WO2021161412A1 - Inverter - Google Patents

Abstract

An inverter having a switching circuit, a control circuit, and a detector that detects the current flowing inside the inverter or the temperature of the inverter. The switching circuit has a first switching element connected between high-potential wiring and output wiring, a second switching element connected between intermediate-potential wiring and the output wiring, a third switching element connected in series to the second switching element between the intermediate-potential wiring and the output wiring, and a fourth switching element connected between the output wiring and low-potential wiring. The inverter reduces the proportion of the periods in which the control circuit controls the switching circuit to the intermediate-potential output state to a greater extent when the detection value of the detector is higher than a reference value than when the detection value is lower than the reference value.

Description

Inverter

The technology disclosed in this specification relates to an inverter.

An inverter is disclosed in Japanese Patent Application Laid-Open No. 2017-093039. This inverter has a high-potential wiring, a low-potential wiring, a medium-potential wiring, and a switching circuit. The switching circuit includes an output wiring, a first switching element, a second switching element, a third switching element, a fourth switching element, a first intermediate diode, and a second intermediate diode. The first switching element is connected between the high potential wiring and the output wiring. The second switching element is connected between the medium potential wiring and the output wiring. The third switching element is connected in series with respect to the third switching element between the medium potential wiring and the output wiring. The fourth switching element is connected between the output wiring and the low potential wiring. The first intermediate diode is connected in parallel to the second switching element with the cathode facing the medium potential wiring side. The second intermediate diode is connected in parallel to the third switching element with the cathode facing the output wiring side. When the first switching element is turned on, a high potential is applied to the output wiring. When the second switching element and the third switching element are turned on, a medium potential is applied to the output wiring. When the fourth switching element is turned on, a low potential is applied to the output wiring. That is, this inverter can control the potential of the output wiring at three levels of high potential, medium potential, and low potential.

When the potential of the output wiring is controlled at three levels as in the inverter of Patent Document 1, the ripple current included in the output current becomes small, and the loss caused by the load (for example, the motor) connected to the inverter can be suppressed. can.

In the inverter that controls the potential of the output wiring at three levels as in Patent Document 1, the switching circuit has four switching elements (first to fourth switching elements). Therefore, the size of the inverter becomes large. This specification proposes a technique for miniaturizing an inverter that controls the potential of the output wiring at three levels.

The first inverter disclosed in the present specification applies a high potential wiring to which a high potential is applied, a low potential wiring to which a low potential is applied, and a medium potential lower than the high potential and higher than the low potential. It has a medium potential wiring to be used, a switching circuit, a detector for detecting the current flowing in the inverter or the temperature of the inverter, and a control circuit. The switching circuit includes an output wiring, a first switching element connected between the high potential wiring and the output wiring, and a second switching element connected between the medium potential wiring and the output wiring. A third switching element connected in series with the second switching element between the medium potential wiring and the output wiring, and a fourth switching connected between the output wiring and the low potential wiring. The element, the first intermediate diode connected in parallel to the second switching element with the cathode facing the medium potential wiring side, and the third switching element with the cathode facing the output wiring side. It has a second intermediate diode, which is connected in parallel to the other. The current capacities of the second switching element and the third switching element are lower than the current capacities of the first switching element and the fourth switching element. The control circuit is configured to control the potentials of the gates of the first switching element, the second switching element, the third switching element, and the fourth switching element. The control circuit turns on the switching circuit in a high potential output state in which the first switching element is turned on and the high potential is applied to the output wiring, and the second switching element and the third switching element are turned on. It is configured to change between the medium potential output state in which the medium potential is applied to the output wiring and the low potential output state in which the fourth switching element is turned on and the low potential is applied to the output wiring. There is. The control circuit is configured to generate an alternating current in the output wiring by changing the switching circuit between the high potential output state, the medium potential output state, and the low potential output state. When the detection value of the detector is larger than the reference value, the control circuit lowers the ratio of the period for controlling the switching circuit to the medium potential output state as compared with the case where the detection value is smaller than the reference value.

In this inverter, the current capacities of the second switching element and the third switching element are lower than the current capacities of the first switching element and the fourth switching element. Therefore, the second switching element and the third switching element can be miniaturized. As a result, the inverter can be miniaturized.

Further, in this inverter, when the value detected by the detector (that is, the current flowing in the inverter or the temperature of the inverter) is smaller than the reference value, the switching circuit has a medium potential than when the detected value is larger than the reference value. The ratio of the period for controlling the output state is high. That is, the ratio of the period during which the second switching element and the third switching element are turned on is high. When the value detected by the detector is low, the operating conditions of each switching element are not so strict. Therefore, even if the ratio of the period during which the second switching element and the third switching element are turned on is increased, so high stress is not applied to the second switching element and the third switching element. Further, by increasing the ratio of the period for controlling the medium potential output state in this way, the ripple current generated in the output current is suppressed, and the loss generated in the load connected to the inverter is reduced.

Further, in this inverter, when the value detected by the detector (that is, the current flowing in the inverter or the temperature of the inverter) is higher than the reference value, the switching circuit is placed at a medium potential than when the detected value is lower than the reference value. The ratio of the period for controlling the output state is low. That is, the ratio of the period during which the second switching element and the third switching element are turned on is low. In this case, the period for controlling the switching circuit to the medium potential output state may be zero. When the value detected by the detector is high, the operating environment for each switching element is severe. In particular, the operating environment is severe for the second switching element and the third switching element having a low current capacity. In this case, the control circuit reduces the ratio of the period during which the second switching element and the third switching element are turned on, thereby reducing the stress applied to the second switching element and the third switching element.

As described above, when the operating environment is not so severe, this inverter increases the ratio of the period during which the second switching element and the third switching element are turned on to suppress the ripple current and reduce the loss. Further, when the operating environment is severe, this inverter protects the second switching element and the third switching element by lowering the ratio of the period during which the second switching element and the third switching element are turned on. Therefore, the stress applied to the second switching element and the third switching element is reduced. Therefore, even if the current capacities of the second switching element and the third switching element are lowered, the inverter can operate when the operating environment is severe. By lowering the current capacities of the second switching element and the third switching element, the inverter can be miniaturized. When the ratio of the period during which the second switching element and the third switching element are turned on is lowered, the ripple current becomes large and the loss generated becomes large. However, since the frequency of operation of the inverter under severe operating conditions is not so high, the effect of the increase in loss due to the ripple current is not so large. Therefore, according to the configuration of this inverter, it is possible to miniaturize the inverter while enabling operation with substantially the same low loss as the conventional one.

The additional features of the first inverter are listed below. In addition, each feature listed below is useful independently.

(Feature 1) The inverter may have three switching circuits. The control circuit is configured to control the potentials of the gates of the first switching element, the second switching element, the third switching element, and the fourth switching element of each of the three switching circuits. You may. The control circuit changes the three switching circuits between the high potential output state, the medium potential output state, and the low potential output state, so that the three phases are between the output wirings of the three switching circuits. It may be configured to generate an AC current. For each of the three switching circuits, when the detection value of the detector is larger than the reference value, the control circuit controls the medium potential output state more than when the detection value is smaller than the reference value. The ratio may be lowered.

According to this configuration, a three-phase inverter can be configured.

(Feature 2) When the detected value is larger than the reference value, the control circuit changes between the high potential output state and the low potential output state without setting the switching circuit in the medium potential output state. You may let me.

According to this configuration, stress applied to the second switching element and the third switching element can be almost eliminated when the operating conditions are severe.

Instead of feature 2, the first inverter may have feature 3.

(Feature 3) When the detected value is larger than the reference value, the control circuit changes the switching circuit between the high potential output state and the low potential output state, and in this case, the high potential When changing from the output state to the low potential output state, the high potential output state is changed to the low potential output state via the medium potential output state, and the low potential output state is changed to the high potential output state. Occasionally, the low potential output state may be changed to the high potential output state via the medium potential output state. When the detected value is larger than the reference value, the control circuit controls the ratio of the period for controlling the medium potential output state to the high potential output state and the ratio for controlling the low potential output state. It may be lower than the ratio of.

According to this configuration, the ripple current can be suppressed to some extent by turning on the second switching element and the third switching element for a short time even when the operating conditions are severe.

(Feature 4) The first switching element may be made of a silicon semiconductor. The second switching element may be made of a compound semiconductor. The third switching element may be made of a compound semiconductor. The fourth switching element may be made of a silicon semiconductor.

Cost can be suppressed by configuring the first switching element and the fourth switching element, which have a large current capacity and are large, with an inexpensive silicon semiconductor. Even if the second switching element and the third switching element, which have a small current capacity and are small in size, are made of an expensive compound semiconductor, the cost is not so high. Further, by forming the second switching element and the third switching element with the compound semiconductor, the loss generated can be effectively reduced.

(Feature 5) The first intermediate diode may be made of a compound semiconductor. The second intermediate diode may be composed of a compound semiconductor.

According to this configuration, the loss that occurs can be reduced more effectively.

The second inverter disclosed in the present specification will be described below.

A three-phase inverter can be configured by providing three phases of switching circuits in which the second switching element and the third switching element are connected in series between the medium potential wiring and the output wiring as in Patent Document 1. In such an inverter, the second switching element and the third switching element may be short-circuited and failed. In this case, the switching circuit having the short-circuit fault element cannot output a specific potential to the output wiring. For example, when the second switching element is short-circuited and the first switching element is turned on, a line short circuit occurs between the high-potential wiring and the medium-potential wiring. Therefore, in this case, the switching circuit cannot output a high potential to the output wiring. Further, when the third switching element fails due to a short circuit, when the fourth switching element is turned on, a line short circuit occurs between the medium potential wiring and the low potential wiring. Therefore, in this case, the switching circuit cannot output a low potential to the output wiring.

Even if a short-circuit faulty element occurs, there are cases where you want to generate a three-phase alternating current with the inverter. In such a case, it is conceivable to prohibit the potential that cannot be output and change the potential of each output wiring with the remaining two potentials.

As described above, if the second switching element fails due to a short circuit, a high potential cannot be output to the output wiring. In this case, the potentials of the three output wirings can be changed between the medium potential and the low potential to generate a three-phase alternating current. However, in this case, the inverter cannot continuously generate the three-phase alternating current. That is, in this operation, since the electric charge stored in the lower capacitor is continuously used, the electric charge of the lower capacitor becomes extremely small after a certain period of time, and the medium potential becomes extremely low. When the medium potential becomes extremely low in this way, it is not possible to properly generate a three-phase alternating current.

As described above, if the third switching element fails due to a short circuit, a low potential cannot be output to the output wiring. In this case, the potentials of the three output wirings can be changed between the high potential and the medium potential to generate a three-phase alternating current. However, in this case, the inverter cannot continuously generate the three-phase alternating current. That is, in this operation, since the electric charge stored in the upper capacitor is continuously used, the electric charge of the upper capacitor becomes extremely small and the medium potential becomes extremely high after a certain period of time has elapsed. When the medium potential becomes extremely high in this way, it is not possible to properly generate a three-phase alternating current.

As described above, with the above-mentioned technique, when the second switching element or the third switching element has a short-circuit failure, it is not possible to continuously generate a three-phase alternating current. In the present specification, an inverter capable of continuously generating a three-phase alternating current in such a case is proposed as a second inverter.

The second inverter disclosed in the present specification includes a high-potential wiring to which a high potential is applied, a low-potential wiring to which a low potential is applied, a medium-potential wiring, and between the high-potential wiring and the medium-potential wiring. The upper capacitor connected to, the lower capacitor connected between the medium-potential wiring and the low-potential wiring, and three switching circuits of a U-phase switching circuit, a V-phase switching circuit, and a W-phase switching circuit. , Has a control circuit. Each of the three switching circuits is connected between the output wiring, the first switching element connected between the high potential wiring and the output wiring, and the medium potential wiring and the output wiring. The two switching elements, the third switching element connected in series with respect to the second switching element between the medium potential wiring and the output wiring, and connected between the output wiring and the low potential wiring. The fourth switching element, the first intermediate diode connected in parallel to the second switching element with the cathode facing the medium potential wiring side, and the first intermediate diode with the cathode facing the output wiring side. It has a second intermediate diode connected in parallel to the three switching elements. The control circuit is configured to control the potentials of the gates of the first switching element, the second switching element, the third switching element, and the fourth switching element of the three switching circuits. The control circuit is in a high potential output state in which the three switching circuits are turned on and the high potential is applied to the corresponding output wiring, the second switching element and the third switching element. Is turned on to apply the medium potential, which is the potential of the medium potential wiring, to the corresponding output wiring, and the fourth switching element is turned on to apply the low potential to the corresponding output wiring. It is configured to vary between low potential output states. The control circuit is the U-phase output wiring which is the output wiring of the U-phase switching circuit, the V-phase output wiring which is the output wiring of the V-phase switching circuit, and the output wiring of the W-phase switching circuit. By changing the respective potentials of the W-phase output wiring between the high potential, the medium potential, and the low potential, the U-phase output wiring, the V-phase output wiring, and the W-phase output wiring It is configured to generate a three-phase AC current between them. The control circuit can execute an emergency operation when any one of the second switching element and the third switching element of the three switching circuits fails in a short circuit. The short-circuit-failed element is called a short-circuit-failed element. The output wiring of one of the three switching circuits including the short-circuit fault element is referred to as a limited output wiring. The output wirings of the two switching circuits that do not include the short-circuit fault element of the three switching circuits are referred to as normal output wirings, respectively. In the emergency operation, the control circuit changes the potential of the limited output wiring between the two potentials other than the prohibited potential among the high potential, the medium potential, and the low potential, and the normal output. Each potential of the wiring is changed between the three potentials of the high potential, the medium potential, and the low potential. When the short-circuit failure element is the second switching element, the prohibited potential is the low potential. When the short-circuit failure element is the third switching element, the prohibited potential is the high potential.

In this inverter, during emergency operation, the output wiring (limited output wiring) of the switching circuit including the short-circuit fault element is changed between the two potentials excluding the prohibited potential, while the normal output wiring is changed between the three potentials. Let me. That is, the limited output wiring is controlled at two levels, while the normal output wiring is controlled at three levels. In this way, when the normal output wiring is controlled at three levels, either the upper capacitor or the lower capacitor is not discharged unilaterally, thus preventing an excessive rise or fall of the medium potential. can do. Therefore, according to this inverter, even if the second switching element or the third switching element fails due to a short circuit, the three-phase alternating current can be continuously generated.

The additional features of the second inverter mentioned above are listed below. In addition, each feature listed below is useful independently.

(Feature 1) It may further have a command circuit that generates a command value of an angle of the voltage vector so that the voltage vector rotates and inputs it to the control circuit. The voltage vector is a vector represented by the parameters Vu, Vv, Vw. The parameter Vu is a value indicating whether the potential of the U-phase output wiring is the high potential, the medium potential, or the low potential. The parameter Vv is a value indicating whether the potential of the V-phase output wiring is the high potential, the medium potential, or the low potential. The parameter Vw is a value indicating whether the potential of the W-phase output wiring is the high potential, the medium potential, or the low potential. In the emergency operation, the voltage vector at which the limited output wiring has the prohibited potential is the prohibited vector, and the angle range including the prohibited vector in the angle range of the voltage vector is the restricted angle range. The angle range outside the limit angle range of the angle range of the voltage vector is the normal angle range. In the emergency operation, when the angle indicated by the command value is within the limiting angle range, the control circuit has the angle indicated by the command value and the three said according to an allowable vector which is not the prohibited vector. The potential of the output wiring may be controlled. In the emergency operation, when the angle indicated by the command value is within the normal angle range, the control circuit selects a specific voltage vector from a plurality of voltage vectors having the angle indicated by the command value. , The potentials of the three output wirings may be controlled according to the selected specific voltage vector. The control circuit may select the voltage vector that raises the medium potential as the specific voltage vector when the medium potential is lower than the reference value. The control circuit may select the voltage vector for lowering the medium potential as the specific voltage vector when the medium potential is higher than the reference value.

