changes in the on / off state of the switching element Tr, the induced voltage Em, the line voltage Ea, and the line current ia when the dead time dn is narrow. As a comparative example, as shown in FIG. 8, when the dead times dn1 and dn2 are narrow, the periods W1 and W2 in which the switching elements are turned on become longer, so the line current ia increases. In this case, the line current Ia may exceed a preset allowable current value Imax.

  On the other hand, in the present embodiment, as shown in FIGS. 6 and 7, by increasing the dead time dw, even if the estimation of the angular velocity of the magnetic pole and thus the induced voltage Em is not accurate, the line current ia Can be prevented from becoming overcurrent, and the inverter circuit 22 and the synchronous motor 21 can be prevented from being damaged.

  When the rough estimation period ends and the routine proceeds to the normal estimation period, the angular velocity and the angular position of the magnetic pole can be estimated with high accuracy, and an inverter voltage based on the induced voltage Em can be applied. Therefore, even if the dead time d is reduced in the normal estimation period, the line voltage Ea can be prevented from increasing, and the line current ia can be prevented from becoming an overcurrent. Further, by using the estimation result from the rough estimation period, the estimation accuracy in the normal estimation period is improved, and the inverter circuit 22 is controlled based on the estimation result, so that even if the current supplied to the synchronous motor 21 is small. The torque that can be generated by the synchronous motor 21 can be increased. That is, the torque of the synchronous motor 21 can be generated efficiently. Further, such an effect can be achieved regardless of the circuit configuration of the inverter circuit. Therefore, the same effect can be obtained even with the three-phase bridge circuit 22 shown in FIG. In addition, other on / off states of the switching element Tr can be selected.

  FIG. 9 is a timing chart showing the on / off state of the switching element Tr and the temporal change of the line current ia in the rough estimation period in the modification of the first embodiment. Also in this modification, it is assumed that the voltage effect by the switching element Tr and the diode D is zero. Further, it is assumed that the induced voltage Em is constant during one PWM carrier period W0.

In the present embodiment, the first on period W21 in which Tr _1U and Tr _2L are in the on state and Tr _1L and Tr _2U are in the off state per one PWM carrier cycle W0, and Tr _1U and Tr _2L are in the off state. And the gate signal is adjusted so that the magnitudes of the second ON period W22 during which Tr _1L and Tr _2U are in the ON state coincide with each other. That is, the PWM modulation rate of the inverter circuit 22 is adjusted to zero.

  Also, a first dead time dw11, which is a period from the end of the first on-period W11 to the start of the second on-period W12, and a period from the end of the second on-period W12 to the start of the first on-period W11. The gate signal is adjusted so that the magnitude of the second dead time dw12, which is the period of time, coincides.

  Also in the present embodiment, in the steady state of the rough estimation period, when the induced voltage Em is a positive value, the line current ia is maintained at a negative value, and when the induced voltage Em is a negative value. The dead time dw is set so that the line current ia is maintained at a positive value.

As shown in FIG. 9 (3), when the induced voltage Em is a negative value (V _Em <0), the line current ia is first in the first on-period W11, and the line current ia is in the first state X11. Increase with time. Next, in the first dead time dw21 and the second ON period w21, the line current ia decreases as time passes as the second state X12. In the second state X12, as time elapses, the line current ia approaches a constant value. Then, the line current ia repeats the first state X11 and the second state X12 in order, and maintains a positive value. Here, the increase in the line current ia means an increase in the current flowing in the first direction, and the decrease in the current means a decrease in the current flowing in the first direction.

As shown in FIG. 9 (4), when the induced voltage Em is a positive value (V _Em > 0), the line current ia is first in the first state during the first on-period W11 and the first dead time dw11. As X21, the line-to-line current ia decreases with time. In the first state X21, as time elapses, the line current ia approaches a constant value. Next, in the second on-period w21, the line current ia increases with time as the second state X22. Next, at the second dead time dw12, the line current ia decreases as time elapses as the third state X23. Then, the line current ia repeats the first state X21 to the third state X23 in order and maintains a negative value. Here, the increase in the line current ia means an increase in the current flowing in the second direction, and the decrease in the current means a decrease in the current flowing in the second direction.

  Even in the on / off state of the switching element Tr shown in such a modification, the angular velocity ω of the magnetic pole can be estimated from the line current ia in the same manner as described above. Further, by expanding each dead time dw11, dw12 as compared with the normal estimation period, it is possible to prevent the line current ia from becoming an overcurrent and to prevent the inverter circuit 22 and the synchronous motor 21 from being damaged. it can.

  Further, by using the estimation result from the initial period, it is possible to increase the estimation accuracy in the normal estimation period, and to increase the torque that can be generated by the synchronous motor 21 even if the current applied to the synchronous motor 21 is small. it can. That is, the torque of the synchronous motor 21 can be generated efficiently. In FIGS. 6 to 9, the direction in which the current flows is maintained constant according to the sign of the induced voltage. However, the present invention is not limited to this.

  FIG. 10 is a graph showing the relationship among the dead time d, the induced voltage Em, and the line current ia. When the average output voltage output from the inverter circuit 22 to the synchronous motor 21 is a predetermined value, the dead time d, the induced voltage Em, and the line current ia are in a relative relationship. Specifically, when the dead time d is constant, the induced voltage Em increases and the line current ia increases. When the induced voltage Em is constant, the dead time d is widened and the line current ia is small.

  Therefore, the line current ia can be suppressed by widening the dead time d. However, if the dead time d is excessively wide, the line current ia cannot flow when the induced voltage Em is large, and the magnetic pole The angular velocity cannot be measured. When the dead time d is narrowed, the line current ia becomes an overcurrent when the value of the induced voltage Em is large.

  Therefore, when the possible range of the induced voltage Em can be predicted in advance, the dead time d is determined based on the prediction, thereby preventing the line current ia from becoming an overcurrent and measuring the angular velocity of the magnetic pole. can do. For example, when the generation induced voltage range of the synchronous motor 21 is known, an appropriate dead time d in the possible generation range may be set as the dead time dw in the rough estimation period. In addition, when the control device 24 can determine the possible generation range of the induced voltage Em at the time of rough estimation from the actual measurement value or the estimated value before the start of the flying control, the relationship as shown in FIG. Accordingly, the dead time dw that can measure the angular velocity of the magnetic pole while preventing the line current ia from becoming an overcurrent may be determined.

For example, when the relationship shown in FIG. 10 is established and an overcurrent occurs when the absolute value of the line current ia exceeds 30 A, the induced voltage Em at the time of rough estimation is expected to be generated in the range of 52 to 60V. Then, the control device 24 sets the dead time d to 30 μsec (3 × 10 ⁻⁵ sec). As a result, the current value of the line-to-line current ia can be over -30 A and less than 0 A, the angular velocity of the magnetic pole can be measured, and an overcurrent can be prevented.