In a 3-level inverter, the output voltage may be represented by a voltage vector. In this case, there are a plurality of voltage vectors having the same angle. For example, the voltage vectors (Vu, Vv, Vw) having an angle of 60 degrees with respect to the U phase include (2,2,0), (2,2,1), and (1,1,0). There is one. Here, as parameters Vu, Vv, and Vw, the numerical value "2" means a high potential, the numerical value "1" means a medium potential, and the numerical value "0" means a low potential.

When the angle indicated by the command value is within the limit angle range, the control circuit controls the potential of the output wiring according to the allowable vector (voltage vector other than the prohibition vector) among the plurality of voltage vectors having that angle. For example, when the high potential is the prohibited potential, (2,2,0), (2,2,0) among (2,2,0), (2,2,1), (1,1,0) described above 2,2,1) are prohibited vectors. In this case, since (1,1,0) is an allowable vector (usable voltage vector), the control circuit controls the potential of each output wiring according to (1,1,0). As described above, when the angle indicated by the command value is within the limit angle range, the control circuit prohibits the use of the prohibited potential by controlling according to the permissible vector.

When the angle range indicated by the command value is within the normal angle range, the control circuit selects a specific voltage vector from a plurality of voltage vectors having that angle, and sets the potential of the output wiring according to the selected specific voltage vector. Control. Here, the control circuit selects a voltage vector that raises the medium potential when the medium potential is lower than the reference value as a specific voltage vector, and selects a voltage vector that lowers the medium potential when the medium potential is higher than the reference value. Select as a specific voltage vector. For example, if the angles (ie, 60 degrees) of (2,2,0), (2,2,1), (1,1,0) described above are within the normal angle range, the control circuit A specific voltage vector is selected from these voltage vectors. Of these voltage vectors, (2,2,1) and (1,1,0) are voltage vectors that change the medium potential. The control circuit operates so as to bring the medium potential closer to the reference value by selecting either (2,2,1) or (1,1,0) according to the operating state of the inverter. In this way, when the angle range indicated by the command value is within the normal angle range, there are multiple voltage vectors that can be used. Therefore, by selecting an appropriate voltage vector from among them, the medium potential can be set to an appropriate value. Can be controlled to. Therefore, the three-phase alternating current can be continuously generated.

(Feature 2) The three output wirings may be configured to be connected to a load. The voltage vector in which the high potential and the medium potential are applied to the load and the low potential is not applied is referred to as an upper vector. The voltage vector in which the medium potential and the low potential are applied to the load and the high potential is not applied is referred to as a lower vector. When the angle indicated by the command value is within the normal angle range, the control circuit may select the specific voltage vector according to the following conditions A to D. A. When the medium potential is lower than the reference value and the current flowing through the load is in the forward direction with respect to the voltage applied to the load, the upper vector is selected as the specific voltage vector. B. When the medium potential is lower than the reference value and the current flowing through the load is in the opposite direction to the voltage applied to the load, the lower vector is selected as the specific voltage vector. .. C. The lower vector is selected as the specific voltage vector when the medium potential is higher than the reference value and the current flowing through the load is in the forward direction with respect to the voltage applied to the load. .. D. When the medium potential is higher than the reference value and the current flowing through the load is in the opposite direction to the voltage applied to the load, the upper vector is selected as the specific voltage vector.

According to this configuration, fluctuations in the medium potential can be further suppressed.

(Feature 3) In the control circuit, the middle potential is higher than the upper limit value higher than the reference value, the angle indicated by the command value is within the limit angle range, and the allowable vector sets the middle potential. In the case of the voltage vector to be raised, the three output wirings may be controlled to have the same potential regardless of the angle indicated by the command value.

In this way, when the difference between the medium potential and the reference value is large, the medium potential can be quickly returned to a value close to the reference value by setting the output wiring to the same potential under the above conditions.

(Feature 4) In the control circuit, the medium potential is lower than the lower limit value lower than the reference value, the angle indicated by the command value is within the limit angle range, and the allowable vector is the medium potential. In the case of the voltage vector that lowers the voltage, the three output wirings may be controlled to have the same potential regardless of the angle indicated by the command value.

In this way, when the difference between the medium potential and the reference value is large, the medium potential can be quickly returned to a value close to the reference value by setting the output wiring to the same potential under the above conditions.

Instead of features 1 to 4, the second inverter may have feature 5.

(Feature 5) A command circuit that generates a command value of an angle of the voltage vector so that the voltage vector rotates and inputs the command value to the control circuit may be further provided. The voltage vector is a vector represented by the parameters Vu, Vv, Vw. The parameter Vu is a value indicating whether the potential of the U-phase output wiring is the high potential, the medium potential, or the low potential. The parameter Vv is a value indicating whether the potential of the V-phase output wiring is the high potential, the medium potential, or the low potential. The parameter Vw is a value indicating whether the potential of the W-phase output wiring is the high potential, the medium potential, or the low potential. In the emergency operation, the voltage vector at which the limited output wiring has the prohibited potential is the prohibited vector, and the angle range including the prohibited vector in the angle range of the voltage vector is the restricted angle range. The angle range outside the limit angle range of the angle range of the voltage vector is the normal angle range. In the emergency operation, when the angle indicated by the command value is within the limiting angle range, the control circuit has the angle indicated by the command value and the three said according to an allowable vector which is not the prohibited vector. The potential of the output wiring may be controlled. In the emergency operation, when the angle indicated by the command value is within the normal angle range, the control circuit selects a specific voltage vector from a plurality of voltage vectors having the angle indicated by the command value. , The potentials of the three output wirings may be controlled according to the selected specific voltage vector. The three output wires may be configured to be connected to the load. The specific voltage vector is composed of a first group composed of a group of voltage vectors in which the high potential and the medium potential are applied to the load and the low potential is not applied to the load, and the medium potential and the low potential are applied to the load. May be selected from any of the second groups composed of the group of voltage vectors in which the high potential is applied and the high potential is not applied. The control circuit may store the control target value of the medium potential. In the control circuit, when the deviation between the medium potential and the control target value increases after the previous control phase, the voltage vector selected in the previous control phase between the first group and the second group can be used. The specific voltage vector may be selected from a group different from the group to which the specific voltage vector belongs.

Even with this configuration, it is possible to continuously generate three-phase alternating current while controlling the medium potential to an appropriate value.

(Feature 6) The control circuit is configured to detect currents flowing through the first switching element, the second switching element, the third switching element, and the fourth switching element of the three switching circuits. You may be. The control circuit may be capable of performing a short-circuit element determination operation on a selective switching circuit selected from the three switching circuits. In the short-circuit element determination operation, the control circuit turns on the selection switching circuit in the first state in which the first switching element and the second switching element are turned on, and in the third state and the fourth switching element. It may be changed over time between the second states. The control circuit determines that the second switching element is a short-circuit fault element when a short-circuit current flows through the third switching element in the second state with respect to the selective switching circuit, and in the first state. When a short-circuit current flows through the second switching element, it may be determined that the third switching element is a short-circuit failure element.

The number of selective switching circuits may be one, two, or three. That is, the short-circuit element determination operation may be executed for two or three selective switching circuits at the same time.

According to this configuration, the short-circuit failure element can be specified.

The circuit diagram of the inverter of Examples 1 to 4. A table showing a high potential output state, a medium potential output state, and a low potential output state. A space vector diagram showing a voltage vector. The figure which shows the voltage vector E2, E3. The graph which shows the potential V60u in three-level operation. The graph which shows the potential V60u in two-level operation. The graph which shows the potential V60u, V60v, V60w in the three-level operation. The graph which shows the potential V60u, V60v, V60w in the three-level operation. The graph which shows the potential V60u, V60v, V60w in the two-level operation. The graph which shows the potential V60u, V60v, V60w in the two-level operation. Graph showing ripple current. The graph which shows the potential V60u in the quasi-two-level operation. The graph which shows the potential V60u, V60v, V60w in the quasi-two-level operation. The graph which shows the potential V60u, V60v, V60w in the quasi-two-level operation. A circuit diagram showing a single-phase inverter. A graph showing the angle of the output voltage vector and the three-phase alternating current. The circuit diagram which shows the current path when (0,0,2) is output. The circuit diagram which shows the current path when (1, 1, 2) is output. The circuit diagram which shows the current path when (0,0,1) is output. The flowchart which shows the short-circuit element determination operation. A table showing the first state, the second state, and the third state. The circuit diagram which shows the path of the short circuit current in the 1st state. The circuit diagram which shows the path of the short circuit current in the 2nd state. The circuit diagram which shows the path of the short circuit current in the 3rd state. The flowchart which shows the emergency operation when the forbidden potential is a low potential VL. A space vector diagram showing an example of a limiting angle range when the forbidden potential is a low potential VL. A table showing the first to fourth rules. The figure which shows an example of the voltage vector which is output by 1st to 4th rule. The flowchart which shows the emergency operation when the prohibition potential is a high potential VH. A space vector diagram showing an example of a limiting angle range when the forbidden potential is a high potential VH. A table showing the 5th to 8th rules. The flowchart which shows the process which the inverter of Example 4 executes. The circuit diagram which shows the modification of Examples 1 to 4.

(Example 1) The inverter of the first embodiment will be described below. Examples 1 and 2 are examples of the first inverter described above.

(Inverter configuration)
FIG. 1 shows a circuit diagram of the inverter 10 of the first embodiment. The inverter 10 is mounted on the vehicle. Further, the vehicle is equipped with a battery 18 and a motor 90. The motor 90 is a three-phase motor and drives the wheels of the vehicle. The inverter 10 is connected to the battery 18 and the motor 90. The inverter 10 converts the DC power supplied from the battery 18 into three-phase AC power, and supplies the three-phase AC power to the motor 90. As a result, the motor 90 is driven and the vehicle travels.

The inverter 10 has a high potential wiring 12, a medium potential wiring 14, a low potential wiring 16, an upper capacitor 20, and a lower capacitor 22. The high potential wiring 12 is connected to the positive electrode of the battery 18. The low potential wiring 16 is connected to the negative electrode of the battery 18. A DC voltage is applied between the high-potential wiring 12 and the low-potential wiring 16 by the battery 18. Therefore, the potential VH of the high-potential wiring 12 is higher than the potential VL (that is, 0V) of the low-potential wiring 16. The upper capacitor 20 is connected between the high-potential wiring 12 and the medium-potential wiring 14. The lower capacitor 22 is connected between the medium potential wiring 14 and the low potential wiring 16. Therefore, the potential VM of the medium-potential wiring 14 is higher than the potential VL of the low-potential wiring 16 and lower than the potential VH of the high-potential wiring 12. In this example, the potential VM is about 50% of the potential VH (ie, VH / 2).

The inverter 10 has three switching circuits 30 of a U-phase switching circuit 30u, a V-phase switching circuit 30v, and a W-phase switching circuit 30w. Each of the switching circuits 30 is connected between the high-potential wiring 12, the low-potential wiring 16, and the medium-potential wiring 14. Each of the switching circuits 30 includes a first switching element 41, a second switching element 42, a third switching element 43, a fourth switching element 44, a first diode 51, a second diode 52, a third diode 53, and a third. It has 4 diodes 54 and an output wiring 60. Since the configurations of the three switching circuits 30 are equal to each other, the configuration of one switching circuit 30 will be described below.

The first switching element 41 is an IGBT (insulated gate bipolar transistor). The first switching element 41 is connected between the high potential wiring 12 and the output wiring 60. The collector of the first switching element 41 is connected to the high potential wiring 12, and the emitter of the first switching element 41 is connected to the output wiring 60.

The first diode 51 is connected in parallel to the first switching element 41. The anode of the first diode 51 is connected to the emitter of the first switching element 41, and the cathode of the first diode 51 is connected to the collector of the first switching element 41.

The fourth switching element 44 is an IGBT. The fourth switching element 44 is connected between the output wiring 60 and the low potential wiring 16. The collector of the fourth switching element 44 is connected to the output wiring 60, and the emitter of the fourth switching element 44 is connected to the low potential wiring 16.

The fourth diode 54 is connected in parallel to the fourth switching element 44. The anode of the fourth diode 54 is connected to the emitter of the fourth switching element 44, and the cathode of the fourth diode 54 is connected to the collector of the fourth switching element 44.

The second switching element 42 and the third switching element 43 are FETs (field effect transistors). The second switching element 42 and the third switching element 43 are connected in series between the output wiring 60 and the medium potential wiring 14. The drain of the second switching element 42 is connected to the medium potential wiring 14. The source of the second switching element 42 is connected to the source of the third switching element 43. The drain of the third switching element 43 is connected to the output wiring 60.

The second diode 52 is connected in parallel to the second switching element 42. The anode of the second diode 52 is connected to the source of the second switching element 42, and the cathode of the second diode 52 is connected to the drain of the second switching element 42. That is, the second diode 52 is connected in parallel to the second switching element 42 with the cathode facing the medium potential wiring 14 side.

The third diode 53 is connected in parallel to the third switching element 43. The anode of the third diode 53 is connected to the source of the third switching element 43, and the cathode of the third diode 53 is connected to the drain of the third switching element 43. That is, the third diode 53 is connected in parallel to the third switching element 43 with the cathode facing the output wiring 60 side.

The switching elements 41 and 44 are made of a silicon semiconductor. The switching elements 42 and 43 are composed of compound semiconductors (for example, silicon carbide semiconductors, gallium nitride semiconductors, etc.). Compound semiconductors have a wider bandgap than silicon semiconductors. The switching elements 42 and 43 made of compound semiconductors are less likely to cause loss than the switching elements 41 and 44 made of silicon semiconductors. Further, the current capacity of the switching elements 42 and 43 is smaller than the current capacity of the switching elements 41 and 44. That is, the maximum value of the current that can be passed through the switching elements 42 and 43 is smaller than the maximum value of the current that can be passed through the switching elements 41 and 44. Therefore, when a current of the same magnitude flows through the switching elements 42 and 43 and the switching elements 41 and 44, the temperature of the switching elements 42 and 43 is more likely to rise than that of the switching elements 41 and 44. Further, the switching elements 42 and 43 are smaller than the switching elements 41 and 44.

The diodes 51 to 54 may be pn diodes or Schottky barrier diodes. The diodes 51 and 54 are made of a silicon semiconductor. For example, the first diode 51 may be formed on a semiconductor substrate common to the first switching element 41 (that is, even if the first switching element 41 and the first diode 51 are formed in one semiconductor substrate). good). Further, the fourth diode 54 may be formed on a semiconductor substrate common to the fourth switching element 44. The diodes 52 and 53 are made of a compound semiconductor. For example, the second diode 52 may be formed on a semiconductor substrate common to the second switching element 42. The second diode 52 may be a parasitic diode incorporated in the second switching element 42 (that is, FET). Further, the third diode 53 may be formed on a semiconductor substrate common to the third switching element 43. The third diode 53 may be a parasitic diode incorporated in the third switching element 43 (that is, FET). The diodes 52 and 53 made of compound semiconductors are less likely to cause loss than the diodes 51 and 54 made of silicon semiconductors. The current capacity of the diodes 52 and 53 is smaller than the current capacity of the diodes 51 and 54. Therefore, when a current of the same magnitude flows through the diodes 52 and 53 and the diodes 51 and 54, the temperature of the diodes 52 and 53 is more likely to rise than that of the diodes 51 and 54. Further, the diodes 52 and 53 are smaller than the diodes 51 and 54.

In the following, the output wiring 60 of the U-phase switching circuit 30u is referred to as a U-phase output wiring 60u, the output wiring 60 of the V-phase switching circuit 30v is referred to as a V-phase output wiring 60v, and the output wiring 60 of the W-phase switching circuit 30w is referred to as W. It is called phase output wiring 60w. Each of the U-phase output wiring 60u, the V-phase output wiring 60v, and the W-phase output wiring 60w is connected to the motor 90.