  In the first embodiment of the present invention, the dead time dw in the rough estimation period is a constant value. However, in order to prevent overcurrent more reliably, the dead time dw in the rough estimation period is set as the second embodiment. The dead time dw may be changed so that the target current flows from the connection path 46. The second embodiment will be described below.

  In the second embodiment, the initial value of the dead time dw in the rough estimation period is set wider than the dead time dn in the normal estimation period. For example, by using a field programmable gate array (FPGA), the dead time can be changed during the rough estimation period. By making the dead time d variable using an FPGA, it is possible to change the dead time d by hardware. As a result, the stability of the control device 24 can be improved as compared with the case where the dead time is changed by software.

  FIG. 11 is a graph for explaining the case where the dead time d is variable. When the average output voltage output from the inverter circuit 22 to the synchronous motor 21 is a predetermined value, the dead time d, the induced voltage Em, and the line current ia are correlated. Specifically, when the current value of the line current ia becomes a predetermined value, the dead time d and the induced voltage Em have a proportional relationship.

As shown in FIG. 11, when the current value of the line current ia becomes zero, the following relational expression is satisfied, where the induced voltage is Em and the dead time is d.
_{Em = K 1 · d ... (} 3)

In the equation (3), K ₁ is a predetermined constant, and is set to 1.7 × 10 ⁶ [V / sec] in FIG. As a result, if the induced voltage Em at any given time is larger than K ₁ times the dead time d at that time, the line current ia flows. If the induced voltage Em is smaller than K ₁ times the dead time d, the line current ia becomes smaller. Not flowing.

Since it cannot be measured if the line current ia does not flow, during the rough estimation period, the rough estimation is performed with the line current ia that does not become an overcurrent flowing. In the present embodiment, rough estimation is performed so that the current value of the line current ia becomes a predetermined set current value. As shown in FIG. 11, when the current value of the line current ia becomes the set current value, the following relationship is satisfied.
Em = K ₁ · d + K ₂ (4)

Note that in (4), _{K 2} is a constant determined in advance. Set current value in FIG. 11 when it is -30A, the _{K 2} is set to 10 [V].

In the present embodiment, the dead time in the first period of the rough estimation period is set as the initial dead time d0, and the dead time d is gradually increased from the initial dead time d0 until the line current ia having a predetermined set current value flows. Narrow. Note that the initial dead time d0 is set to 1/2 or less of the PWM period without generating an overcurrent even when the assumed maximum induced voltage Em is generated, for example, 5 × 10 ⁻⁵ seconds (50 μsec). Set as The initial dead time d0 is set wider than at least the dead time dw in the normal estimation period. Further, the set current value exceeds 0 and falls within the allowable current value range, and is set to, for example, 30A as an absolute value.

When the actual induced voltage Em is 50V, as shown in FIG. 11, the dead time d is gradually narrowed from 5 × 10 ⁻⁵ seconds, which is the initial dead time d0, to about 2.9 × 10 ⁻⁵ seconds. When it is narrowed, a current starts to flow, and when it becomes about 2.3 × 10 ⁻⁵ seconds, the current value of the line current ia becomes −30A. When the dead time when the current value of the line-to-line current ia is −30 A is obtained in this way, the induced voltage Em is estimated to be 50 V by substituting the dead time d into the equation (4). be able to. Further, the angular velocity and the angular position of the magnetic pole can be obtained based on the estimated induced voltage Em.

Although the case where the induced voltage Em is positive has been described, the graph when the induced voltage Em is negative is a graph in which the graph of FIG. 11 is arranged symmetrically with respect to the Y axis, and satisfies the relationship shown below.
−Em = −K ₁ · d−K ₂ (5)

  In the present embodiment, when the dead time is gradually changed in the rough estimation period, the control device 24 actively changes the dead time d based on the current value of the line current ia. Then, after reaching the set current value of the line current ia, the dead time is adjusted to be constant. When the control device 24 adjusts the dead time to be constant and a predetermined period elapses, the controller 24 determines that the accuracy of the angular velocity and angular position of the magnetic pole estimated at that time is high, and shifts to the normal estimation period. Also good.

  FIG. 12 is a block diagram illustrating a control block configuration in the rough estimation period of the control device 24 according to the second embodiment of the present invention. In the present embodiment shown in FIG. 12, the dead time d is actively changed during the rough estimation period. Since the other configuration is the same as that of the control device 24 of the first embodiment, the description of the same configuration is omitted.

  The control device 24 uses a d-axis current parallel to the magnetic flux generated by the magnetic pole on the rotating coordinate system rotating at the same speed as the magnetic pole, from the line current ia given from the two current sensors 28a and 28b, respectively ( π / 2) Conversion into q-axis current with advanced phase. Then, the control device 24 manipulates the dead time based on the q-axis current in the rough estimation period. In addition, the effective value by a current sensor may be used for q-axis current.

Specifically, the control unit 24, the actually measured q-axis current actual measurement value I _{[delta] r} determined from the current as an input value, the q-axis current command value I _[delta] the predetermined target value. Further, the q-axis current command value I _γr is set to −30 A that is the set current value of the line-to-line current ia. The initial dead time d0 is also set to a predetermined value.

The control device 24 includes a first subtractor 60, a first arithmetic unit 61, and a first adder 62 as a control unit 70 for controlling the inverter circuit. First subtractor 60, a q-axis current command value I _[delta], to obtain the q-axis current actual measurement value I _{[delta] r,} subtracts the q-axis current actual measurement value I _{[delta] r} from the q-axis current command value I _[delta], the calculation The result is given to the first calculator 61. The first arithmetic unit 61 performs arithmetic processing on the arithmetic result of the first subtractor 60 according to a predetermined arithmetic expression, and calculates an operation amount for changing the dead time. In the present embodiment, the first computing unit 61 preliminarily sets a proportional value obtained by multiplying the computation result of the first subtractor 60 by a predetermined proportional gain K _qp and an integral value obtained by integrating the computation result of the first subtractor 60. An integral value obtained by multiplying a _{predetermined} integral gain K _qi is obtained.

  The first adder 62 acquires the initial dead time d0, the proportional value and the integral value calculated by the first calculator 61, adds them, and sets the calculation result as a new dead time. Then, a gate signal is given to the inverter circuit 22 so that the set dead time is reached, and the inverter circuit 22 is controlled.

If this by the q-axis current command value I _[delta] and q-axis current actual measurement value I _{[delta] r} do not match, changing the dead time, to match the q-axis current actual measurement value I _{[delta] r} in the q-axis current command value I _[delta] over time Can be dead time. In addition, by including the integral calculation in the first calculator 61, it is possible to prevent a steady deviation from occurring between the q-axis current actual measurement value I _δr and the q-axis current command value I _δ, and more accurately the q-axis current. It is possible to calculate a dead time in which the command value I _δ and the q-axis current actual measurement value I _δr can be matched.

As shown in FIG. 11, the induced voltage Em can be obtained by obtaining the dead time when the current of the q-axis current command value _Iδ flows. The control device 24 has a scale converter 67, a second subtractor 64, a second calculator 65, a second adder 66, an integral as an estimation unit 71 for estimating the angular velocity and the angular position of the magnetic pole. Instrument 68.

The scale converter 67 acquires the dead time calculated by the control unit 70, and multiplies the acquired dead time by a predetermined scale conversion coefficient _Kwt . The scale conversion coefficient K _wt is a coefficient for calculating the angular velocity of the magnetic pole from the dead time, and can be obtained based on the above-described equation (4) or (5) and equation (2). The scale converter 67 outputs the calculation result as an estimated value of the magnetic pole angular velocity.

Second subtractor 64 obtains the d-axis current command value I _gamma, and d-axis current actual measurement value I _.gamma.r, subtracts the d-axis current actual measurement value I _.gamma.r from d-axis current command value I _gamma, the operation The result is given to the second calculator 65. For example the d-axis current command value I _gamma is set to 0A. The second calculator 65 performs calculation processing on the calculation result of the second subtractor 64 according to a predetermined calculation formula, and calculates a calculation amount for estimating the angle of the magnetic pole. In the present embodiment, the second computing unit 65 preliminarily sets a proportional value obtained by multiplying the computation result of the second subtractor 64 by a predetermined proportional gain K _dp and an integral value obtained by integrating the computation result of the second subtractor 64. A calculation result obtained by adding the integral value obtained by multiplying the integral gain _Kdi to be determined is supplied to the second adder 66.

  The integrator 68 acquires the estimated angular velocity value of the magnetic pole provided from the scale conversion unit 67, integrates the estimated value, and provides the integrated value to the second adder 66. The second adder 66 adds the calculation result given from the second calculator 65 and the calculation result given from the integrator 68, and outputs the result as an estimated value of the magnetic pole angle.

  In this way, the control device 24 can estimate the angular velocity and the angular position of the magnetic pole during the rough estimation period. Note that the operations of the blocks of the control unit 70 and the estimation unit 71 described above may be realized by hardware or software. When realized by software, the control device 24 executes a control program corresponding to the block diagram of FIG. 12 to realize operations corresponding to the respective blocks. The block configuration described above is an example of the embodiment. Therefore, even if a part of each arithmetic unit is different or the addition and subtraction of the adder and subtracter are different, they are included in this embodiment.

  FIG. 13 is a block diagram illustrating a control block configuration in the normal estimation period of the control device 24. The control device 24 varies the estimation method between the rough estimation period and the normal estimation period. In the normal estimation period, the dead time dn is fixed to a predetermined value. For the estimation of the magnetic pole angular velocity and angle and the control of the inverter circuit 22 during the normal estimation period, known estimation methods and control methods can be used.

In the present embodiment, the control device 24 includes an estimation unit 72 for estimating the angular velocity and the angular position of the magnetic pole, and a control unit 73 for controlling the inverter circuit 22. The estimation unit 72 includes a dq converter 80 for calculating the d-axis current actual measurement value I _γr and the q-axis current actual measurement value I _δr , a model device 81 having a model arithmetic expression representing a model of the synchronous motor 21, and a third subtraction. And an estimator 82 having an estimation algorithm for estimating the magnetic pole angular velocity, angular position, and induced voltage based on the output of the modeler 81.

Based on the current values Iu and Iv given from the current sensor and the angular position θm of the magnetic pole estimated by the estimation unit 82, the dq converter 80 _{measures the} d-axis current measured value I _γr and the q-axis current measured value I _δr. And the calculation result is given to the model device 81.

Model 81, together with the given d-axis current actual measurement value I _.gamma.r and q-axis current actual measurement value I _{[delta] r} from dq converter 80, the control unit 73 d-axis voltage operation value V _gamma and the q-axis voltage operation value V _[delta] And is given. An induced voltage Em estimated by the estimator 82 is given.

The model device 81 substitutes the given information into the model arithmetic expression to solve the model arithmetic expression. Then, the d-axis model actual current value I _Mγ and the q-axis model actual current value I _Mδ that are current values that will flow to the synchronous motor 21 are calculated, and the calculation result is given to the third subtractor 83. The third subtracter 83 is supplied with the d-axis model actual current value I _Mγ and the q-axis model actual current value I _Mδ from the model unit 81, and from the dq converter 80 with the measured d-axis current value I _γr and the q-axis current. An actual measurement value I _δr is given, and a deviation between them is calculated. Specifically, by subtracting the d-axis model actual current value I _Emuganma from d-axis current actual measurement value I _.gamma.r, obtains the d-axis current error [Delta] I _gamma, q-axis current actual measurement value I _{[delta] r} from the q-axis model actual current value I the subtracted _Emuderuta, obtains the q-axis current error [Delta] I _[delta]. The third subtracter 83 gives the calculation result to the estimator 82.

The estimator 82 is provided with the current errors ΔI _γ and ΔI _δ from the third subtractor 83, and based on the current errors ΔI _γ and ΔI _δ and a predetermined estimation algorithm, the induced voltage Em and the magnetic pole The angular position θm and the magnetic pole angular velocity ωm are estimated. The estimator 82 gives the estimated induced voltage Em to the model unit 81. The estimated angular position θm of the magnetic pole is given to the dq converter 80 and the inverse dq converter 87. Further, the estimated angular velocity ωm of the magnetic pole is given to the control unit 73.

The control unit 73 includes a fourth subtracter 84, a speed controller 85, a current controller 86, an inverse dq converter 87, and a PWM pulse generator 88. The fourth subtracter 84 acquires a preset speed command value ωr and an estimated angular speed ωm. Then, the estimated angular speed ωm is subtracted from the speed command value ωr, and the calculation result is given to the speed controller 85. The speed controller 85 calculates the q-axis current command value I _δ based on the calculation result given from the fourth subtracter 84 and gives the calculation result to the current controller 86.

The current controller 86 is supplied with the q-axis current command value I _δ given from the speed controller 85 and the d-axis current command value I _γ set according to the torque, speed, and the like. From these pieces of information, the d-axis voltage operation value V _γ and the q-axis voltage operation value V _δ are calculated. The calculation result is given to the model unit 81 and the inverse dp converter 87. The inverse dp converter 87 converts the given voltage operation values V _γ and V _δ into three-phase voltage operation values and supplies them to the PWM pulse generator 88. The PWM pulse generator 88 performs pulse width modulation on the three-phase voltage operation value supplied from the inverse converter 87 and supplies it to the PWM inverter circuit 22 as a gate signal.