The inverter 10 has a control circuit 70 and a command circuit 72. The command circuit 72 generates a command value according to the operating state of the motor 90, and inputs the generated command value to the control circuit 70. Although not shown, the control circuit 70 is connected to the gates of the switching elements 41 to 44 of each of the U-phase switching circuit 30u, the V-phase switching circuit 30v, and the W-phase switching circuit 30w. That is, the control circuit 70 is connected to the gates of the 12 switching elements shown in FIG. The control circuit 70 turns each switching element on and off based on the command value input from the command circuit 72. As a result, a three-phase alternating current is generated between the three output wirings 60. By supplying the three-phase alternating current to the motor 90, the motor 90 is driven and the vehicle travels.

Further, the inverter 10 has a current sensor 74 that detects the main current flowing in the inverter 10. The current detected by the current sensor 74 may be an input current supplied from the battery 18 to the inverter 10 (that is, a current flowing through the high potential wiring 12 or the low potential wiring 16), or is supplied from the inverter 10 to the motor 90. The current to be generated (that is, the current flowing through one or more of the output wirings 60u, 60v, 60w) or the main current flowing through any of the switching elements (for example, the collector of the switching elements 41, 44). -Emitor-to-emitter current, drain-source current of switching elements 42 and 43) may be used. The detected value of the current sensor 74 is input to the control circuit 70.

(Electric potential of output wiring)
Next, the potential applied to each output wiring 60 will be described. The control circuit 70 controls each switching circuit 30 to one of the high potential output state, the medium potential output state, and the low potential output state shown in FIG.

In the high potential output state, the first switching element 41 is controlled to be on, the second switching element 42 to be off, the third switching element 43 to be off, and the fourth switching element 44 to be off. In the high potential output state, the output wiring 60 is connected to the high potential wiring 12 via the first switching element 41. Therefore, in the high potential output state, the potential of the output wiring 60 is the same potential VH as that of the high potential wiring 12.

In the medium potential output state, the first switching element 41 is turned off, the second switching element 42 is turned on, the third switching element 43 is turned on, and the fourth switching element 44 is turned off. In the medium potential output state, the output wiring 60 is connected to the medium potential wiring 14 via the second switching element 42 and the third switching element 43. Therefore, in the medium potential output state, the potential of the output wiring 60 is the same medium potential VM as the medium potential wiring 14.

In the low potential output state, the first switching element 41 is turned off, the second switching element 42 is turned off, the third switching element 43 is turned off, and the fourth switching element 44 is turned on. In the low potential output state, the output wiring 60 is connected to the low potential wiring 16 via the fourth switching element 44. Therefore, in the low potential output state, the potential of the output wiring 60 is the same low potential VL as that of the low potential wiring 16.

By changing the state of each switching circuit 30 between the high potential output state, the medium potential output state, and the low potential output state, the potential of each output wiring 60 is among the high potential VH, the medium potential VM, and the low potential VL. Change. The control circuit 70 generates a three-phase alternating current in the output wiring 60 by controlling the potential of each output wiring 60.

(Voltage vector)
FIG. 3 is a space vector diagram showing a voltage vector showing a potential applied to each of the output wirings 60. In FIG. 3, the voltage vector E1 is illustrated. The voltage vector is represented by three parameters (Vu, Vv, Vw). The parameter Vu is a value indicating the potential of the U-phase output wiring 60u. The parameter Vv is a value indicating the potential of the V-phase output wiring 60v. The parameter Vw is a value indicating the potential of the W-phase output wiring 60w. The parameters Vu, Vv, Vw are numerical values between 0 and 2. The numerical value "0" indicates that the low potential VL is applied to the corresponding output wiring 60, the numerical value "1" indicates that the medium potential VM is applied to the corresponding output wiring 60, and the numerical value "2" corresponds. It is shown that the high potential VH is applied to the output wiring 60 to be processed. For example, since the voltage vector E1 illustrated in FIG. 3 is (2,2,0), a high potential VH is applied to the U-phase output wiring 60u, and a high potential VH is applied to the V-phase output wiring 60v. This means that a low potential VL is applied to the W-phase output wiring 60w.

The command circuit 72 generates command values of potentials to be applied to the three output wirings 60. The command circuit 72 generates a command value by a voltage vector (that is, three parameters (Vu, Vv, Vw)). The command value generated by the command circuit 72 is input to the control circuit 70. Hereinafter, the command value (voltage vector) input from the command circuit 72 to the control circuit 70 is referred to as a command value vector.

The control circuit 70 controls the inverter 10 according to the command value vector. For example, when the parameter Vu of the command value vector is "0", the control circuit 70 controls the U-phase switching circuit 30u to the low-potential output state and applies the low-potential VL to the U-phase output wiring 60u. When the parameter Vu of the command value vector is "1", the control circuit 70 controls the U-phase switching circuit 30u to the medium-potential output state and applies the medium-potential VM to the U-phase output wiring 60u. When the parameter Vu of the command value vector is "2", the control circuit 70 controls the U-phase switching circuit 30u to the high-potential output state and applies the high-potential VH to the U-phase output wiring 60u. When the parameter Vv of the command value vector is "0", the control circuit 70 controls the V-phase switching circuit 30v to the low-potential output state and applies the low-potential VL to the V-phase output wiring 60v. When the parameter Vv of the command value vector is "1", the control circuit 70 controls the V-phase switching circuit 30v to the medium-potential output state and applies the medium-potential VM to the V-phase output wiring 60v. When the parameter Vv of the command value vector is "2", the control circuit 70 controls the V-phase switching circuit 30v to the high-potential output state and applies the high-potential VH to the V-phase output wiring 60v. When the parameter Vw of the command value vector is "0", the control circuit 70 controls the W-phase switching circuit 30w to the low-potential output state and applies the low-potential VL to the W-phase output wiring 60w. When the parameter Vw of the command value vector is "1", the control circuit 70 controls the W-phase switching circuit 30w to the medium-potential output state and applies the medium-potential VM to the W-phase output wiring 60w. When the parameter Vw of the command value vector is "2", the control circuit 70 controls the W-phase switching circuit 30w to the high-potential output state and applies the high-potential VH to the W-phase output wiring 60w. In this way, the control circuit 70 controls the potentials of the three output wirings 60 according to the command value vector. In the following, controlling the potentials of the three output wirings 60 by the control circuit 70 may be referred to as “outputting a voltage vector”.

Further, the command circuit 72 may generate a command value vector including a decimal as the parameters Vu, Vv, and Vw. For example, the voltage vector E2 in FIG. 4 may be generated as a command value vector. When the parameters of the command value vector include a decimal number, the control circuit 70 synthesizes these voltage vectors by outputting a plurality of voltage vectors (voltage vectors whose parameters are integers) close to the command value vector with a time lag. .. The control circuit 70 controls the output voltage vector so that the combined voltage vector matches the command value vector.

The command circuit 72 sequentially generates a command value vector so that the command value vector rotates as shown by an arrow 102 in FIG. 3, and inputs the command value vector to the control circuit 70. The control circuit 70 outputs a voltage vector according to the input command value vector. Therefore, the output voltage vector rotates as shown by the arrow 102. As a result, a three-phase alternating current is generated between the three output wirings 60, and the magnetic field generated inside the motor 90 rotates. As a result, the rotor of the motor 90 rotates.

(Switching between 2-level operation and 3-level operation)
As described above, the current value detected by the current sensor 74 is input to the control circuit 70. The control circuit 70 switches the operation of the inverter 10 between the two-level operation and the three-level operation based on the input current value. When the current value detected by the current sensor 74 is smaller than the reference value, the control circuit 70 executes a three-level operation. In the three-level operation, the control circuit 70 controls the potential of each output wiring 60 at three levels of high potential VH, medium potential VM, and low potential VL. For example, FIG. 5 illustrates a change in the potential V60u of the U-phase output wiring 60u in a three-level operation. As shown in FIG. 5, in the three-level operation, the potential V60u changes among the three potentials VH, VM, and VL according to the target potential. Similarly, in the three-level operation, the potential V60v of the V-phase output wiring 60v and the potential V60w of the W-phase output wiring 60w also change between the three potentials VH, VM, and VL according to the target potential. Further, when the current value detected by the current sensor 74 is larger than the reference value, the control circuit 70 executes a two-level operation. In the two-level operation, the control circuit 70 controls the potential of each output wiring 60 at two levels of high potential VH and low potential VL. That is, in the two-level operation, the control circuit 70 does not output the medium potential VM to each output wiring 60. For example, FIG. 6 shows a change in the potential V60u of the U-phase output wiring 60u in the two-level operation. As shown in FIG. 6, in the two-level operation, the potential V60u changes between the high potential VH and the low potential VL according to the target potential. That is, in the two-level operation, the U-phase switching circuit 30u does not enter the medium potential output state. Similarly, in the two-level operation, the potential V60v of the V-phase output wiring 60v and the potential V60w of the W-phase output wiring 60w also change between the high potential VH and the low potential VL according to the target potential. The three-level operation and the two-level operation will be described in detail below.

(3 level operation)
The three-level operation will be described by taking the voltage vectors E2 and E3 of FIG. 4 as an example. For example, consider the case where the command value vector is the voltage vector E2 in FIG. In the three-level operation, the control circuit 70 shifts the three voltage vectors (0,0,0), (1,0,0), and (1,1,0) closest to the voltage vector E2 in time. And output. FIG. 7 shows the changes in the potentials V60u, V60v, and V60w in this case. As shown in FIG. 7, when (0,0,0), (1,0,0), and (1,1,0) are output with a time lag, the potential V60u and the potential V60v become medium potentials. It varies between VM and low potential VL. The potential V60w is maintained at the low potential VL.

Further, for example, consider the case where the command value vector is the voltage vector E3 in FIG. In the three-level operation, the control circuit 70 temporally shifts the three voltage vectors (2,2,1), (2,2,0), and (1,2,0) closest to the voltage vector E3. And output. FIG. 8 shows the changes in the potentials V60u, V60v, and V60w in this case. As shown in FIG. 8, when (2,2,1), (2,2,0), and (1,2,0) are output with a time lag, the potential V60u becomes a high potential VH and a medium potential. It changes between VMs, the potential V60v is maintained at high potential VH, and the potential V60w changes between medium potential VM and low potential VL.

As described above, in the three-level operation, the potentials V60u, V60v, and V60w can be the three potentials VH, VM, and VL. Therefore, as shown in FIG. 5, the potential V60u changes among the three potentials VH, VM, and VL. As shown in FIG. 5, in the three-level operation, the potential V60u changes between the high potential VH and the medium potential VM, or between the medium potential VM and the low potential VL. In other words, in the three-level operation, the potential V60u does not directly change from the high potential VH to the low potential VL, and does not directly change from the low potential VL to the high potential VH. Similarly, the potentials V60v and V60w also change among the three potentials VH, VM and VL.

(2 level operation)
The two-level operation will be described by taking the voltage vectors E2 and E3 of FIG. 4 as an example. For example, consider the case where the command value vector is the voltage vector E2 in FIG. In the two-level operation, the control circuit 70 does not control each switching circuit 30 to the medium potential output state. Therefore, in the two-level operation, the control circuit 70 is the three voltage vectors closest to the voltage vector E2 from the voltage vectors that do not include the parameter “1” (0,0,0), (2,0,0). ), (2,2,0) are selected, and these three voltage vectors are output with a time lag. FIG. 9 shows the changes in the potentials V60u, V60v, and V60w in this case. As shown in FIG. 9, when (0,0,0), (2,0,0), and (2,20) are output with a time lag, the potential V60u and the potential V60v become high potentials. It varies between VH and low potential VL. The potential V60w is maintained at the low potential VL.

Further, for example, consider the case where the command value vector is the voltage vector E3 in FIG. In the two-level operation, the control circuit 70 is the three voltage vectors closest to the voltage vector E3 among the voltage vectors not including the parameter “1” (0,0,0), (2,2,0). (0,2,0) is selected, and these three voltage vectors are output with a time lag. FIG. 10 shows the changes in the potentials V60u, V60v, and V60w in this case. As shown in FIG. 10, when the voltage vectors (0,0,0), (2,2,0), and (0,2,0) are output with a time lag, the potential V60u and the potential V60v Varies between high potential VH and low potential VL, and potential V60w is maintained at low potential VL.

As described above, in the two-level operation, the potentials V60u, V60v, and V60w can be the two potentials VH and VL, but are not controlled by the medium potential VM. Therefore, as shown in FIG. 6, the potential V60u changes between the two potentials VH and VL. As shown in FIG. 6, in the two-level operation, the potential V60u directly changes from the high potential VH to the low potential VL, and directly changes from the low potential VL to the high potential VH. Similarly, the potentials V60v and V60w also change between the two potentials VH and VL.

(Advantages of Inverter of Example 1)
FIG. 11 shows a current I (current flowing through any one of the output wirings 60u, 60v, and 60w) supplied from the output wiring 60 to the motor 90. As shown in FIG. 11, while the current I draws a sine curve, the current I is superposed with a ripple current that changes minutely at high frequencies. The ripple current is generated due to a change in the potential of the output wiring 60. As described above with reference to FIG. 6, in the two-level operation, the potential of each output wiring 60 changes directly between the high potential VH and the low potential VL. Therefore, the amount of change ΔV when the potential of each output wiring 60 changes is large. On the other hand, in the three-level operation, the potential of each output wiring 60 does not change directly between the high potential VH and the low potential VL. As shown in FIG. 5, in the three-level operation, the potential of each output wiring 60 changes between the high potential VH and the medium potential VM, or between the medium potential VM and the low potential VL. In the three-level operation, the amount of change ΔV when the potential of the output wiring 60 changes is smaller than that in the two-level operation. Therefore, in the three-level operation, the amplitude W of the ripple current (see FIG. 11) can be made smaller than that in the two-level operation. When the ripple current is suppressed in this way, the loss generated in the motor 90 is suppressed. That is, the loss generated in the motor 90 can be suppressed in the three-level operation as compared with the two-level operation.

Further, in the three-level operation, since the amount of change ΔV of the potential in each output wiring 60 is small, the amount of change in the voltage applied to each of the switching elements 41 to 44 during switching of each of the switching elements 41 to 44 is small. Therefore, in the three-level operation, the switching loss generated in each of the switching elements 41 to 44 is smaller than in the two-level operation. Further, when the amount of change in the voltage applied to each of the switching elements 41 to 44 is small, the amount of change in the voltage applied to each of the diodes 51 to 54 is also small. Therefore, in the three-level operation, the loss (recovery loss) generated when each diode 51 to 54 is in the recovery operation is smaller than that in the two-level operation.

Further, in the three-level operation, since the medium potential VM is output to each output wiring 60, a current flows through the switching elements 42 and 43. In the two-level operation, since the medium potential VM is not output to each output wiring 60, the switching elements 42 and 43 are kept off, and no current flows through the switching elements 42 and 43. The switching elements 42 and 43 are made of compound semiconductors, while the switching elements 41 and 44 are made of silicon semiconductors. Therefore, the loss that occurs when a current flows through the switching elements 42 and 43 is smaller than the loss that occurs when a current flows through the switching elements 41 and 44. Therefore, the three-level operation of switching the switching elements 42 and 43 is less likely to cause a loss than the two-level operation of keeping the switching elements 42 and 43 off. Similarly, the diodes 52 and 53 are made of compound semiconductors, while the diodes 51 and 54 are made of silicon semiconductors. Therefore, the loss that occurs when a current flows through the diodes 52 and 53 is smaller than the loss that occurs when a current flows through the diodes 51 and 54. Therefore, the three-level operation in which the current flows through the diodes 51 and 54 is less likely to cause a loss than the two-level operation in which the current does not flow through the diodes 51 and 54.