Thus, in the normal estimation period, the control device 24 can rotate the rotor of the synchronous motor 21 at a predetermined rotation speed based on the speed command ωr. Further, the control device 24 can apply a current of a d-axis component to the synchronous motor 21 by performing current control based on the d-axis current command value _Iγ . Thus, field weakening control can be performed. That is, when the induced voltage Em increases due to the high-speed rotation of the rotor, it is possible to suppress a decrease in the generated torque of the synchronous motor 21. In addition, in this embodiment, it has the estimation part 72 which estimates the angular velocity and angle of a magnetic pole, what is called an observer. Therefore, a sensor for measuring the angular velocity of the magnetic pole such as an encoder and a resolver is not required. As a result, the electric motor drive control system 20 can be configured at low cost, the structure can be simplified, and maintainability can be improved.

  Further, when shifting from the rough estimation period to the normal estimation period, the angular velocity and the angular position of the magnetic pole estimated in the rough estimation period are set as initial values of the estimated values in the normal estimation period. Specifically, it is given to the estimated angular velocity ωm given to the speed controller 85, the estimated angle θm given to the dq converter 80 and the inverse dq converter 87, and the induced voltage Em given to the model unit 81. Further, it is appropriately given as an initial value to some of the plurality of integrators included in the estimation unit 72 and the control unit 73.

  FIG. 14 is a flowchart showing the operation of the control device 24. When the control device 24 determines in step s0 that power is supplied from the voltage sensor 29 or the like, the control device 24 proceeds to step s1 and starts the control operation of the PWM inverter circuit 22.

  In step s1, the control device 24 is initialized, and the process proceeds to step s2. In step s2, estimation of the angular velocity and the angular position of the magnetic pole is started by the estimation method in the rough estimation period described above. Specifically, the current value of the line current ia is adjusted to be the set current ia while gradually reducing the dead time from the initial dead time, and the process proceeds to step s3.

  In step s3, when it is determined that the line current ia is adjusted to the set current value and the dead time d is constant, the angular velocity, angular position, and induced voltage of the magnetic pole at that time are stored, and in step s4 move on. In step s4, the magnetic pole angular velocity, angular position, and induced voltage stored in step s3 are substituted as the initial value of the estimated value at the start of the normal estimation period. Then, while estimating the angular velocity and the angular position of the magnetic pole by the estimation method in the normal estimation period, a gate signal is given to the PWM inverter circuit 22 so that the magnetic pole is angularly displaced at a preset speed. When an end command is given, the process proceeds to step s5, and the operation ends.

  FIG. 15 is a timing chart showing time variations of the dead time d, the magnetic pole angular velocity ω, the magnetic pole angular position θ, and the line current ia. When estimation is first started, a rough estimation period is set, and then a normal estimation period is set.

  At the start of control, the dead time is the initial dead time d0, which is narrowed with time and is adjusted to a constant value depending on the induced voltage Em. The line current ia starts from a state where the current value is zero or very small because the initial dead time is set to a wide value. Then, after a lapse of time, the current starts to flow after the dead time becomes narrow to some extent. After the current flows, the absolute value of the current value increases with time. The estimation of the angular velocity and the angular position of the magnetic pole changes as the dead time changes.

  When the current value of the line current ia reaches a predetermined set current value, the angular velocity and the angular position of the magnetic pole are accurately estimated. The dead time becomes a constant value after the current value of the line current ia reaches a predetermined set current value. When the control device 24 determines that the dead time is constant for a predetermined time, the normal estimation period starts.

  When the normal estimation period starts, the dead time is adjusted to the dead time in the normal estimation period. Also, estimation of the normal estimation period is started with the angular velocity and angular position of the magnetic pole estimated in the rough estimation period as initial values. Further, the magnitude of the line current ia is changed due to the control of the synchronous motor.

  As described above, according to the second embodiment, the dead time dw in the rough estimation period is made wider than the dead time dn in the normal estimation period. As a result, the current flowing from the inverter circuit 22 to the synchronous motor 21 can be reduced in the initial period of the estimation start. Thereby, in the initial period when the estimation of the inverter circuit 22 is started, the estimation accuracy of the magnetic pole angular velocity and angular position is low, and the phase of the output voltage of the inverter circuit 22 and the induced voltage generated by the synchronous motor is shifted. In the case where the induced voltage is large, it is possible to prevent the current flowing from the inverter circuit 22 to the synchronous motor 21 from being an overcurrent. In the normal estimation period, the estimation accuracy is increased by substituting the angular velocity and angle position of the magnetic pole estimated in the rough estimation period as initial values related to the estimation in the normal estimation period. Even if the current flowing through the electric motor 21 is increased, the current flowing through the inverter circuit 22 and the synchronous motor 21 can be prevented from becoming an overcurrent. Thereby, even if sensorless control is started in the normal estimation period, the synchronous motor 21 can be rotated smoothly and shocklessly. Further, by using the estimation result from the rough estimation period, it is possible to improve the estimation accuracy in the normal estimation period, and to increase the torque that can be generated by the synchronous motor 21 even if the current applied to the synchronous motor 21 is small. Can do. That is, the torque of the synchronous motor 21 can be generated efficiently.

  Thus, according to the present embodiment, since it is not necessary to use a sensor for measuring the angular position of the magnetic pole, the structure of the electric motor drive control system 20 can be simplified, and a significant cost reduction can be expected. Maintainability can be improved. Moreover, even if conditions such as dust, vibration, and temperature are severe as environmental conditions in which the synchronous motor is used and the sensor cannot be used, the synchronous motor can be controlled without a sensor.

  In addition, even if the estimation accuracy in the initial period when the estimation is started is low, it is possible to prevent an overcurrent, so that an excessive load is not applied to the inverter circuit 22 and the synchronous motor 21, and they are damaged. Thus, the life of the motor drive control system 20 can be extended. In addition, since the control of the inverter circuit 22 can be started without waiting for the rotor to stop, time loss can be eliminated. Further, by changing the dead time, the current can be actively reduced, and overcurrent can be surely prevented regardless of the influence of the estimation method set by the control device 24, the feedback period, and the like. Accordingly, the present embodiment can be applied over a wide range.

  In addition, the current flowing from the inverter circuit 22 is different because the dead time is different between the rough estimation period and the normal estimation period. In the present embodiment, as shown in FIG. 12 and FIG. 13, by using different estimation methods for the estimation unit between the normal estimation period and the rough estimation period, the magnetic poles can be obtained from the current values in the rough estimation period and the normal estimation period. The angular velocity and the angular position can be estimated with high accuracy, and the estimation accuracy can be improved. Specifically, the estimation accuracy can be improved by changing the model of the inverter circuit 22 used for estimation in the rough estimation period and the normal estimation period. FIG. 12 schematically shows a control block configuration of the control device 24 shown in FIGS. 12 and 13, and shows a part of the control device 24. Therefore, the control device 24 may have other configurations than the entire configurations of FIGS. 12 and 13. For example, the control blocks shown in FIGS. 12 and 13 can be realized by the operation program stored in the storage unit of the control device 24 being executed by the calculation unit of the control device 24, respectively.