As described above, the loss generated in the inverter 10 and the motor 90 can be reduced in the three-level operation as compared with the two-level operation.

Further, as described above, the current does not flow through the switching elements 42 and 43 and the diodes 52 and 53 in the two-level operation, while the current flows through the switching elements 42 and 43 and the diodes 52 and 53 in the three-level operation. The switching elements 42 and 43 have a smaller current capacity than the switching elements 41 and 44, and the diodes 52 and 53 have a smaller current capacity than the diodes 51 and 54. Therefore, the two-level operation can operate with a larger current than the three-level operation. That is, in the two-level operation, a larger current can be supplied to the motor 90 than in the three-level operation, and the motor 90 can be operated with a higher torque.

As described above, in the 3-level operation, the loss can be reduced as compared with the 2-level operation, while in the 2-level operation, a larger current can be supplied to the motor 90 than in the 3-level operation. The control circuit 70 executes a three-level operation when the current value detected by the current sensor 74 is smaller than the reference value. Therefore, when the current value is smaller than the reference value, the loss can be reduced by the three-level operation. Further, the control circuit 70 executes a two-level operation when the current value detected by the current sensor 74 is larger than the reference value. This prevents a large current from flowing through the switching elements 42, 43 and the diodes 52, 53, which have a small current capacity. Further, according to the two-level operation, a large current can be supplied to the motor 90 by the switching elements 41 and 44 and the diodes 51 and 54 having a large current capacity. Therefore, the inverter 10 can operate even at a large current.

If the current capacities of the switching elements 42 and 43 and the diodes 52 and 53 are increased according to the maximum current flowing through the inverter 10, it is possible to perform three-level operation at a large current. However, in that case, the switching elements 42 and 43 and the diodes 52 and 53 become large, and the inverter 10 becomes large. On the other hand, as in the first embodiment, the current capacities of the switching elements 42, 43 and the diodes 52, 53 are reduced, and the switching elements 42, 43 and the diodes 52, 53 are operated by executing the two-level operation at the time of a large current. It is possible to operate the inverter 10 with a large current while protecting it from a large current. Further, this makes it possible to reduce the size of the inverter 10. Further, the frequency of operating the motor 90 with high torque is not so high, and the current flowing through the inverter 10 is not so high in the standard operating state. That is, in a standard operating state, the inverter 10 executes three-level operation, and the inverter 10 is operated only in a limited situation where a high torque is required by the motor 90 (for example, a situation where the tire of the vehicle is locked). Performs a two-level operation. Therefore, the effect of the loss increase due to the two-level operation is limited, and in most situations, the benefit of the loss reduction due to the three-level operation can be obtained. Therefore, according to the configuration of the inverter 10, the inverter 10 can be miniaturized with almost no increase in loss as compared with the conventional case.

Further, since the switching elements 42 and 43 and the diodes 52 and 53 are composed of the compound semiconductor as described above, the loss in the three-level operation can be further reduced. In particular, since the three-level operation is executed in the standard operating state, it is executed more frequently than the two-level operation. Therefore, the switching elements 42 and 43 and the diodes 52 and 53 made of the compound semiconductor can be used with high frequency, and the loss can be efficiently reduced. Also, compound semiconductors are generally expensive. However, since the switching elements 42 and 43 and the diodes 52 and 53 have a small current capacity and are small in size, even if an element composed of a compound semiconductor is adopted as the switching elements 42 and 43 and the diodes 52 and 53, the cost is so high. It doesn't become. As described above, by adopting an element composed of a compound semiconductor as the switching elements 42 and 43 and the diodes 52 and 53, the loss can be reduced at low cost.

(Example 2)
Next, the inverter of the second embodiment will be described. In the above-described inverter of the first embodiment, when the current value detected by the current sensor 74 is larger than the reference value, the two-level operation is executed. On the other hand, in the inverter of the second embodiment, when the current value detected by the current sensor 74 is larger than the reference value, the quasi-two level operation is executed. In other respects, the configuration of the inverter of the second embodiment is the same as the configuration of the inverter of the first embodiment. The quasi-two-level operation will be described below.

In the above-mentioned two-level operation, as shown in FIG. 6, the potential of each output wiring 60 is directly changed between the high potential VH and the low potential VL without controlling each switching circuit 30 to the medium potential output state. .. On the other hand, in the quasi-two-level operation, as shown in FIG. 12, the switching circuit 30 is instantaneously controlled to the medium potential output state while the potential of each output wiring 60 is changed from the low potential VL to the high potential VH. And output the medium potential VM. Similarly, in the quasi-two-level operation, while changing the potential of each output wiring 60 from the high potential VH to the low potential VL, the switching circuit 30 is instantaneously controlled to the medium potential output state to output the medium potential VM. ..

For example, consider the case where the command value vector is the voltage vector E2 in FIG. In the quasi-two-level operation, as in the two-level operation, the control circuit 70 is the three voltage vectors closest to the voltage vector E2 from among the voltage vectors not including the parameter "1" (0,0,0). (2,0,0) and (2,2,0) are selected, and these three voltage vectors are output with a time lag. However, in the quasi-two-level operation, while changing the output voltage vector, the voltage vector intermediate between the two voltage vectors before and after the change is instantaneously output. That is, as shown in FIG. 13, the control circuit 70 momentarily uses an intermediate voltage vector between them while changing the output voltage vector from (0,0,0) to (2,0,0). Output a certain (1,0,0). That is, the output voltage vector is changed in the order of (0,0,0), (1,0,0), (2,0,0). The period for outputting (1,0,0) is extremely shorter than the period for outputting (0,0,0) and (2,0,0). Similarly, the control circuit 70 momentarily has an intermediate voltage vector (2,1) while changing the output voltage vector from (2,0,0) to (2,2,0). , 0) is output. The period for outputting (2,1,0) is extremely shorter than the period for outputting (2,0,0) and (2,2,0). Similarly, the control circuit 70 is momentarily an intermediate voltage vector (1,1) while changing the output voltage vector from (2,2,0) to (0,0,0). , 0) is output. The period for outputting (1,1,0) is extremely shorter than the period for outputting (2,2,0) and (0,0,0). When the voltage vector changes in this way, as shown in FIG. 13, the medium potential VM is instantaneously output while the potential of each output wiring 60 is changing between the high potential VH and the low potential VL.

Further, for example, consider the case where the command value vector is the voltage vector E3 in FIG. In the quasi-two-level operation, as in the two-level operation, the control circuit 70 is the three voltage vectors closest to the voltage vector E3 among the voltage vectors not including the parameter “1” (0,0,0). (2,2,0) and (0,2,0) are selected, and these three voltage vectors are output with a time lag. However, in the quasi-two-level operation, while changing the output voltage vector, the voltage vector intermediate between the two voltage vectors before and after the change is instantaneously output. That is, as shown in FIG. 14, the control circuit 70 momentarily uses an intermediate voltage vector between them while changing the output voltage vector from (0,0,0) to (2,2,0). Output a certain (1,1,0). Further, the control circuit 70 momentarily has an intermediate voltage vector (1,2,0) while changing the output voltage vector from (2,2,0) to (0,2,0). ) Is output. Further, the control circuit 70 momentarily has an intermediate voltage vector (0,1,0) while changing the output voltage vector from (0,2,0) to (0,0,0). ) Is output. When the voltage vector changes in this way, as shown in FIG. 14, the medium potential VM is instantaneously output while the potential of each output wiring 60 is changing between the high potential VH and the low potential VL.

Since the voltage vector is output in this way, as shown in FIG. 12, in the quasi-two-level operation, the potential of the output wiring 60 is momentarily medium potential while changing between the high potential VH and the low potential VL. VM is output.

As described above, in the quasi-two-level operation, the medium potential VM is output to the output wiring 60 while the potential of each output wiring 60 is changing between the high potential VH and the low potential VL. Therefore, in the quasi-two-level operation, the rate of change when the potential of the output wiring 60 changes between the high potential VH and the low potential VL is slower than in the two-level operation. Therefore, in the quasi-two-level operation, the ripple current can be suppressed and the loss generated in the motor 90 can be suppressed as compared with the two-level operation. Further, in the quasi-two-level operation, the loss generated in each of the switching elements 41 to 44 can be reduced and the recovery loss generated in each diode 51 to 54 can be reduced as compared with the two-level operation. That is, in the quasi-two-level operation, the loss generated in the inverter and the motor 90 can be reduced as compared with the two-level operation. Further, in the quasi-two-level operation, the switching elements 42 and 43 are turned on during the period of outputting the medium potential VM, and a high current flows through the switching elements 42 and 43. However, in the quasi-two-level operation, the period for outputting the medium potential VM is extremely short. As is clear from comparing FIGS. 5 and 12, the period for outputting the medium potential VM (that is, the period during which the switching elements 42 and 43 are turned on) is shorter in the quasi-two-level operation than in the three-level operation. Therefore, it is possible to perform quasi-two-level operation without applying so high stress to the switching elements 42 and 43.

In Examples 1 and 2 described above, the medium potential VM was held at a potential lower than the high potential VH and higher than the low potential VL by the capacitors 20 and 22. However, if the potential of the medium potential wiring 14 can be controlled by another configuration, the capacitors 20 and 22 may be omitted. For example, the first battery may be connected between the high-potential wiring 12 and the medium-potential wiring 14, and the second battery may be connected between the medium-potential wiring 14 and the low-potential wiring 16.

In Examples 1 and 2 described above, the three-level operation and the two-level operation (or the quasi-two-level operation) are switched based on the current detected by the current sensor 74 (that is, the current flowing in the inverter 10). And executed. However, when a temperature sensor for detecting the temperature of the inverter 10 is provided, a three-level operation is executed when the temperature detected by the temperature sensor is lower than the reference value, and the temperature detected by the temperature sensor is higher than the reference value. Two-level operation (or quasi-two-level operation) may be performed. When a current is passed through the switching elements 42 and 43 while the temperature of the switching elements 42 and 43 is high, a high stress is applied to the switching elements 42 and 43. When the temperature of the inverter 10 is high, the current flowing through the switching elements 42 and 43 can be suppressed and the stress on the switching elements 42 and 43 can be suppressed by performing the two-level operation (or the quasi-two-level operation). .. Further, the temperature of the inverter 10 becomes high when a large current is flowing through the inverter 10, which is a limited situation in which a high torque is required for the motor 90. In the standard operating state, the temperature of the inverter 10 is not so high, and three-level operation is executed. Therefore, in most situations, you can benefit from the loss reduction of the three-level operation. The temperature sensor may detect the temperature at any position of the inverter 10, but if the temperature of the switching elements 42 and 43 is detected, the stress on the switching elements 42 and 43 can be suppressed more reliably. ..

Further, in Examples 1 and 2 described above, the switching elements 41 and 44 were IGBTs composed of silicon semiconductors. However, the switching elements 41 and 42 may be composed of FETs composed of compound semiconductors, similarly to the switching elements 42 and 43. Even in this case, the size of the inverter 10 can be reduced by making the current capacities of the switching elements 42 and 43 smaller than the current capacities of the switching elements 41 and 44.

Further, in Examples 1 and 2 described above, the switching elements 42 and 43 were FETs composed of compound semiconductors. However, the switching elements 42 and 43 may be switching elements (for example, IGBTs, FETs, etc.) made of silicon semiconductors. Even in this case, the inverter 10 can be miniaturized by making the current capacities of the switching elements 42 and 43 smaller than the current capacities of the switching elements 41 and 44.

Further, although the three-phase inverter has been described in Examples 1 and 2 described above, the techniques of Examples 1 and 2 may be applied to the single-layer inverter shown in FIG. 15, for example.

(Example 3) The inverter of the third embodiment will be described below. In addition, Example 3 is an Example of the above-mentioned second inverter.

(Structure of Inverter) The inverter of the third embodiment has the configuration of FIG. 1 as in the first and second embodiments. The inverter of the third embodiment is characterized by an emergency operation executed when the second switching element 42 or the third switching element 43 fails in a short circuit. In the normal state (that is, when the second switching element 42 and the third switching element 43 are not short-circuited), the operation of the inverter of the third embodiment is the same as the operation of the inverters of the first and second embodiments. It may be different or it may be different. The inverter of the third embodiment does not have to have the current sensor 74.

In the third embodiment, similarly to the first and second embodiments, the upper capacitor 20 is connected between the high potential wiring 12 and the medium potential wiring 14, and the lower capacitor 22 is the medium potential wiring 14 and the low potential wiring 16. It is connected in between. Therefore, the medium potential VM is higher than the low potential VL and lower than the high potential VH. The medium potential VM varies depending on the amount of electric charge stored in the upper capacitor 20 and the amount of electric charge stored in the lower capacitor 22. When the upper capacitor 20 is discharged or the lower capacitor 22 is charged, the medium potential VM rises. When the upper capacitor 20 is charged or the lower capacitor 22 is discharged, the medium potential VM is lowered.

Although not shown, in the third embodiment, the control circuit 70 and the command circuit 72 are connected to the medium potential wiring 14. The control circuit 70 and the command circuit 72 can detect the medium potential VM.

Although not shown, the inverter 10 has a current sensor that detects the current flowing through each output wiring 60. The detected value of each current sensor is input to the control circuit 70.

Also in the third embodiment, the command circuit 72 generates a command value vector. The command circuit 72 sequentially generates a command value vector so that the command value vector rotates as shown by an arrow 102 in FIG. 3, and inputs the command value vector to the control circuit 70. In normal operation, the control circuit 70 outputs a voltage vector according to a command value vector. In the following, as shown in FIG. 3, the angle of the voltage vector is shown by the angle θ with respect to the Vu axis. For example, the angle θ of (2,2,0) is 60 °. Since the command value vector is represented by (Vu, Vv, Vw), the command value vector contains information on the angle θ.

Note that there are multiple voltage vectors at a specific angle θ. For example, as shown in FIG. 3, when the angle θ is 60 °, there are three voltage vectors (2,2,0), (2,2,1), and (1,1,0). When (2,2,0) is output, the motor 90 can be operated with a higher torque than when (2,2,1) and (1,1,0) are output. The command circuit 72 sets (2,2,0) as a command value vector when the torque required by the motor 90 is high. Further, the command circuit 72 sets (2,2,1) or (1,1,0) as a command value vector when the torque required by the motor 90 is low. Further, when (2,2,1) is output, the upper capacitor 20 is charged or discharged. Further, when (1,1,0) is output, the lower capacitor 22 is charged or discharged. Therefore, when (2,2,1) or (1,1,0) is output, the medium potential VM fluctuates. The fluctuation of the medium potential VM will be described in detail later. The command circuit 72 detects the medium potential VM, and (2,2,1) or (1,1,0) so that the medium potential VM becomes a target value (for example, a value halved of the high potential VH). Select one of the above and use it as the command value vector. As described above, at the angle θ where a plurality of voltage vectors exist, the command circuit 72 selects one voltage vector from the plurality of voltage vectors and uses it as the command value vector.

Further, the command circuit 72 may be configured to generate a command value vector including a decimal as the parameters Vu, Vv, and Vw. For example, the voltage vector E2 of FIG. 4 may be generated as a command value vector. In this case, the control circuit 70 outputs a voltage vector close to the voltage vector E2 with a time lag. Further, the three voltage vectors (2,2,2), (1,1,1) and (0,0,0) are so-called zero vectors, and the three output wirings 60 have the same potential. means.

As described above, in the normal operation, the command circuit 72 rotates the command value vector, and the control circuit 70 controls the switching circuits 30u, 30v, and 30w according to the command value vector. Therefore, the three output wirings 60 A three-phase alternating current is generated in. FIG. 16 shows the relationship between the currents Iu, Iv, and Iw flowing through the three output wirings 60u, 60v, and 60w and the angle θ of the output voltage vector. As shown in FIG. 16, the phase of the angle θ of the voltage vector is shifted by about 90 ° with respect to the phase of the current Iu. However, the phase difference between the angle θ and the current Iu may further change from FIG. 16 due to the influence of the parasitic resistance of the circuit or the like. Further, when the frequency of the three-phase alternating current is changed, the phase difference between the angle θ and the current Iu may change.