  As shown in FIG. 6, the first embodiment of the present invention is based on the current value measured when a fixed dead time is used in which the dead time is widened in the rough estimation period rather than the dead time in the normal estimation period. Thus, the induced voltage may be estimated. However, when the dead time is constant in the rough estimation period, there is a possibility that an overcurrent occurs if the induced voltage Em is large, and no current flows if the induced voltage Em is small.

  In contrast, in the second embodiment of the present invention, the dead time is gradually narrowed from the initial dead time d0 in the rough estimation period. In this case, the current value gradually changes and reaches the set current value. As a result, a current having a current value that allows estimation of the angular velocity and the angular position of the magnetic pole can be passed before an overcurrent occurs, and the estimation can be performed reliably without an overcurrent. Further, by setting the dead time as the manipulated variable, the dead time reaching the set current value can be set in a short time, and the rough estimation period can be shortened. Thereby, it is possible to quickly shift to the normal estimation period.

  As shown in FIG. 11, in the present embodiment, the induced voltage is estimated based on the dead time when a predetermined set current value flows in a state where the output average voltage of the inverter circuit 22 is zero. Then, according to the block configuration shown in FIG. 12, the angular velocity and the angular position of the magnetic pole are estimated from the induced voltage Em. Thereby, the angular velocity and the angular position of the magnetic pole can be easily estimated regardless of the change in the dead time. Further, by controlling the dead time so that the actual measured value becomes the current command value, the current value flowing from the inverter circuit 22 to the synchronous motor can be prevented from exceeding the command value, and the overcurrent can be reliably suppressed.

  Such an electric motor drive control system 20 of the present embodiment can be suitably applied to the field in which the rotor of the synchronous motor rotates at high speed, and is preferably applied to, for example, an air-conditioning ventilator mounted on a train car. When the vehicle passes through the section and power is supplied again, the control of the synchronous motor can be resumed quickly without load.

  FIG. 16 is a flowchart showing another operation of the control device 24. When the angular velocity of the magnetic pole at the start of estimation is equal to or lower than a predetermined start speed, the control device 24 may start control the synchronous motor so as to reach the start speed. In this case, in steps s10 to s13, the same operations as in steps s0 to s3 shown in FIG. 14 are performed. When it is determined in step s13 that the rough estimation is completed, the process proceeds to step s14. In step s14, the control device 24 determines whether or not the estimated angular velocity of the magnetic pole is equal to or higher than a predetermined starting speed. If the controller 24 determines that the estimated angular velocity of the magnetic pole is less than the starting velocity, the process proceeds to step s15. If it is determined that the estimated angular velocity of the magnetic pole is equal to or higher than the starting velocity, the process proceeds to step s16.

  In step s15, according to a predetermined startup sequence, the inverter device 22 is controlled to rotate the synchronous motor, and the process proceeds to step s16. Also in step s16, the angular velocity, angular position, and induced voltage of the magnetic pole estimated in step s13 are substituted as initial values in the starting sequence. If the angular velocity of the magnetic pole is equal to or higher than the starting velocity, the magnetic pole's angular velocity, angular position, and induced voltage at that time are stored, and the process proceeds to step s16.

  In step s16, the angular velocity, angular position, and induced voltage of the magnetic pole stored in the previous step are substituted as the initial value of the estimated value at the start of the normal estimation period. Then, a gate signal is given to the PWM inverter 22 so that the magnetic pole is angularly displaced at a preset speed while estimating the angular velocity and the angular position of the magnetic pole by the estimation method in the normal estimation period. When an end command is given, the process proceeds to step s17 to end the operation.

  Even when such an operation is performed, the same effect as described above can be obtained. Further, by performing the start sequence using the estimated value in the rough estimation period as an initial value, it is possible to suppress the occurrence of overcurrent even during the start sequence operation. In addition, at the time of starting, the synchronous motor 21 can be rotated smoothly and shocklessly. As a starting sequence, for example, there is a method of keeping the output voltage and frequency of the inverter constant until a predetermined speed is reached.

FIG. 17 is a block diagram showing a control block configuration in the rough estimation period of the control device 24 in the third embodiment of the present invention. The third embodiment has a configuration similar to that of the second embodiment. In the third embodiment, the description of the configuration corresponding to the configuration of the second embodiment is omitted, and the same reference numerals are given. In the second embodiment, the dead time is changed according to the measured q current value I _δr . However, in the third embodiment, the dead time is elapsed from the initial dead time d0 regardless of the measured q current value I _δr. Set to narrow together. Therefore, the change in the dead time d is determined by the time function f (t). Further, when the dead time dn in the rough estimation period is narrowed with time and the dead time in the rough estimation period becomes equal to the dead time dw in the normal estimation period, the change of the dead time is stopped.

  The control unit 70 of the control device 24 includes a rough estimation period from when the estimation of the inverter circuit 22 is started until a predetermined condition is satisfied, and a normal estimation period after the predetermined condition is satisfied after the estimation is started. Judging.

  The control unit 70 makes the initial dead time d0 when the estimation of the inverter circuit 22 starts wider than the dead time dn in the normal estimation period. Then, the dead time dw is changed so as to narrow with the passage of time in the rough estimation period. In the present embodiment, the control unit makes the dead time dw immediately before the end of the rough estimation period substantially equal to the dead time dn of the normal estimation period.

Control unit 70 also changes the PWM modulation rate of inverter circuit 22 based on q-axis current actual measurement value _Iδr . The estimation unit 71 estimates the angular velocity of the magnetic pole based on the PWM modulation rate output from the control unit 70. Further, the angular position of the magnetic pole is estimated based on the estimated angular velocity of the magnetic pole and the actually measured d-axis current value _Iγr . For example, a relational expression between the PWM modulation rate and the induced voltage when the dead time dw of the normal estimation period is reached has already been acquired, and the induced voltage is estimated by substituting the modulation rate into this relational expression.

Specifically, the control device 24 includes a first subtractor 60, a first arithmetic unit 61, and a first adder 62, as in the second embodiment. The difference from the second embodiment is that the first calculator 61 calculates an operation amount for changing the PWM voltage modulation rate and the inverter frequency of the inverter circuit 22. Further, the first adder 62 integrates a proportional value obtained by multiplying the calculation result of the first subtractor 60 by a predetermined proportional gain K _qp and an integration gain K _qi predetermined by an integral value obtained by integrating the calculation result of the first subtractor 60. Find the integrated value multiplied by.