(Variation of medium potential VM)
Next, the fluctuation of the medium potential VM will be described. Among the voltage vectors shown in FIG. 3, in the voltage vector not including the numerical value "1", the medium potential VM is not applied to any of the three output wirings 60. In this case, the fluctuation of the medium potential VM does not occur. For example, when (0, 0, 2) is output, the output wirings 60u and 60v are connected to the low potential wiring 16 and the output wiring 60w is connected to the high potential wiring 12 as shown in FIG. Depending on the operating state of the motor 90, a current flows in the same direction as the voltage applied to the motor 90 (hereinafter referred to as the forward direction) and a direction opposite to the voltage applied to the motor 90 (hereinafter referred to as the reverse direction). ) May flow current. When a current flows in the forward direction, as shown by an arrow 200 in FIG. 17, a current flows from the high potential wiring 12 to the motor 90 via the output wiring 60w. The current flowing into the motor 90 flows to the low potential wiring 16 via the output wirings 60u and 60v. When the current flows in the opposite direction, the current flows in the opposite direction of the arrow 200. In any of these cases, the inflow of electric charge to the medium-potential wiring 14 and the outflow of electric charge from the medium-potential wiring 14 do not occur. Therefore, in this case, the fluctuation of the medium potential VM does not occur. The same applies when (2,0,0), (2,2,0), (0,2,0), (0,2,2), (2,0,2) are output as voltage vectors. Therefore, the fluctuation of the medium potential VM does not occur.

When outputting a voltage vector including the numerical value "1" among the voltage vectors shown in FIG. 3, since the medium potential wiring 14 is connected to at least one of the three output wirings 60, the fluctuation of the medium potential VM changes. Occurs.

For example, when (1, 1, 2) is output, the output wirings 60u and 60v are connected to the medium potential wiring 14 and the output wiring 60w is connected to the high potential wiring 12 as shown in FIG. When a current flows in the forward direction, as shown by an arrow 202 in FIG. 18, a current flows from the high potential wiring 12 to the motor 90 via the output wiring 60w. The current flowing into the motor 90 flows to the medium potential wiring 14 via the output wirings 60u and 60v. In this case, since the upper capacitor 20 is discharged, the medium potential VM rises. When the current flows in the opposite direction, the current flows in the opposite direction of the arrow 202. In this case, since the upper capacitor 20 is charged, the medium potential VM is lowered. As described above, when (1, 1, 2) is output, the medium potential VM increases when the current is in the forward direction, and decreases when the current is in the reverse direction. The same applies when (2,1,1), (2,2,1), (1,2,1), (1,2,2), (2,1,2) are output as voltage vectors. Then, when the current is in the forward direction, the medium potential VM rises, and when the current is in the reverse direction, the medium potential VM decreases.

Further, for example, when (0, 0, 1) is output, the output wirings 60u and 60v are connected to the low potential wiring 16 and the output wiring 60w is connected to the medium potential wiring 14 as shown in FIG. NS. When a current flows in the forward direction, as shown by arrow 204 in FIG. 19, a current flows from the medium potential wiring 14 to the motor 90 via the output wiring 60w. The current flowing into the motor 90 flows to the low potential wiring 16 via the output wirings 60u and 60v. In this case, since the lower capacitor 22 is discharged, the medium potential VM is lowered. When the current flows in the opposite direction, the current flows in the opposite direction of the arrow 204. In this case, since the lower capacitor 22 is charged, the medium potential VM rises. In this way, when (0, 0, 1) is output, the medium potential VM decreases when the current is in the forward direction, and increases when the current is in the reverse direction. The same applies when (1,0,0), (1,1,0), (0,1,0), (0,1,1), (1,0,1) are output as voltage vectors. Then, when the current is in the forward direction, the medium potential VM decreases, and when the current is in the reverse direction, the medium potential VM increases.

Further, among the voltage vectors shown in FIG. 3, (2,1,0), (1,2,0), (0,2,1), (0,1,2), (1,0,2) , (2, 0, 1) is also output, so that the charge flows in and out of the medium potential wiring 14, so that the medium potential VM fluctuates.

As described above, the command circuit 72 changes the command value vector according to the medium potential VM when the torque required by the motor 90 is low. For example, when the medium potential VM is lower than the control target value, a command value vector for raising the medium potential VM is generated. Further, for example, when the medium potential VM is higher than the control target value, a command value vector for lowering the medium potential VM is generated. As described above, in the normal operation, the control circuit 70 controls the potentials of the three output wirings 60 according to the command value vector. Therefore, the three-phase alternating current can be supplied to the motor 90 while controlling the medium potential VM to a value close to the target value.

(Short-circuit element judgment operation)
Next, the short-circuit element determination operation will be described. The control circuit 70 periodically executes a short-circuit element determination operation when the vehicle is not traveling. In the short-circuit element determination operation, it is determined whether or not the switching elements 41 to 44 are short-circuited or not for each of the switching circuits 30u, 30v, and 30w. The short-circuit failure means a failure mode in which the switching element is turned on regardless of the potential of the gate of the switching element. The control circuit 70 selects one of the three switching circuits 30u, 30v, and 30w, and executes a short-circuit element determination operation for the selected switching circuit 30. The control circuit 70 may select all three switching circuits 30 and execute the short-circuit element determination operation for all the selected switching circuits 30 at the same time. Since the short-circuit element determination operation for the switching circuits 30u, 30v, and 30w is the same, the short-circuit element determination operation for one switching circuit 30 will be described below.

Each of the switching elements 41 to 44 is provided with a current detection terminal. The control circuit 70 is connected to the current detection terminals of the switching elements 41 to 44. The control circuit 70 can detect the main current (collector-emitter current or drain-source current) of the switching elements 41 to 44 from the potential of the current detection terminal.

FIG. 20 is a flowchart of the short-circuit element determination operation. As shown in FIG. 20, when the short-circuit element determination operation is started, the control circuit 70 controls the inverter to the first state, the second state, and the third state in steps S2 to S6. FIG. 21 shows a first state, a second state, and a third state. The first state is a state in which the first switching element 41 and the second switching element 42 are on, and the third switching element 43 and the fourth switching element 44 are off. The second state is a state in which the third switching element 43 and the fourth switching element 44 are on, and the first switching element 41 and the second switching element 42 are off. The third state is a state in which the second switching element 42 and the third switching element 43 are on, and the first switching element 41 and the fourth switching element 44 are off.

In step S2, the control circuit 70 controls the switching circuit 30 to the first state, and detects the main current and the medium potential VM of the switching elements 41 to 44. In step S2 (that is, the first state), if the fourth switching element 44 has a short-circuit failure, a line short-circuit occurs between the high-potential wiring 12 and the low-potential wiring 16 as shown by the arrow 300 in FIG. .. In this case, the short-circuit current (overcurrent) is detected by the first switching element 41. Further, in step S2 (that is, in the first state), if the third switching element 43 is short-circuited, a short-circuit between the high-potential wiring 12 and the medium-potential wiring 14 is performed as shown by an arrow 302 in FIG. Occurs. In this case, the short-circuit current is detected in the first switching element 41 and the second switching element 42, and the increase in the medium potential VM is detected.

In step S4, the control circuit 70 controls the switching circuit 30 to the second state, and detects the main current and the medium potential VM of the switching elements 41 to 44. In step S4 (that is, the second state), if the first switching element 41 has a short-circuit failure, a line short-circuit occurs between the high-potential wiring 12 and the low-potential wiring 16 as shown by the arrow 304 in FIG. .. In this case, the short-circuit current is detected by the fourth switching element 44. Further, in step S4 (that is, in the second state), if the second switching element 42 has a short-circuit failure, a short-circuit between the lines is short-circuited between the medium-potential wiring 14 and the low-potential wiring 16 as shown by the arrow 306 in FIG. Occurs. In this case, the short-circuit current is detected in the third switching element 43 and the fourth switching element 44, and the decrease in the medium potential VM is detected.

In step S6, the control circuit 70 controls the switching circuit 30 to the third state, and detects the main current and the medium potential VM of the switching elements 41 to 44. In step S6 (that is, the third state), if the first switching element 41 has a short-circuit failure, a line short-circuit occurs between the high-potential wiring 12 and the medium-potential wiring 14 as shown by arrow 308 in FIG. .. In this case, the short-circuit current is detected in the second switching element 42 and the third switching element 43, and the increase in the medium potential VM is detected. Further, in step S6 (that is, in the third state), if the fourth switching element 44 is short-circuited, a short-circuit is performed between the medium-potential wiring 14 and the low-potential wiring 16 as shown by the arrow 310 in FIG. Occurs. In this case, the short-circuit current is detected in the second switching element 42 and the third switching element 43, and the decrease in the medium potential VM is detected.

The control circuit 70 identifies a switching element that has a short-circuit failure based on the results of steps S2 to S6. If no short-circuit current is detected in any of the switching elements in steps S2 to S6, the control circuit 70 determines that all of the switching elements 41 to 44 are normal.

In the control circuit 70, a short-circuit current is detected in the second switching element 42 and the third switching element 43 in step S6, an increase in the medium potential VM is detected, and a short-circuit current is detected in the fourth switching element 44 in step S4. When it is detected, it is determined that the first switching element 41 has a short-circuit failure.

When the short-circuit current is detected in the third switching element 43 in step S4, the control circuit 70 determines that the second switching element 42 has a short-circuit failure.

When the short-circuit current is detected in the second switching element 42 in step S2, the control circuit 70 determines that the third switching element 43 has a short-circuit failure.

In the control circuit 70, the short-circuit current is detected in the first switching element 41 in step S2, the short-circuit current is detected in the second switching element 42 and the third switching element 43 in step S6, and the medium potential VM is lowered. When it is detected, it is determined that the fourth switching element 44 has a short-circuit failure.

(Prohibited potential)
The inverter 10 can perform an emergency operation when the second switching element 42 or the third switching element 43 has a short-circuit failure. If the first switching element 41 or the fourth switching element 44 has a short-circuit failure, the inverter 10 cannot perform an emergency operation. The control circuit 70 executes an emergency operation when it is necessary to drive the motor 90 in a state where the second switching element 42 or the third switching element 43 has a short-circuit failure. In the following, the switching circuit 30 having a short-circuit fault element will be referred to as a limiting switching circuit 30x. Further, the output wiring 60 of the limiting switching circuit 30x is referred to as a limiting output wiring 60x. Further, the switching circuit 30 other than the limiting switching circuit 30x is referred to as a normal switching circuit 30y. Further, the output wiring 60 of the normal switching circuit 30y is referred to as a normal output wiring 60y. The emergency operation is executed when there is one limiting switching circuit 30x and the short-circuit failure element is the second switching element 42 or the third switching element 43. In the emergency operation, the limiting switching circuit 30x is controlled so that the prohibited potential is not applied to the limiting output wiring 60x. First, the prohibited potential will be described.

The prohibited potential means a voltage that cannot be applied to the limited output wiring 60x because a line short circuit occurs in the limited switching circuit 30x. The prohibited potential differs depending on the type of short-circuit fault element in the limiting switching circuit 30x.

When the second switching element 42 is short-circuited and failed, when the fourth switching element 44 is turned on, the low potential is low from the medium potential wiring 14 via the second switching element 42, the third diode 53, and the fourth switching element 44. A short-circuit current flows through the wiring 16. Therefore, when the second switching element 42 has a short-circuit failure, the fourth switching element 44 cannot be turned on. Therefore, when the second switching element 42 has a short-circuit failure, the prohibited potential is the low potential VL.

When the third switching element 43 is short-circuited and failed, when the first switching element 41 is turned on, the high potential wiring 12 passes through the first switching element 41, the third switching element 43, and the second diode 52 to generate a medium potential. A short-circuit current flows through the wiring 14. Therefore, when the third switching element 43 has a short-circuit failure, the first switching element 41 cannot be turned on. Therefore, when the third switching element 43 has a short-circuit failure, the prohibited potential is a high potential VH.

(Emergency operation)
In the emergency operation, the control circuit 70 selects one voltage vector satisfying a specific condition from a plurality of voltage vectors having the same angle θ as the command value vector input from the command circuit 72, and the selected voltage vector. Is output. In the following, the voltage vector selected by the control circuit 70 may be referred to as a specific voltage vector. The command value vector and the specific voltage vector may be the same or different. The rule that the control circuit 70 selects a specific voltage vector changes according to the prohibited potential.

(A. Emergency operation when the prohibited potential is low potential VL)
When the prohibited potential is the low potential VL (that is, when the short-circuit fault element is the second switching element 42), the control circuit 70 changes the rule for selecting the specific voltage vector according to the medium potential VM. The control circuit 70 stores the upper limit value Vt1, the reference value Vt2, and the lower limit value Vt3 as reference values for controlling the medium potential VM. The reference value Vt2 is a control target value of the medium potential VM. For example, the reference value Vt2 can be halved of the high potential VH. The upper limit value Vt1 is higher than the reference value Vt2, and the lower limit value Vt3 is lower than the reference value Vt2. The control circuit 70 changes the rule for selecting the specific voltage vector between the first rule and the fourth rule according to the flowchart shown in FIG. The control circuit 70 executes the flowchart shown in FIG. 25 each time a command value vector is received from the command circuit 72.

In step S10 of FIG. 25, the control circuit 70 detects the medium potential VM and determines the detected medium potential VM. The control circuit 70 adopts the first rule when the medium potential VM is equal to or higher than the upper limit value Vt1, and adopts the second rule when the medium potential VM is lower than the upper limit value Vt1 and is equal to or higher than the reference value Vt2. The third rule is adopted when the medium potential VM is lower than the reference value Vt2 and is equal to or higher than the lower limit value Vt3, and the fourth rule is adopted when the medium potential VM is lower than the lower limit value Vt3. The second and third rules are adopted when the medium potential VM is close to the reference value Vt2 (control target value), and the first and fourth rules are the rules in which the medium potential VM is the reference value Vt2. It is a rule adopted when it deviates greatly from.

In any of the first rule to the fourth rule, the control circuit 70 determines whether the angle θ of the received command value vector is within the limit angle range or the normal angle range. Therefore, the limiting angle range and the normal angle range will be described below. As described above, in the emergency operation, the prohibited potential cannot be applied to the limited output wiring 60x, so that a part of the voltage vector cannot be output. In the following, a voltage vector that cannot be output in this way (that is, a voltage vector in which the limited output wiring 60x has a prohibited potential) is referred to as a prohibited vector. The limiting angle range means an angle range including a prohibition vector. For example, FIG. 26 shows the limit angle range by reference numeral 400 in the space vector diagram. Note that FIG. 26 shows, as an example, the limiting angle range when the W-phase output wiring 60w is the limiting output wiring 60x and the prohibited potential is the low potential VL. In FIG. 26, the voltage vector in which Vw is “0” (that is, (0,0,0), (1,0,0), (1,1,0), (0,1,0), (2,0) 0,0), (2,1,0), (2,2,0), (1,2,0) and (0,2,0)) are prohibited vectors. The range of the angle θ in which these prohibition vectors exist is the limit angle range. In FIG. 26, the angle range of 0 ° ≦ θ ≦ 120 ° is the limiting angle range. Even within the limiting angle range, a voltage vector whose Vw is not "0" can be output. In the following, the voltage vector that can be output within the limit angle range is referred to as an allowable vector. For example, in FIG. 26, (1,1,1), (2,2,2), (2,1,1), (2,2,1) and (1,2,1) are permissible vectors. .. Further, the angle range outside the limiting angle range is the normal angle range. In FIG. 26, reference number 402 indicates the normal angle range. All voltage vectors within the normal angle range can be output.

Further, in any of the first rule to the fourth rule, the control circuit 70 determines whether the current flowing in the traveling motor is in the forward direction or the reverse direction based on the current flowing in the limited output wiring 60x. As described above, the forward direction means that the current flows in the same direction as the voltage applied to the motor 90, and the reverse direction means that the current flows in the direction opposite to the voltage applied to the motor 90. means.

Further, in the following, a voltage vector composed of only the numerical values "2" and the numerical value "1" is referred to as an upper vector, and a voltage vector composed of only the numerical values "1" and the numerical value "0" is referred to as a lower vector. .. As shown in FIG. 26, the upper vectors are (2,1,1), (2,2,1), (1,2,1), (1,2,2), (1,1,2). ), (2,1,2), and the lower vectors are (1,0,0), (1,1,0), (0,1,0), (0,1,1), ( 0,0,1), (1,0,1). When the upper vector is output, the high potential wiring 12 and the medium potential wiring 14 are connected to the motor 90, so that the upper capacitor 20 is charged or discharged. For example, when the upper vector (1, 1, 2) is output, a current flows along the arrow 202 in FIG. 18 to discharge the upper capacitor 20, and a current flows in the opposite direction of the arrow 202. The upper capacitor 20 may be charged. Further, for example, when the lower vector (0, 0, 1) is output, a current flows along the arrow 204 in FIG. 19 to discharge the lower capacitor 22, and the opposite direction of the arrow 204. A current may flow through the capacitor 22 to charge the lower capacitor 22. Within the normal angle range, the upper vector and the lower vector exist as a pair. For example, as shown in FIG. 26, (1,2,2) and (0,1,1) exist as a pair at θ = 180 °, and (1,1,2) at θ = 240 °. (0,0,1) exists as a pair, and (2,1,2) and (1,0,1) exist as a pair at θ = 300 °.

(A-1. Second rule)
In the second rule, the control circuit 70 selects a specific voltage vector according to Table 2 of FIG. 27 according to the angle θ of the received command value vector and the direction of the current flowing through the limited output wiring 60x.

As shown in Table 2 of FIG. 27, when the angle θ of the command value vector is within the limit angle range, the control circuit 70 sets the allowable vector to a specific voltage regardless of the direction of the current flowing through the limit output wiring 60x. Select as a vector. Here, the control circuit 70 selects an allowable vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. For example, in FIG. 26, when the command value vector is (2,2,0), (2,2,1) is selected as the specific voltage vector. Further, in FIG. 26, when the command value vector is (1,1,0), (2,2,1) is selected as the specific voltage vector. Further, in FIG. 26, when the command value vector is (2,2,1), the same (2,2,1) is selected as the specific voltage vector. Then, the control circuit 70 controls the potentials of the three output wirings 60 according to the specific voltage vector. As described above, when the angle θ of the command value vector is within the limit angle range, an allowable vector having the same angle θ as the command value vector is output. This makes it possible to output a voltage vector having the same angle θ as the command value vector while preventing the output of the prohibition vector.

When the prohibited potential is low potential VL, the permissible vector is always the upper vector. If the allowable vector (upper vector) is output when the current flowing through the motor 90 is in the forward direction, the upper capacitor 20 is discharged and the medium potential VM rises. If the allowable vector (upper vector) is output when the current flowing through the motor 90 is in the opposite direction, the upper capacitor 20 is charged and the medium potential VM is lowered.

As shown in Table 2 of FIG. 27, the control circuit 70 is in the direction of the current flowing through the limited output wiring 60x (that is, the current flowing through the motor 90) when the angle θ of the command value vector is within the normal angle range. Select a specific voltage vector according to. When the direction of the current flowing through the limited output wiring 60x is in the forward direction, the control circuit 70 selects a lower vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. For example, in FIG. 26, when the command value vector is (0,0,2), (0,0,1) is selected as the specific voltage vector. Further, in FIG. 26, when the command value vector is (1, 1, 2), (0, 0, 1) is selected as the specific voltage vector. Further, in FIG. 26, when the command value vector is (0,0,1), the same (0,0,1) is selected as the specific voltage vector. Further, when the direction of the current flowing through the limited output wiring 60x is opposite, the control circuit 70 selects an upper vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. For example, in FIG. 26, when the command value vector is (0,0,2), (1,1,2) is selected as the specific voltage vector. Further, in FIG. 26, when the command value vector is (0, 0, 1), (1, 1, 2) is selected as the specific voltage vector. Further, in FIG. 26, when the command value vector is (1,1,2), the same (1,1,2) is selected as the specific voltage vector. The control circuit 70 controls the potentials of the three output wirings 60 according to the specific voltage vector.

In this way, in the normal angle range, the lower vector is output when the current is in the forward direction, and the upper vector is output when the current is in the reverse direction. When the lower vector is output when the current is in the forward direction, the lower capacitor 22 is discharged and the medium potential VM is lowered. Further, when the upper vector is output when the current is in the opposite direction, the upper capacitor 20 is charged and the medium potential VM is lowered. Thus, in the second rule, the voltage vector at which the medium potential VM decreases is selected in the normal angle range.

As explained above, in the second rule, the control circuit 70 outputs an allowable vector in the limit angle range. As described above, when the permissible vector is output, the medium potential VM may increase or decrease. Further, according to the second rule, the control circuit 70 outputs a voltage vector that lowers the medium potential VM in the normal angle range. Therefore, the medium potential VM tends to decrease as a whole during the period in which the second rule is adopted. As described above, the second rule is adopted when the medium potential VM is higher than the reference value Vt2 (control target value). By adopting the second rule, the medium potential VM higher than the reference value Vt2 can be pulled back to a value close to the reference value Vt2.

(A-2. Third rule)
In the third rule, the control circuit 70 selects a specific voltage vector according to Table 3 of FIG. 27 according to the angle θ of the received command value vector and the direction of the current flowing through the limited output wiring 60x.

As shown in Table 3 of FIG. 27, when the angle θ of the command value vector is within the limit angle range, the control circuit 70 sets the allowable vector to a specific voltage regardless of the direction of the current flowing through the limit output wiring 60x. Select as a vector. Here, the control circuit 70 selects an allowable vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. That is, the third rule in the limiting angle range is equal to the second rule in the limiting angle range. As described above, when the angle θ of the command value vector is within the limit angle range, an allowable vector having the same angle θ as the command value vector is output. This makes it possible to output a voltage vector having the same angle θ as the command value vector while preventing the output of the prohibition vector.

When the prohibited potential is low potential VL, the permissible vector is always the upper vector. If the allowable vector (upper vector) is output when the current flowing through the limited output wiring 60x is in the forward direction, the upper capacitor 20 is discharged and the medium potential VM rises. If the allowable vector (upper vector) is output when the current flowing through the limited output wiring 60x is in the opposite direction, the upper capacitor 20 is charged and the medium potential VM is lowered.

As shown in Table 3 of FIG. 27, when the angle θ of the command value vector is within the normal angle range, the control circuit 70 selects a specific voltage vector according to the direction of the current flowing through the limited output wiring 60x. .. When the direction of the current flowing through the limited output wiring 60x is in the forward direction, the control circuit 70 selects an upper vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. For example, in FIG. 26, when the command value vector is (0,0,2), (1,1,2) is selected as the specific voltage vector. Further, in FIG. 26, when the command value vector is (0, 0, 1), (1, 1, 2) is selected as the specific voltage vector. Further, in FIG. 26, when the command value vector is (1,1,2), the same (1,1,2) is selected as the specific voltage vector. Further, when the direction of the current flowing through the limited output wiring 60x is opposite, the control circuit 70 selects a lower vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. For example, in FIG. 26, when the command value vector is (0,0,2), (0,0,1) is selected as the specific voltage vector. Further, in FIG. 26, when the command value vector is (1, 1, 2), (0, 0, 1) is selected as the specific voltage vector. Further, in FIG. 26, when the command value vector is (0,0,1), the same (0,0,1) is selected as the specific voltage vector. The control circuit 70 controls the potentials of the three output wirings 60 according to the specific voltage vector.

In this way, in the normal angle range, the upper vector is output when the current is in the forward direction, and the lower vector is output when the current is in the reverse direction. When the upper vector is output when the current is in the forward direction, the upper capacitor 20 is discharged and the medium potential VM rises. Further, when the lower vector is output when the current is in the opposite direction, the lower capacitor 22 is charged and the medium potential VM rises. Thus, in the third rule, the voltage vector in which the medium potential VM rises is selected in the normal angle range.

As explained above, in the third rule, the control circuit 70 outputs an allowable vector in the limit angle range. As described above, when the permissible vector is output, the medium potential VM may increase or decrease. Further, according to the third rule, the control circuit 70 outputs a voltage vector that raises the medium potential VM in the normal angle range. Therefore, the medium potential VM tends to increase as a whole during the period in which the third rule is adopted. As described above, the third rule is adopted when the medium potential VM is lower than the reference value Vt2 (control target value). By adopting the third rule, the medium potential VM lower than the reference value Vt2 can be pulled back to a value close to the reference value Vt2.

(A-3. First rule)
In the first rule, the control circuit 70 selects a specific voltage vector according to Table 1 of FIG. 27 according to the angle θ of the command value vector and the direction of the current flowing through the limited output wiring 60x.

As is clear from comparing Table 1 and Table 2 of FIG. 27, the first rule is when the current flowing through the limited output wiring 60x is in the forward direction and the angle θ of the command value vector is within the limited angle range. (Table 1) is different from the second rule (Table 2). In other cases, the first rule (Table 1) is equal to the second rule (Table 2).

In the first rule, when the current flowing through the limit output wiring 60x is in the forward direction and the angle θ of the command value vector is within the limit angle range, the control circuit 70 selects a zero vector as the specific voltage vector. The zero vector is a voltage vector in which the potentials of the three output wirings 60 are the same potential. The cello vector includes (0,0,0), (1,1,1), (2,2,2). Here, the control circuit 70 selects a zero vector (that is, (1,1,1) or (2,2,2)) that does not include the prohibited potential (low potential VL) as the specific voltage vector. The control circuit 70 outputs a specific voltage vector (that is, a zero vector).

As described above with respect to the second rule, when the forbidden potential is the low potential VL, the medium potential VM rises when the permissible vector is output when the current is in the forward direction. The first rule is adopted when the medium potential VM is extremely high (that is, when the medium potential VM is higher than the upper limit value Vt1). As described above, when the medium potential VM is extremely high, it is preferable to reduce the medium potential VM as quickly as possible. Therefore, the control circuit 70 stops the output of the permissible vector at the timing when the medium potential VM rises when the permissible vector is output (that is, the timing when the current is in the forward direction) and outputs the zero vector. Therefore, in the first rule, the medium potential VM can be lowered faster than in the second rule, and the medium potential VM can be pulled back to a value closer to the reference value Vt2. In this way, in the first rule adopted when the medium potential VM rises extremely, the medium potential VM is controlled to an appropriate value by giving priority to lowering the medium potential VM over the power supply to the motor 90. do.

(A-4. Fourth rule)
In the fourth rule, the control circuit 70 selects a specific voltage vector according to Table 4 of FIG. 27 according to the angle θ of the command value vector and the direction of the current flowing through the limited output wiring 60x.

As is clear from comparing Tables 4 and 3 of FIG. 27, the fourth rule is when the current flowing through the limited output wiring 60x is in the opposite direction and the angle θ of the command value vector is within the limited angle range. (Table 4) is different from the third rule (Table 3). In other cases, Rule 4 (Table 4) is equal to Rule 3 (Table 3).

In the fourth rule, when the current flowing through the limit output wiring 60x is in the opposite direction and the angle θ of the command value vector is within the limit angle range, the control circuit 70 selects a zero vector as the specific voltage vector. Here, the control circuit 70 selects a zero vector (that is, (1,1,1) or (2,2,2)) that does not include the prohibited potential (low potential VL) as the specific voltage vector. The control circuit 70 outputs a specific voltage vector (that is, a zero vector).

As described above with respect to the third rule, when the forbidden potential is the low potential VL, the medium potential VM decreases when the allowable vector is output when the current is in the opposite direction. The fourth rule is adopted when the medium potential VM is extremely low (that is, when the medium potential VM is lower than the lower limit value Vt3). As described above, when the medium potential VM is extremely low, it is preferable to raise the medium potential VM as quickly as possible. Therefore, the control circuit 70 stops the output of the permissible vector at the timing when the medium potential VM decreases when the permissible vector is output (that is, the timing when the current is in the opposite direction) and outputs the zero vector. Therefore, in the fourth rule, the medium potential VM can be raised faster than in the third rule, and the medium potential VM can be pulled back to a value closer to the reference value Vt2. In this way, in the fourth rule adopted when the medium potential VM is extremely lowered, the medium potential VM is controlled to an appropriate value by giving priority to raising the medium potential VM over the power supply to the motor 90. do.

As described above, when the prohibited potential is the low potential VL, the medium potential VM is controlled to a value near the reference value Vt2 by adopting the first to fourth rules according to the medium potential VM. However, the voltage vector is output so as to rotate. Therefore, the three-phase alternating current can be continuously generated to drive the motor 90 continuously.

In the emergency operation when the prohibited potential is the low potential VL, the potential of the limiting output wiring 60x is set to two levels of the high potential VH and the medium potential VM by selecting the specific voltage vector according to the first to fourth rules. Controlled, the potentials of the two normal output wirings 60y are controlled at three levels: high potential VH, medium potential VM, and low potential VL.

When the command value vector includes a fraction as a parameter as in the voltage vectors E2 and E3 of FIG. 4, the control circuit 70 has a plurality of specific voltage vectors around the command value vector according to the first to fourth rules. Is selected, and a plurality of selected specific voltage vectors are output with a time lag, so that these specific voltage vectors may be combined and a voltage vector having the same angle θ as the command value vector may be output.

(A-5. Example of control)
FIG. 28 shows an example of an emergency operation when the prohibited potential is a low potential VL. In the example of FIG. 28, the limited output wiring 60x is the W phase output wiring 60w. The graph of FIG. 28 shows the sine (sin θ) of the angle θ of the command value vector. 0 ° ≤ θ ≤ 120 ° is the limiting angle range. Further, the voltage vector (specific voltage vector) output by each of the first rule, the second rule, the third rule, and the fourth rule is shown in the table at the bottom of FIG. 28.

In the first rule, at the timing of θ = 0 °, the angle θ is within the limit angle range, and the motor current (current flowing through the limit output wiring 60x) is in the forward direction, so that it is a zero vector (1,1,1). 1) (or (2,2,2)) is output. At the timing of θ = 60 °, since the angle θ is within the limiting angle range and the motor current is in the opposite direction, the allowable vector (2, 2, 1) is output. At the timing of θ = 120 °, the angle θ is within the limit angle range and the motor current is in the opposite direction, so that the allowable vector (1, 2, 1) is output. At the timing of θ = 180 °, since the angle θ is within the normal angle range and the motor current is in the opposite direction, the upper vector (1, 2, 2) is output. At the timing of θ = 240 °, since the angle θ is within the normal angle range and the motor current is in the forward direction, the lower vector (0, 0, 1) is output. At the timing of θ = 300 °, since the angle θ is within the normal angle range and the motor current is in the forward direction, the lower vector (1, 0, 1) is output.

In the second rule, at the timing of θ = 0 °, the angle θ is within the limit angle range and the motor current is in the forward direction, so that the allowable vector (2,1,1) is output. At the timing of θ = 60 °, since the angle θ is within the limiting angle range and the motor current is in the opposite direction, the allowable vector (2, 2, 1) is output. At the timing of θ = 120 °, the angle θ is within the limit angle range and the motor current is in the opposite direction, so that the allowable vector (1, 2, 1) is output. At the timing of θ = 180 °, since the angle θ is within the normal angle range and the motor current is in the opposite direction, the upper vector (1, 2, 2) is output. At the timing of θ = 240 °, since the angle θ is within the normal angle range and the motor current is in the forward direction, the lower vector (0, 0, 1) is output. At the timing of θ = 300 °, since the angle θ is within the normal angle range and the motor current is in the forward direction, the lower vector (1, 0, 1) is output.