  The first adder 62 acquires the proportional value and the integral value calculated by the first calculator 61, adds them, and sets the calculation result as a new PWM modulation rate. And the control apparatus 24 controls the inverter circuit 22 so that it may become the computed dead time and PWM modulation factor. Since the modulation factor is a value that substantially depends on the induced voltage Em, the angular velocity and the angular position of the magnetic pole can be obtained from the modulation factor. As in the second embodiment, the control device 24 includes a scale converter 67, a second subtracter 64, a second calculator 65, a second adder 66, an integrator 68, and the like as an estimation unit 71. Have

The scale converter 67 acquires the PWM modulation rate calculated by the control unit 70 and multiplies the acquired PWM modulation rate by a predetermined scale conversion coefficient K _WV . This scale conversion coefficient is a coefficient for calculating the angular velocity of the magnetic pole from the modulation rate. The scale converter 67 outputs the calculation result as an estimated value of the magnetic pole angular velocity. Note that the second subtractor 64, the second arithmetic unit 65, the second adder 66, and the integrator 68 are the same as those in the second embodiment, and a description thereof will be omitted. In the present embodiment, the PWM modulation rate is changed to bring the measured q-axis current value I _{δr closer} to the q-axis current command value I _δ , and the induced voltage Em is obtained based on the PWM modulation rate at that time. In addition, by changing parameters other than the PWM modulation rate, the induced voltage Em can be obtained based on the parameter value when the q-axis current command value _Iδ is reached. This parameter is a parameter that has a one-to-one correspondence with the induced voltage Em at a preset current value determined by the line current ia, and is a parameter that can be controlled by the control device 24.

  FIG. 18 is a timing chart showing temporal changes of the dead time d, the magnetic pole angular velocity ω, the magnetic pole angular position θ, and the line current ia. First, when the control operation is started, it becomes a rough estimation period and then a normal estimation period.

  At the start of control, the dead time is the initial dead time d0, which becomes narrower as time elapses, and stops changing when it becomes equal to the dead time dn of the normal estimation period. Since the initial dead time d0 is set wide, the line current ia starts from a state where the current value is zero or very small. And when time passes and a dead time becomes narrow to some extent, an electric current will begin to flow. Then, the PWM voltage modulation rate, the inverter frequency, and the estimated angle are adjusted so that the actual measured current value converges to a predetermined current command value. For example, in FIG. 18, the current command value is set to zero. The estimation of the angular velocity and the angular position of the magnetic pole changes with the change of the PWM modulation rate.

  Also in such an embodiment, as the estimation time becomes longer, the accuracy of estimation of the angular velocity and the angular position of the magnetic pole is improved, and in a state where the dead time dw in the rough estimation period reaches the dead time dn in the normal estimation period, The angular velocity and angular position of the magnetic pole are accurately estimated. When the control device 24 determines that a predetermined time has elapsed after becoming equal to the dead time dn in the normal estimation period, the control device 24 shifts to the normal estimation period and starts control of the synchronous motor 21. When the control of the synchronous motor 21 is started, estimation of the normal estimation period is started by using the angular velocity and the angular position of the magnetic pole estimated during the rough estimation period as initial values. Further, the magnitude of the line current ia changes due to the control of the synchronous motor. For example, the rough estimation period is set within ¼ of the rotation period of the synchronous motor.

  As described above, according to the third embodiment, the same effect as that of the second embodiment can be obtained. Further, by causing the dead time dw to change during the rough estimation period regardless of the actual measured current value, it is not necessary to actively change the dead time d, and the dead time d can be easily set. . Further, the dead time dw is set to be equal to the dead time dn of the normal estimation period immediately before the end of the rough estimation period. Therefore, even when the rough estimation period shifts to the normal estimation period, the shock due to the change of the dead time d Can be eliminated. Therefore, even if sensorless control is started in the normal estimation period, the synchronous motor 21 can be rotated smoothly and shocklessly.

  In addition, in the rough estimation period, a change period for changing the dead time d and a steady period having a dead time dn equal to the normal estimation period are provided. By providing the steady period, the angular velocity and the angular position of the magnetic pole can be accurately estimated at the end of the rough estimation period even when both the modulation factor and the dead time change.

  FIG. 19 is a block diagram showing a power generation control system 120 according to the fourth embodiment of the present invention. The power generation control system 120 according to the fourth embodiment of the present invention generates DC power using the synchronous generator 121 and is used for, for example, a wind power generator. In this case, a windmill rotating by wind is connected to the rotor of the synchronous generator 121. The control system 120 starts rectification control of the electric power generated from the synchronous generator 121 in a state where the rotor of the synchronous generator 121 is rotated.

  The control system 120 estimates the angular velocity and angular position of the magnetic poles of the rotor of the synchronous generator 121 and adjusts the operation of the PWM converter circuit 122 that is a rectifier so as to follow the estimated angular velocity and angular position of the magnetic poles. To do. The power generation control system 120 includes a synchronous generator 121, a PWM converter circuit 122, a load circuit unit 123, and a control device 124.

  The synchronous power generation 121 is a three-phase synchronous generator, and is configured sensorlessly without an angle sensor for detecting the angular position of the rotor. The load circuit 123 is supplied with the DC power rectified by the PWM converter circuit 122 and is provided with a voltage sensor 129. The voltage sensor 129 gives a DC bus voltage signal applied from the converter circuit 122 to the load circuit 123 to the control device 124.

  The PWM converter circuit 122 rectifies the AC power generated by the synchronous generator 122 into DC power, and includes, for example, a three-phase bridge circuit. The configuration of the converter circuit 122 is substantially the same as that of the inverter circuit 22 shown in FIG. Current sensors 128 are interposed in two of the three connection paths 127 that connect the converter circuit 122 and the synchronous generator 122. Each current sensor 128 gives an actual measured value of the current to the control device 124.

  The control device 124 acquires the DC bus voltage sensor signal given from the voltage sensor 129 and each current sensor signal given from the current sensor 128. The control device 124 determines an actual measurement value of the current flowing through the synchronous generator 121 based on the acquired current sensor signal. Further, the control device 124 determines the timing of the estimation start based on a control start command from another device.

  The control device 124 estimates the angular velocity and angular position of the magnetic poles of the rotor of the synchronous generator 121 based on the measured current value flowing through the synchronous generator 21, and the angular velocity of the magnetic poles estimated by the estimating unit. And a control unit that applies a gate signal to each switching element Tr of the converter circuit 122 so that the angular velocity of the magnetic pole becomes a predetermined angular velocity based on the angular position.

  The present embodiment can also be applied to such a power generation system. The control unit starts the estimation of the angular velocity and the angular position of the magnetic pole by the estimation unit and then the rough estimation period until the predetermined condition is satisfied, and after the predetermined condition is satisfied after the estimation by the estimation unit is started. The normal estimation period is determined. The control unit gives a gate signal to each switching element Tr of the converter circuit 122 so that the dead time in the rough estimation period becomes wider than the dead time in the normal estimation period. In the present embodiment, the control unit starts control of the synchronous motor 21 via the converter circuit 122 based on the estimation result of the estimation unit in the normal estimation period.

  FIG. 20 is a timing chart showing temporal changes in the dead time d, the magnetic pole angular velocity ω, the magnetic pole angular position θ, and the line current ia in the fourth embodiment. In the fourth embodiment, the configuration of the control device 124 is similar to that of the second embodiment.

The control device 124 has control blocks corresponding to FIGS. 12 and 13 and makes the dead time in the rough estimation period wider than the dead time in the normal estimation period. The control device 124 gradually narrows the dead time so that the measured q-axis current value I _δr becomes the q-axis current command value I _δ during the rough estimation period. Based on the changed dead time, the angular velocity of the magnetic poles of the rotor of the synchronous generator 121 is estimated. Further, the angular position of the magnetic pole is estimated based on the estimated angular velocity and the actually measured d-axis current value _Iδ .

As shown in FIG. 20, at the start of control, the dead time is the initial dead time d0, which becomes narrower based on the deviation between the q-axis current command value I _δ and the measured q-axis current value I _δr, and becomes the induced voltage Em. It is adjusted to a constant value that depends on it. Since the initial dead time d0 is set to a wide value, the line current ia starts from a state where the current value is zero or very small. Then, after a lapse of time, the current starts to flow after the dead time becomes narrow to some extent. After the current flows, the absolute value of the current value increases with time. The angular velocity and angular position of the magnetic pole are estimated as the dead time changes.

  After the current value of the line-to-line current ia reaches a predetermined set current value, the dead time becomes a constant value. In this case, the estimation of the angular velocity and the angular position of the magnetic pole is performed with high accuracy. When the control device 124 determines that the dead time is constant for a predetermined time, the normal estimation period starts.

  When the normal estimation period starts, the dead time d is adjusted to the dead time dn in the normal estimation period. The magnetic pole angular velocity and angular position estimated in the rough estimation period are set as initial values, and estimation of the magnetic pole angular velocity and angular position in the normal estimation period is started. Then, the converter circuit 122 is controlled so that the operation following the angular velocity and the angular position of the magnetic pole is performed.

  FIG. 21 is a timing chart showing temporal changes of the dead time d, the magnetic pole angular velocity ω, the magnetic pole angular position θ, and the line current ia in the fifth embodiment of the present invention. In the fifth embodiment of the present invention, the configuration of the control device 124 is similar to that of the third embodiment.

The control device 124 has control blocks corresponding to FIGS. 13 and 17, and makes the initial dead time d0 in the rough estimation period wider than the dead time dn in the normal estimation period. The control device 124 gradually narrows the dead time d as time elapses after starting the rough estimation period. Further, during the rough estimation period, the PWM voltage modulation rate is adjusted so that the measured q-axis current value I _δr becomes the q-axis current command value I _δ . Based on the changed modulation factor, the angular velocity of the magnetic poles of the rotor of the synchronous generator 121 is estimated. Further, the angular position of the magnetic pole is estimated based on the estimated angular velocity and the actually measured d-axis current value _Iδr .

  As shown in FIG. 21, at the start of control, the dead time d is the initial dead time d0, which is narrowed based on the passage of time and is adjusted to be equal to the dead time dn in the normal estimation period. Since the initial dead time d0 is set to a wide value, the line current ia starts from a state where the current value is zero or very small. Then, after a lapse of time, the current starts to flow after the dead time becomes narrow to some extent. After the current flows, the absolute value of the current value increases with time. The current value changes as the dead time increases, but is adjusted to a predetermined value by adjusting the PWM modulation rate. The estimation of the angular velocity and the angular position of the magnetic pole is estimated as the PWM modulation rate changes.

  After the dead time dw becomes equal to the dead time dn of the normal estimation period in the rough estimation period, the angular velocity and the angular position of the magnetic pole are accurately estimated when a predetermined period elapses. When the control device 124 determines that a predetermined period has elapsed since it became equal to the dead time, the process shifts to a normal estimation period.

  When the normal estimation period starts, the dead time d is adjusted to the dead time dn in the normal estimation period. The magnetic pole angular velocity and angular position estimated in the rough estimation period are set as initial values, and estimation of the magnetic pole angular velocity and angular position in the normal estimation period is started. Then, the converter circuit 122 is controlled so that the operation following the angular velocity and the angular position of the magnetic pole is performed.

  According to the fourth and fifth embodiments, the current flowing from the converter circuit 122 to the synchronous generator 121 is reduced in the initial period of control start, and the estimation of the angular velocity and the angular position of the magnetic pole in the initial period is accurate. Even if it does not, it can suppress that the converter circuit 122 and the synchronous generator 121 become overcurrent. In the normal estimation period, since the estimation accuracy has increased with the passage of time from the start of control, the converter circuit 122 and the synchronization can be synchronized even if the current flowing from the converter circuit 122 to the synchronous generator 121 is increased by narrowing the dead time. The generator 121 is prevented from being overcurrent. Thereby, the same effects as those of the first to third embodiments described above can be obtained.

  Each embodiment of the present invention described above is merely an example of the present invention, and the configuration can be changed within the scope of the invention. For example, the dead time may be widened in the rough estimation period, and the estimation method may be the same in the rough estimation period and the normal estimation period. In the second to fifth embodiments, the dead time dw is changed from a wide state to a narrow state in the rough estimation period, but may be a constant value. In this case, the dead time in the rough estimation period is set wider than at least the dead time in the normal estimation period. Further, the case where the estimation method is made different between the rough estimation period and the normal estimation period is not limited to the above-described method.

  In addition, the control unit determines that the normal estimation period starts when a predetermined time elapses after the estimation unit starts the estimation. However, the control unit is not limited to this, and the predetermined condition is satisfied. The start may be determined. For example, during the rough estimation period, the physical quantity used for estimation of the angular velocity and angle position of the magnetic pole, such as the current value flowing through the synchronous motor or the synchronous generator, dead time, voltage modulation rate, etc. You may determine the start of. Further, the start of the normal estimation period may be determined when the deviation between the estimated value and the actually measured value reaches a predetermined range.

  In the first embodiment, the dead time only needs to be shortened in the rough estimation period, and when the induced voltage Em is either positive or negative, the dead time is adjusted so that the line current ia becomes either positive or negative. It does not have to be done. Further, when the first on-period W11 and the second on-period W12 are different, the first dead time dw1 and the second dead time dw2 may not be set to be the same.

  The above-described electronic circuit and block diagram are examples of the present invention, and may be appropriately changed as long as the same effect can be obtained. Although the full bridge type is used as the PWM inverter / converter circuit, a half bridge type can be realized in the same manner, and any so-called self-excited inverter / converter circuit may be used. Further, the calculation contents of the calculators 61 and 65 shown in FIGS. 12 and 17 are not limited to this, and may have other configurations. Further, in the present embodiment, an example in which the present invention is applied to a vehicle ventilation device and a wind power generation device has been described, but the present invention can be applied to all control devices that start control of a synchronous machine in a rotated state. The synchronous machine includes a synchronous motor and a synchronous generator. In the present embodiment, the control unit starts control of the synchronous machine via the inverter / converter circuit during the normal estimation period. However, even in the rough estimation period, based on the estimation result of the estimation unit, the inverter / converter The synchronous machine may be controlled through a circuit.