According to the third rule, at the timing of θ = 0 °, the angle θ is within the limit angle range and the motor current is in the forward direction, so that the allowable vector (2,1,1) is output. At the timing of θ = 60 °, since the angle θ is within the limiting angle range and the motor current is in the opposite direction, the allowable vector (2, 2, 1) is output. At the timing of θ = 120 °, the angle θ is within the limit angle range and the motor current is in the opposite direction, so that the allowable vector (1, 2, 1) is output. At the timing of θ = 180 °, since the angle θ is within the normal angle range and the motor current is in the opposite direction, the lower vector (0,1,1) is output. At the timing of θ = 240 °, since the angle θ is within the normal angle range and the motor current is in the forward direction, the upper vector (1, 1, 2) is output. At the timing of θ = 300 °, since the angle θ is within the normal angle range and the motor current is in the forward direction, the upper vector (2, 1, 2) is output.

According to the fourth rule, at the timing of θ = 0 °, the angle θ is within the limit angle range and the motor current is in the forward direction, so that the allowable vector (2,1,1) is output. At the timing of θ = 60 °, since the angle θ is within the limit angle range and the motor current is in the opposite direction, a zero vector (1,1,1) is output. At the timing of θ = 120 °, since the angle θ is within the limit angle range and the motor current is in the opposite direction, a zero vector (1,1,1) is output. At the timing of θ = 180 °, since the angle θ is within the normal angle range and the motor current is in the opposite direction, the lower vector (0,1,1) is output. At the timing of θ = 240 °, since the angle θ is within the normal angle range and the motor current is in the forward direction, the upper vector (1, 1, 2) is output. At the timing of θ = 300 °, since the angle θ is within the normal angle range and the motor current is in the forward direction, the upper vector (2, 1, 2) is output.

As explained above, the medium potential VM is controlled to a value close to the reference value Vt2 by selecting and outputting the specific voltage vector according to the rules.

(B. Emergency operation when the prohibited potential is high potential VH)
When the prohibited potential is the high potential VH (that is, when the short-circuit fault element is the third switching element 43), the control circuit 70 changes the rule for selecting the specific voltage vector according to the medium potential VM. The control circuit 70 changes the rule for selecting the specific voltage vector between the fifth rule and the eighth rule according to the flowchart shown in FIG. The control circuit 70 executes the flowchart shown in FIG. 29 each time a command value vector is received from the command circuit 72.

In step S20 of FIG. 29, the control circuit 70 detects the medium potential VM and determines the detected medium potential VM. The control circuit 70 adopts the fifth rule when the medium potential VM is the upper limit value Vt1 or more, and adopts the sixth rule when the medium potential VM is lower than the upper limit value Vt1 and is the reference value Vt2 or more. The seventh rule is adopted when the medium potential VM is lower than the reference value Vt2 and is equal to or higher than the lower limit value Vt3, and the eighth rule is adopted when the medium potential VM is lower than the lower limit value Vt3. The sixth and seventh rules are adopted when the medium potential VM is close to the reference value Vt2 (control target value), and the fifth and eighth rules are the rules in which the medium potential VM is the reference value Vt2. It is a rule adopted when it deviates greatly from.

In any of the fifth to eighth rules, the control circuit 70 determines whether the angle θ of the received command value vector is within the limit angle range or the normal angle range. FIG. 30 shows the limiting angle range 406 and the normal angle range 408 when the prohibited potential is the high potential VH. Note that FIG. 30 shows, as an example, a case where the W-phase output wiring 60w is the limited output wiring 60x. In FIG. 30, the voltage vector in which Vw is “2” (that is, (2,2,2), (1,2,2), (1,1,2), (2,1,2), (0, 2,2), (0,1,2), (0,0,2), (1,0,2) and (2,0,2)) are prohibited vectors. The range of the angle θ in which these prohibition vectors exist is the limit angle range 406. In FIG. 30, the angle range of 180 ° ≦ θ ≦ 300 ° is the limiting angle range 406. In FIG. 30, the permissible vectors are (1,1,1), (0,0,0), (0,1,1), (0,0,1) and (1,0,1).

Further, in any of the fifth to eighth rules, the control circuit 70 determines whether the current flowing through the limited output wiring 60x is in the forward direction or the reverse direction.

(B-1. 6th rule)
In the sixth rule, the control circuit 70 selects a specific voltage vector according to Table 6 of FIG. 31 according to the angle θ of the received command value vector and the direction of the current flowing through the limited output wiring 60x.

As shown in Table 6 of FIG. 31, when the angle θ of the command value vector is within the limit angle range, the control circuit 70 sets the allowable vector to a specific voltage regardless of the direction of the current flowing through the limit output wiring 60x. Select as a vector. Here, the control circuit 70 selects an allowable vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. Then, the control circuit 70 controls the potentials of the three output wirings 60 according to the specific voltage vector. As described above, when the angle θ of the command value vector is within the limit angle range, an allowable vector having the same angle θ as the command value vector is output. This makes it possible to output a voltage vector having the same angle θ as the command value vector while preventing the output of the prohibition vector.

When the prohibited potential is high potential VH, the permissible vector is always the lower vector. If the allowable vector (lower vector) is output when the current flowing through the limited output wiring 60x is in the forward direction, the lower capacitor 22 is discharged and the medium potential VM is lowered. If the allowable vector (lower vector) is output when the current flowing through the limited output wiring 60x is in the opposite direction, the lower capacitor 22 is charged and the medium potential VM rises.

As shown in Table 6 of FIG. 31, when the angle θ of the command value vector is within the normal angle range, the control circuit 70 selects a specific voltage vector according to the direction of the current flowing through the limited output wiring 60x. .. When the direction of the current flowing through the limited output wiring 60x is in the forward direction, the control circuit 70 selects a lower vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. When the direction of the current flowing through the limited output wiring 60x is opposite, the control circuit 70 selects an upper vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. The control circuit 70 controls the potentials of the three output wirings 60 according to the specific voltage vector.

In this way, in the normal angle range, the lower vector is output when the current is in the forward direction, and the upper vector is output when the current is in the reverse direction. When the lower vector is output when the current is in the forward direction, the lower capacitor 22 is discharged and the medium potential VM is lowered. Further, when the upper vector is output when the current is in the opposite direction, the upper capacitor 20 is charged and the medium potential VM is lowered. Thus, in the sixth rule, the voltage vector at which the medium potential VM decreases is selected in the normal angle range.

As explained above, in the sixth rule, the control circuit 70 outputs an allowable vector in the limit angle range. As described above, when the permissible vector is output, the medium potential VM may increase or decrease. Further, according to the sixth rule, the control circuit 70 outputs a voltage vector that lowers the medium potential VM in the normal angle range. Therefore, the medium potential VM tends to decrease as a whole during the period in which the sixth rule is adopted. As described above, the sixth rule is adopted when the medium potential VM is higher than the reference value Vt2 (control target value). By adopting the sixth rule, the medium potential VM higher than the reference value Vt2 can be pulled back to a value close to the reference value Vt2.

(B-2. 7th rule)
In the seventh rule, the control circuit 70 selects a specific voltage vector according to Table 7 of FIG. 31 according to the angle θ of the received command value vector and the direction of the current flowing through the limited output wiring 60x.

As shown in Table 7 of FIG. 31, when the angle θ of the command value vector is within the limit angle range, the control circuit 70 sets the allowable vector to a specific voltage regardless of the direction of the current flowing through the limit output wiring 60x. Select as a vector. Here, the control circuit 70 selects an allowable vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. That is, the seventh rule in the limiting angle range is equal to the sixth rule in the limiting angle range. As described above, when the angle θ of the command value vector is within the limit angle range, an allowable vector having the same angle θ as the command value vector is output. This makes it possible to output a voltage vector having the same angle as the command value vector while preventing the output of the prohibition vector.

If the allowable vector (lower vector) is output when the current flowing through the limited output wiring 60x is in the forward direction, the lower capacitor 22 is discharged and the medium potential VM is lowered. If the allowable vector (lower vector) is output when the current flowing through the limited output wiring 60x is in the opposite direction, the lower capacitor 22 is charged and the medium potential VM rises.

As shown in Table 7 of FIG. 31, when the angle θ of the command value vector is within the normal angle range, the control circuit 70 selects a specific voltage vector according to the direction of the current flowing through the limited output wiring 60x. .. When the direction of the current flowing through the limited output wiring 60x is in the forward direction, the control circuit 70 selects an upper vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. Further, when the direction of the current flowing through the limited output wiring 60x is opposite, the control circuit 70 selects a lower vector having the same angle θ as the angle θ of the command value vector as the specific voltage vector. The control circuit 70 controls the potentials of the three output wirings 60 according to the specific voltage vector.

In this way, in the normal angle range, the upper vector is output when the current is in the forward direction, and the lower vector is output when the current is in the reverse direction. When the upper vector is output when the current is in the forward direction, the upper capacitor 20 is discharged and the medium potential VM rises. Further, when the lower vector is output when the current is in the opposite direction, the lower capacitor 22 is charged and the medium potential VM rises. Thus, in the seventh rule, the voltage vector in which the medium potential VM rises is selected in the normal angle range.

As explained above, in the seventh rule, the control circuit 70 outputs an allowable vector in the limit angle range. As described above, when the permissible vector is output, the medium potential VM may increase or decrease. Further, according to the seventh rule, the control circuit 70 outputs a voltage vector that raises the medium potential VM in the normal angle range. Therefore, the medium potential VM tends to increase as a whole during the period in which the seventh rule is adopted. As described above, the seventh rule is adopted when the medium potential VM is lower than the reference value Vt2 (control target value). By adopting the seventh rule, the medium potential VM lower than the reference value Vt2 can be pulled back to a value close to the reference value Vt2.

(B-3. Rule 5)
In the fifth rule, the control circuit 70 selects a specific voltage vector according to Table 5 of FIG. 31 according to the angle θ of the command value vector and the direction of the current flowing through the limited output wiring 60x.

As is clear from comparing Tables 5 and 6 of FIG. 31, when the current flowing through the limited output wiring 60x is in the opposite direction and the angle θ of the command value vector is within the limited angle range, the fifth rule ( Table 5) is different from the sixth rule (Table 6). In other cases, Rule 5 (Table 5) is equal to Rule 6 (Table 6).

In the fifth rule, when the current flowing through the limit output wiring 60x is in the opposite direction and the angle θ of the command value vector is within the limit angle range, the control circuit 70 selects a zero vector as the specific voltage vector. Here, the control circuit 70 selects a zero vector (that is, (0,0,0) or (1,1,1)) that does not include the prohibited potential (high potential VH) as the specific voltage vector. The control circuit 70 outputs a specific voltage vector (that is, a zero vector).

As described above with respect to the sixth rule, in the limiting angle range, if the allowable vector is output when the current is in the opposite direction, the medium potential VM rises. The fifth rule is adopted when the medium potential VM is extremely high (that is, when the medium potential VM is higher than the upper limit value Vt1). As described above, when the medium potential VM is extremely high, it is preferable to reduce the medium potential VM as quickly as possible. Therefore, the control circuit 70 stops the output of the permissible vector at the timing when the medium potential VM rises when the permissible vector is output (that is, the timing when the current is in the opposite direction) and outputs the zero vector. Therefore, in the fifth rule, the medium potential VM can be lowered faster than in the sixth rule, and the medium potential VM can be pulled back to a value closer to the reference value Vt2. In this way, in the fifth rule adopted when the medium potential VM rises extremely, the medium potential VM is controlled to an appropriate value by giving priority to lowering the medium potential VM over the power supply to the motor 90. do.

(B-4. 8th rule)
In the eighth rule, the control circuit 70 selects a specific voltage vector according to Table 8 of FIG. 31 according to the angle θ of the command value vector and the direction of the current flowing through the limited output wiring 60x.

As is clear from comparing Tables 8 and 7 of FIG. 31, the eighth rule is when the current flowing through the limited output wiring 60x is in the forward direction and the angle θ of the command value vector is within the limited angle range. (Table 8) is different from the seventh rule (Table 7). In other cases, Rule 8 (Table 8) is equivalent to Rule 7 (Table 7).

In the eighth rule, when the current flowing through the limit output wiring 60x is in the forward direction and the angle θ of the command value vector is within the limit angle range, the control circuit 70 selects a zero vector as the specific voltage vector. Here, the control circuit 70 selects a zero vector (that is, (0,0,0) or (1,1,1)) that does not include the prohibited potential (high potential VH) as the specific voltage vector. The control circuit 70 outputs a specific voltage vector (that is, a zero vector).

As described above with respect to the seventh rule, when the prohibited potential is the high potential VH, the medium potential VM decreases when the allowable vector is output when the current is in the forward direction. The eighth rule is adopted when the medium potential VM is extremely low (that is, when the medium potential VM is lower than the lower limit value Vt3). As described above, when the medium potential VM is extremely low, it is preferable to raise the medium potential VM as quickly as possible. Therefore, the control circuit 70 stops the output of the permissible vector at the timing when the medium potential VM decreases when the permissible vector is output (that is, the timing when the current is in the forward direction) and outputs the zero vector. Therefore, in the eighth rule, the medium potential VM can be raised faster than in the seventh rule, and the medium potential VM can be pulled back to a value closer to the reference value Vt2. In this way, in the eighth rule adopted when the medium potential VM is extremely lowered, the medium potential VM is controlled to an appropriate value by giving priority to raising the medium potential VM over the power supply to the motor 90. do.

As described above, when the prohibited potential is the high potential VH, the medium potential VM is controlled to a value near the reference value Vt2 by adopting the fifth to eighth rules according to the medium potential VM. However, the voltage vector is output so as to rotate. Therefore, the three-phase alternating current can be continuously generated to drive the motor 90 continuously.

In the emergency operation when the prohibited potential is the high potential VH, the potential of the limiting output wiring 60x is set to two levels of the medium potential VM and the low potential VL by selecting the specific voltage vector according to the fifth to eighth rules. Controlled, the potentials of the two normal output wirings 60y are controlled at three levels: high potential VH, medium potential VM, and low potential VL.

When the command value vector includes a fraction as a parameter as in the voltage vectors E2 and E3 of FIG. 4, the control circuit 70 has a plurality of specific voltage vectors around the command value vector according to the fifth to eighth rules. Is selected, and a plurality of selected specific voltage vectors are output with a time lag, so that these specific voltage vectors may be combined and a voltage vector having the same angle θ as the command value vector may be output.

In the third embodiment described above, the case where the current flowing through the limited output wiring 60x is in the forward direction and the reverse direction has been described. However, during operation, a current may flow in a forward direction to some of the output wirings 60, and a current may flow in a reverse direction to the remaining output wirings 60. Even in such a case, the control circuit 70 appropriately selects either a voltage vector that raises the medium potential VM or a voltage vector that lowers the medium potential VM according to the direction and magnitude of the current flowing through each output wiring 60. And output, and the medium potential VM can be controlled to a value close to the reference value Vt2.

(Example 4)
In the inverter of the fourth embodiment, the emergency operation is different from that of the second embodiment. More specifically, in the inverter of the fourth embodiment, the operation of the control circuit 70 when the angle θ of the command value vector is within the normal range is different from that of the third embodiment. The operation of the control circuit 70 when the angle θ of the command value vector is within the limit range is the same in the fourth embodiment and the third embodiment.