  The above-described method for estimating the angular velocity and angular position of the rotor of the synchronous machine, the control method for the synchronous machine, and the control method for controlling the inverter / converter circuit used for controlling the synchronous machine are also included in the present invention. The present invention also includes a program for causing the control device to execute these methods and a storage medium storing the program.

1 is a block diagram showing an electric motor drive control system 20 according to a first embodiment of the present invention. 3 is a diagram illustrating an example of a circuit configuration of a PWM inverter circuit 22. FIG. It is a circuit diagram which shows a part of inverter circuit 22 as 22 A of single phase full bridge circuits. It is a timing chart which shows the on-off state of each switching element Tr in a normal estimation period. It is a timing chart which shows the on-off state of each switching element Tr in a rough estimation period. It is a timing chart which shows the time change of the on-off state of the switching element Tr in the rough estimation period, the induced voltage Em, the line voltage Ea, and the line current ia. It is a timing chart which shows the time change of the on-off state of the switching element Tr in the rough estimation period, the induced voltage Em, the line voltage Ea, and the line current ia. 6 is a timing chart showing the on / off state of a switching element Tr, the time variation of the line voltage Ea and the line current ia when the dead time dn is narrow. It is a timing chart which shows the time change of the ON / OFF state of switching element Tr and the line current ia in the rough presumption period in the modification of a 1st embodiment. It is a graph which shows the relationship between the dead time d, the induced voltage Em, and the line current ia. It is a graph for demonstrating the case where dead time d is made variable. It is a block diagram which shows the control block structure in the rough estimation period of the control apparatus 24 of 2nd Embodiment of this invention. It is a block diagram which shows the control block structure in the normal estimation period of the control apparatus. 3 is a flowchart showing the operation of a control device 24. It is a timing chart showing a time change of dead time d, magnetic pole angular velocity ω, magnetic pole angular position θ, and line current ia. 7 is a flowchart showing another operation of the control device 24. It is a block diagram which shows the control block structure in the rough estimation period of the control apparatus 24 in 3rd Embodiment of this invention. It is a timing chart showing a time change of dead time d, magnetic pole angular velocity ω, magnetic pole angular position θ, and line current ia. It is a block diagram which shows the electric power generation control system 120 which is 4th Embodiment of this invention. It is a timing chart which shows the time change of dead time d in 4th Embodiment, angular velocity (omega) of a magnetic pole, angular position (theta) of a magnetic pole, and line current ia.

It is a timing chart which shows the time change of the dead time d in 5th Embodiment of this invention, the angular velocity (omega) of a magnetic pole, the angular position (theta) of a magnetic pole, and the line current ia.

Explanation of symbols

DESCRIPTION OF SYMBOLS 20 Motor drive control system 21 Synchronous motor 22 Inverter circuit 24 Control apparatus 70,73 Control part 71,72 Estimation part 120 Power generation control system 121 Synchronous generator 122 Converter circuit

Claims (10)

A control device for a synchronous motor that rotationally drives a synchronous motor by applying AC power from an inverter circuit,
An estimation unit that estimates the angular velocity and angular position of the magnetic poles of the rotor of the synchronous motor based on the current flowing through the synchronous motor;
A control unit that provides a gate signal to the inverter circuit so that the angular velocity of the magnetic pole becomes a predetermined angular velocity based on the angular velocity and the angular position of the magnetic pole estimated by the estimation unit;
The control unit is configured such that the dead time in the inverter circuit when the estimation unit starts estimation is wider than the dead time in the inverter circuit during a normal estimation period after a predetermined condition is satisfied after the estimation unit starts estimation. A control apparatus for a synchronous motor, characterized in that:   The estimation unit is characterized in that the estimation method is different between a rough estimation period from the start of estimation until a predetermined condition is satisfied and a normal estimation period after the predetermined condition is satisfied after the estimation is started. The control apparatus for a synchronous motor according to claim 1. During the rough estimation period,
The control unit adjusts the dead time so that the current flowing through the synchronous motor becomes a predetermined set current value from a state in which the dead time is extended,
3. The synchronous motor control device according to claim 2, wherein the estimation unit estimates the angular velocity and the angular position of the magnetic pole based on the dead time. During the rough estimation period,
3. The synchronous motor control device according to claim 1, wherein the control unit changes the dead time from a state in which the dead time is widened so that the dead time is narrowed over time. A control device for a synchronous generator that rectifies AC power supplied from a synchronous generator by a converter circuit,
An estimation unit that estimates the angular velocity and the angular position of the magnetic poles of the rotor of the synchronous generator based on the current flowing through the synchronous generator;
A control unit that applies a gate signal to the converter circuit so that the operation of the converter circuit follows the angular velocity and angular position of the magnetic pole estimated by the estimation unit;
The control unit is configured such that the dead time in the converter circuit when the estimation unit starts estimation is wider than the dead time of the converter circuit in a normal estimation period after a predetermined condition is satisfied after the estimation unit starts estimation. A control device for a synchronous generator.   The estimation unit varies the estimation method between a rough estimation period from the start of estimation until a predetermined condition is satisfied, and a normal estimation period after the predetermined condition is satisfied after the estimation is started. 6. The control device for a synchronous generator according to claim 5, wherein: During the rough estimation period,
The control unit adjusts the dead time so that the current flowing from the synchronous generator becomes a predetermined set current value from the state in which the dead time is extended,
The synchronous generator control device according to claim 6, wherein the estimation unit estimates an angular velocity and an angular position of the magnetic pole based on a dead time. During the rough estimation period,
The control device for a synchronous generator according to claim 5 or 6, wherein the control unit changes the dead time from a state in which the dead time is widened so that the dead time is narrowed as time elapses. A method for controlling a synchronous motor that rotationally drives a synchronous motor by applying AC power from an inverter circuit,
Estimate the angular velocity and angular position of the magnetic poles of the rotor of the synchronous motor based on the current flowing through the synchronous motor, and establish a predetermined condition after starting the estimation of the dead time in the inverter circuit when the estimation starts After the dead time of the inverter circuit in the normal estimation period after
A control method for a synchronous motor, wherein a gate signal is given to an inverter circuit based on an estimation result. A method of controlling a synchronous generator that rectifies AC power supplied from a synchronous generator by a converter circuit,
Estimate the angular velocity and angular position of the magnetic pole of the rotor of the synchronous generator based on the measured current value flowing through the synchronous generator, and estimate the dead time in the converter circuit when the estimation is started. It is wider than the dead time of the converter circuit in the normal estimation period after the specified condition is satisfied,
A control method for a synchronous generator, wherein a gate signal is given to a converter circuit based on an estimation result.

Images