In the fourth embodiment, the operation of the control circuit 70 when the angle θ of the command value vector is within the normal range is common to the first to eighth rules. In the fourth embodiment, the control circuit 70 selects one of the upper vector group and the lower vector group, and selects and outputs a voltage vector having the same command value vector and angle θ in the selected group. do. FIG. 32 shows the operation of the control circuit 70 when the angle θ of the command value vector is within the normal range in the fourth embodiment. The control circuit 70 repeatedly executes the process shown in FIG. 32. In step S40, the control circuit 70 detects the medium potential VM and calculates the difference ΔVM between the medium potential VM and the reference value Vt2 (control target value). The difference ΔVM is calculated as an absolute value. The control circuit 70 stores the calculated difference ΔVM. After the difference ΔVM is calculated, steps S42 to S48 are executed. As a result of executing steps S42 to S48, the medium potential VM fluctuates. In step S40 of the next control phase, the difference ΔVM is calculated again. Hereinafter, the difference ΔVM calculated in the current control phase is referred to as a difference ΔVM1, and the difference ΔVM calculated in the previous control phase is referred to as a difference ΔVM2. In step S42, the control circuit 70 determines whether or not the difference ΔVM1 calculated in the current control phase is within the permissible value α. If YES in step S42, the control circuit 70 is in the same group as the voltage vector output in the previous control phase in step S48 (that is, either the upper vector group or the lower vector group). Selects a voltage vector having the same angle θ as the command value vector as a specific voltage vector and outputs it. If NO in step S42, in step S44, the control circuit 70 determines whether or not the difference ΔVM1 calculated in the current control phase is equal to or less than the difference ΔVM2 calculated in the previous control phase. That is, the control circuit 70 determines whether or not the medium potential VM is closer to the reference value Vt2 (control target value) than the previous control phase. If YES in step S44, step S48 is executed. If NO in step S44, in step S46, the control circuit 70 is in a group different from the voltage vector output in the previous control phase (that is, either the upper vector group or the lower vector group). Among them, a voltage vector having the same angle θ as the command value vector is selected as a specific voltage vector and output.

In Example 4, when the medium potential VM deviates from the reference value Vt2 to the allowable value α or more, the voltage vector is selected so that the difference ΔVM is reduced. As a result, it is possible to prevent the medium potential VM from deviating significantly from the reference value Vt2. Further, according to the fourth embodiment, the voltage vector can be appropriately output regardless of the detected value of the current of the output wiring 60. Even when the current detection speed is not sufficient with respect to the control speed, the medium potential VM can be controlled to an appropriate value according to the configuration of the fourth embodiment.

Note that, in a configuration other than the above-mentioned Examples 3 and 4, a three-phase alternating current may be generated while controlling the medium potential VM during an emergency operation. By controlling the potential of the limited output wiring at two levels excluding the prohibited potential and controlling the potential of the normal output wiring at three levels of high potential VH, medium potential VM, and low potential VL, other than Examples 1 and 2. Even if the voltage vector is output according to the rule of, it is possible to appropriately control the medium potential VM. For example, in a part of the normal angle range, a voltage vector consisting of only the numerical values "0" and the numerical value "2" and a voltage vector including all the numerical values "0", the numerical value "1", and the numerical value "2" are output. You may.

Further, in the above-described first and second embodiments, the command circuit 72 inputs the command value vector to the control circuit 70, but the command circuit 72 may input only the command value of the angle θ of the voltage vector to the control circuit 70. ..

Note that, in the above-described Examples 1 to 4, as shown in FIG. 33, the positions of the second switching element 42 and the second diode 52 may be replaced with the positions of the third switching element 43 and the third diode 53.

The second diode 52 of Examples 1 to 4 described above is an example of the first intermediate diode of the claim. Further, the third diode 53 of Examples 1 to 4 described above is an example of the second intermediate diode of the claim.

Although the embodiments have been described in detail above, these are merely examples and do not limit the scope of claims. The techniques described in the claims include various modifications and modifications of the specific examples illustrated above. The technical elements described in the present specification or the drawings exhibit technical usefulness alone or in various combinations, and are not limited to the combinations described in the claims at the time of filing. In addition, the techniques illustrated in the present specification or drawings achieve a plurality of objectives at the same time, and achieving one of the objectives itself has technical usefulness.

Claims (13)

1. It ’s an inverter,
   High potential wiring to which high potential is applied and
   Low potential wiring to which low potential is applied and
   A medium potential wiring to which a medium potential lower than the high potential and higher than the low potential is applied, and
   Switching circuit and
   A detector that detects the current flowing in the inverter or the temperature of the inverter,
   Control circuit,
   Have,
   The switching circuit
   Output wiring and
   A first switching element connected between the high-potential wiring and the output wiring,
   A second switching element connected between the medium potential wiring and the output wiring,
   A third switching element connected in series with the second switching element between the medium potential wiring and the output wiring,
   A fourth switching element connected between the output wiring and the low potential wiring,
   A first intermediate diode connected in parallel to the second switching element with the cathode facing the medium potential wiring side.
   A second intermediate diode, which is connected in parallel to the third switching element with the cathode facing the output wiring side.
   Have,
   The current capacities of the second switching element and the third switching element are lower than the current capacities of the first switching element and the fourth switching element.
   The control circuit is configured to control the potentials of the gates of the first switching element, the second switching element, the third switching element, and the fourth switching element.
   The control circuit turns on the switching circuit in a high potential output state in which the first switching element is turned on and the high potential is applied to the output wiring, and the second switching element and the third switching element are turned on. It is configured to change between the medium potential output state in which the medium potential is applied to the output wiring and the low potential output state in which the fourth switching element is turned on and the low potential is applied to the output wiring. Wiring,
   The control circuit is configured to generate an alternating current in the output wiring by changing the switching circuit between the high potential output state, the medium potential output state, and the low potential output state.
   When the detection value of the detector is larger than the reference value, the control circuit lowers the ratio of the period for controlling the switching circuit to the medium potential output state as compared with the case where the detection value is smaller than the reference value.
   Inverter.
2. It has three switching circuits.
   The control circuit is configured to control the potentials of the gates of the first switching element, the second switching element, the third switching element, and the fourth switching element of each of the three switching circuits. Ori,
   The control circuit changes the three switching circuits between the high potential output state, the medium potential output state, and the low potential output state, so that the three phases are between the output wirings of the three switching circuits. It is configured to generate an AC current and is configured to generate an AC current.
   For each of the three switching circuits, when the detection value of the detector is larger than the reference value, the control circuit controls the medium potential output state more than when the detection value is smaller than the reference value. Lower the ratio,
   The inverter of claim 1.
3. A claim that the control circuit changes the switching circuit between the high potential output state and the low potential output state without making the switching circuit into the medium potential output state when the detected value is larger than the reference value. 1 or 2 inverters.
4. When the detected value is larger than the reference value, the control circuit changes the switching circuit between the high potential output state and the low potential output state, and in this case, the high potential output state is changed to the said. When changing to the low potential output state, the high potential output state is changed to the low potential output state via the medium potential output state, and when changing from the low potential output state to the high potential output state, the low potential is changed. The output state is changed to the high potential output state via the medium potential output state,
   When the detected value is larger than the reference value, the control circuit controls the ratio of the period for controlling the medium potential output state to the high potential output state and the ratio for controlling the low potential output state. Lower than the ratio of
   The inverter of claim 1 or 2.
5. The first switching element is made of a silicon semiconductor.
   The second switching element is composed of a compound semiconductor.
   The third switching element is composed of a compound semiconductor.
   The fourth switching element is made of a silicon semiconductor.
   The inverter according to any one of claims 1 to 4.
6. The first intermediate diode is composed of a compound semiconductor.
   The second intermediate diode is composed of a compound semiconductor.
   The inverter of claim 5.
7. It ’s an inverter,
   High potential wiring to which high potential is applied and
   Low potential wiring to which low potential is applied and
   Medium potential wiring and
   An upper capacitor connected between the high-potential wiring and the medium-potential wiring,
   The lower capacitor connected between the medium-potential wiring and the low-potential wiring,
   Three switching circuits, a U-phase switching circuit, a V-phase switching circuit, and a W-phase switching circuit,
   Control circuit,
   Have,
   Each of the three switching circuits
   Output wiring and
   A first switching element connected between the high-potential wiring and the output wiring,
   A second switching element connected between the medium potential wiring and the output wiring,
   A third switching element connected in series with the second switching element between the medium potential wiring and the output wiring,
   A fourth switching element connected between the output wiring and the low potential wiring,
   A first intermediate diode connected in parallel to the second switching element with the cathode facing the medium potential wiring side.
   A second intermediate diode, which is connected in parallel to the third switching element with the cathode facing the output wiring side.
   Have,
   The control circuit is configured to control the potentials of the gates of the first switching element, the second switching element, the third switching element, and the fourth switching element of the three switching circuits.
   The control circuit is in a high potential output state in which the three switching circuits are turned on and the high potential is applied to the corresponding output wiring, the second switching element and the third switching element. Is turned on to apply the medium potential, which is the potential of the medium potential wiring, to the corresponding output wiring, and the fourth switching element is turned on to apply the low potential to the corresponding output wiring. It is configured to vary between low potential output states.
   The control circuit is the U-phase output wiring which is the output wiring of the U-phase switching circuit, the V-phase output wiring which is the output wiring of the V-phase switching circuit, and the output wiring of the W-phase switching circuit. By changing the respective potentials of the W-phase output wiring between the high potential, the medium potential, and the low potential, the U-phase output wiring, the V-phase output wiring, and the W-phase output wiring It is configured to generate a three-phase AC current between them.
   The control circuit can execute an emergency operation when any of the second switching element and the third switching element of the three switching circuits has a short-circuit failure.
   The element with a short-circuit failure is called a short-circuit failure element.
   The output wiring of one of the three switching circuits including the short-circuit fault element is referred to as a limited output wiring.
   The output wiring of the two switching circuits that do not include the short-circuit fault element of the three switching circuits is referred to as a normal output wiring, respectively.
   In the emergency operation, the control circuit changes the potential of the limited output wiring between the two potentials other than the prohibited potential among the high potential, the medium potential, and the low potential, and the normal output. Each potential of the wiring is changed between the three potentials of the high potential, the medium potential, and the low potential.
   When the short-circuit failure element is the second switching element, the prohibited potential is the low potential.
   When the short-circuit failure element is the third switching element, the prohibited potential is the high potential.
   Inverter.
8. It also has a command circuit that generates a command value for the angle of the voltage vector so that the voltage vector rotates and inputs it to the control circuit.
   The voltage vector is a vector represented by the parameters Vu, Vv, and Vw.
   The parameter Vu is a value indicating whether the potential of the U-phase output wiring is the high potential, the medium potential, or the low potential.
   The parameter Vv is a value indicating whether the potential of the V-phase output wiring is the high potential, the medium potential, or the low potential.
   The parameter Vw is a value indicating whether the potential of the W-phase output wiring is the high potential, the medium potential, or the low potential.
   In the emergency operation, the voltage vector at which the limited output wiring has the prohibited potential is the prohibited vector, and the angle range including the prohibited vector in the angle range of the voltage vector is the restricted angle range. The angle range outside the limit angle range of the angle range of the voltage vector is the normal angle range.
   In the emergency operation, when the angle indicated by the command value is within the limiting angle range, the control circuit has the angle indicated by the command value and the three said according to an allowable vector which is not the prohibited vector. Control the potential of the output wiring,
   In the emergency operation, when the angle indicated by the command value is within the normal angle range, the control circuit selects a specific voltage vector from a plurality of voltage vectors having the angle indicated by the command value. , Control the potentials of the three output wirings according to the selected specific voltage vector,
   The control circuit selects the voltage vector that raises the medium potential as the specific voltage vector when the medium potential is lower than the reference value.
   The control circuit selects the voltage vector that lowers the medium potential as the specific voltage vector when the medium potential is higher than the reference value.
   The inverter of claim 7.
9. The three output wires are configured to be connected to the load.
   The voltage vector in which the high potential and the medium potential are applied to the load and the low potential is not applied is referred to as an upper vector.
   The voltage vector in which the medium potential and the low potential are applied to the load and the high potential is not applied is referred to as a lower vector.
   When the angle indicated by the command value is within the normal angle range, the control circuit has the following conditions A to D, that is,
   A. When the medium potential is lower than the reference value and the current flowing through the load is in the forward direction with respect to the voltage applied to the load, the upper vector is selected as the specific voltage vector.
   B. When the medium potential is lower than the reference value and the current flowing through the load is in the opposite direction to the voltage applied to the load, the lower vector is selected as the specific voltage vector. ,
   C. The lower vector is selected as the specific voltage vector when the medium potential is higher than the reference value and the current flowing through the load is in the forward direction with respect to the voltage applied to the load. ,
   D. When the medium potential is higher than the reference value and the current flowing through the load is in the opposite direction to the voltage applied to the load, the upper vector is selected as the specific voltage vector.
   The inverter according to claim 8, wherein the specific voltage vector is selected according to the above condition.
10. In the control circuit, the voltage at which the medium potential is higher than the upper limit value higher than the reference value, the angle indicated by the command value is within the limit angle range, and the allowable vector raises the medium potential. In the case of a vector, the inverter according to claim 9, which controls the three output wirings to the same potential regardless of the angle indicated by the command value.
11. In the control circuit, the medium potential is lower than the lower limit value lower than the reference value, the angle indicated by the command value is within the limit angle range, and the allowable vector lowers the medium potential. In the case of a voltage vector, the inverter according to claim 9 or 10, which controls the three output wirings to the same potential regardless of the angle indicated by the command value.
12. It also has a command circuit that generates a command value for the angle of the voltage vector so that the voltage vector rotates and inputs it to the control circuit.
    The voltage vector is a vector represented by the parameters Vu, Vv, and Vw.
    The parameter Vu is a value indicating whether the potential of the U-phase output wiring is the high potential, the medium potential, or the low potential.
    The parameter Vv is a value indicating whether the potential of the V-phase output wiring is the high potential, the medium potential, or the low potential.
    The parameter Vw is a value indicating whether the potential of the W-phase output wiring is the high potential, the medium potential, or the low potential.
    In the emergency operation, the voltage vector at which the limited output wiring has the prohibited potential is the prohibited vector, and the angle range including the prohibited vector in the angle range of the voltage vector is the restricted angle range. The angle range outside the limit angle range of the angle range of the voltage vector is the normal angle range.
    In the emergency operation, when the angle indicated by the command value is within the limiting angle range, the control circuit has the angle indicated by the command value and the three said according to an allowable vector which is not the prohibited vector. Control the potential of the output wiring,
    In the emergency operation, when the angle indicated by the command value is within the normal angle range, the control circuit selects a specific voltage vector from a plurality of voltage vectors having the angle indicated by the command value. , Control the potentials of the three output wirings according to the selected specific voltage vector,
    The three output wires are configured to be connected to the load.
    The specific voltage vector is composed of a first group composed of a group of voltage vectors in which the high potential and the medium potential are applied to the load and the low potential is not applied to the load, and the medium potential and the low potential are applied to the load. Is selected from one of the second groups composed of the group of voltage vectors in which the high potential is applied and the high potential is not applied.
    The control circuit stores the control target value of the medium potential, and
    In the control circuit, when the deviation between the medium potential and the control target value increases after the previous control phase, the voltage vector selected in the previous control phase between the first group and the second group can be used. Select the specific voltage vector from a group different from the group to which it belongs.
    The inverter of claim 7.
13. The control circuit is configured to detect currents flowing through the second switching element and the third switching element of the three switching circuits.
    The control circuit can execute a short-circuit element determination operation for a selection switching circuit selected from the three switching circuits.
    In the short-circuit element determination operation, the control circuit turns on the selection switching circuit in the first state in which the first switching element and the second switching element are turned on, and in the third state and the fourth switching element. Change over time between the second states
    The control circuit is relative to the selective switching circuit.
    When a short-circuit current flows through the third switching element in the second state, it is determined that the second switching element is a short-circuit failure element.
    When a short-circuit current flows through the second switching element in the first state, it is determined that the third switching element is a short-circuit failure element.
    The inverter according to any one of claims 7 to 12.

Images