JP6716039B2 - Power converter - Google Patents

Description

The present invention relates to a power converter capable of four-quadrant operation, and more particularly, to controlling the power converter in a discontinuous current mode.

An inverter is known as a power conversion device capable of four-quadrant operation. An inverter capable of four-quadrant operation generally operates in a current continuous mode in which a current continuously flows through a load. It is known that in the continuous current mode, a current ripple occurs at the timing (current zero cross) at which the direction (polarity) of the current flowing through the load is reversed.

For example, as in the inverter disclosed in Japanese Patent Laid-Open No. 60-156280 (Patent Document 1), by operating the inverter in a current discontinuous mode in which there is a time zone in which almost no current flows through the load, a current ripple is generated. It is possible to reduce the loss due to.

JP, 60-156280, A

In JP-A-60-156280 (Patent Document 1), no consideration is given to the difference in control model between the current discontinuous mode and the current continuous mode. If the current control of the current continuous mode is applied to the current discontinuous mode as it is, the current control may become unstable. As a result, stable power supply by the power conversion device may be difficult.

The present invention has been made to solve the above problems, and an object thereof is to stabilize the power supply of a power conversion device in a current discontinuous mode.

The power converter according to the present invention converts the DC voltage from the DC power supply into an AC voltage in the current discontinuous mode. The power conversion device includes first to fourth terminals, a full bridge circuit, a reactor, a current controller, a non-linear compensator, and a PWM controller. The first and second terminals are connected to the positive electrode and the negative electrode of the DC power supply, respectively. The third and fourth terminals output an AC voltage. The full bridge circuit includes first and second legs connected in parallel between the first terminal and the second terminal. The reactor is connected between a third connection terminal and a first connection point between the upper arm and the lower arm of the first leg. The current controller calculates a target voltage value across the reactor so that the current value flowing through the reactor becomes the target current value. The non-linear compensator receives the target voltage value and the output voltage value output between the third terminal and the fourth terminal, and calculates the target conduction ratio. The PWM controller receives the target conduction ratio from the non-linear compensator and outputs a PWM signal to the full bridge circuit to operate the full bridge circuit in the first or second switching mode. The fourth terminal is connected to a second connection point between the upper arm and the lower arm of the second leg. In the discontinuous current mode, there is a time period during which no current flows in the reactor. In the first switching mode, the upper arm of the first leg and the lower arm of the second leg are switched in synchronization with each other, and the lower arm of the first leg and the upper arm of the second leg are made non-conductive. .. In the second switching mode, the lower arm of the first leg and the upper arm of the second leg are switched in synchronization with each other, and the upper arm of the first leg and the lower arm of the second leg are made non-conductive. .. In the current discontinuous mode, the non-linear compensator uses the target voltage value, the target conduction ratio before one sampling time, and the output voltage value before one sampling time to calculate the target conduction ratio at this sampling time.

According to the power converter of the present invention, the non-linear compensator uses the target voltage value, the target conduction ratio before one sampling time, and the output voltage value before one sampling time to obtain the target conduction ratio for the current sampling time. By calculating, the current control can be stabilized in the current discontinuous mode. As a result, power can be stably supplied in the current discontinuous mode.

It is a functional block diagram which shows the functional structure of the power converter device which concerns on Embodiment 1. It is a time chart of the reactor current in the first switching mode. It is a figure which shows the conduction|electrical_connection state of the full bridge circuit in the time zone when a reactor current increases. It is a figure which shows the conduction|electrical_connection state of the full bridge circuit in the time zone when a reactor current reduces. It is a figure which shows the conduction state of a full bridge circuit in the time zone when current does not flow into the reactor L. It is a block diagram corresponding to the circuit model of a power converter when operating in the 1st switching mode. It is a block diagram corresponding to a micro change model of a power converter when operating in the 1st switching mode. FIG. 8 is a control block diagram corresponding to the minute change model shown in FIG. 7. It is a time chart of the reactor current in the second switching mode. It is a figure which shows the conduction|electrical_connection state of the full bridge circuit in the time zone when a reactor current reduces. It is a figure which shows the conduction|electrical_connection state of the full bridge circuit in the time zone when a reactor current increases. It is a figure which shows the conduction state of a full bridge circuit in the time zone when current does not flow into the reactor L. It is a block diagram corresponding to the circuit model of a power converter when operating in the 2nd switching mode. It is a block diagram corresponding to a micro change model of a power converter when operating in the 2nd switching mode. FIG. 15 is a control block diagram corresponding to the minute change model shown in FIG. 14. It is a functional block diagram which shows the functional structure of a non-linear compensator. It is a figure which shows together the time chart of each of a conduction ratio command, a reference triangular wave, a polarity determination signal, and a PWM signal. It is a figure which also shows the time chart of each of a conduction ratio command, a reactor current, a current value of a reactor current which has passed through a low-pass filter, a current command, an output voltage of an inverter, and a system current. 5 is a functional block diagram of a non-linear compensator of the power conversion device according to the second embodiment. FIG. It is a functional block diagram which shows the functional structure of a conduction ratio selector. It is a figure which shows together the time chart of each of a conduction ratio command and a polarity determination signal. It is a figure which shows together the time chart of each of a conduction ratio command, a reference triangular wave, a polarity determination signal, and a PWM signal. It is a figure which also shows the time chart of each of a conduction ratio command, a reactor current, a current value of a reactor current which has passed through a low-pass filter, a current command, an output voltage of an inverter, and a system current. It is a functional block diagram which shows the functional structure of the power converter device which concerns on Embodiment 3. It is a functional block diagram which shows the functional structure of a non-linear compensator. It is a functional block diagram which shows the functional structure of a conduction ratio selector. It is a time chart of the conduction ratio command in the current discontinuous mode. It is a figure which shows together the time chart of each of a conduction ratio command, a reference triangular wave, an automatic switching signal, and a PWM signal. It is a figure which also shows the time chart of each of a conduction ratio command, a reactor current, a current value of a reactor current which has passed through a low-pass filter, a current command, an output voltage of an inverter, and a system current. 6 is a time chart of a conduction ratio command when the current discontinuous mode and the current continuous mode are selectively switched. It is a figure which shows together the time chart of each of a conduction ratio command, a reference triangular wave, an automatic switching signal, and a PWM signal. It is a figure which also shows the time chart of each of a conduction ratio command, a reactor current, a current value of a reactor current which has passed through a low-pass filter, a current command, an output voltage of an inverter, and a system current.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the drawings, the same or corresponding parts are designated by the same reference numerals and the description thereof will not be repeated in principle.

Embodiment 1.
FIG. 1 is a functional block diagram showing a functional configuration of the power conversion device 100 according to the first embodiment. As shown in FIG. 1, the power conversion device 100 includes a first terminal P1, a second terminal P2, a third terminal P3, a fourth terminal P4, an inverter 1, an LC filter 2, and a low-pass filter LPF1. , LPF 2, a switching mode switch 5, a current controller 6, a non-linear compensator 701, and a PWM (Pulse Width Modulation) controller 801.

In the current discontinuous mode, the power conversion device 100 converts the DC voltage from the DC power supply 3 into an AC voltage and outputs the AC voltage to the AC power supply 4. The current discontinuous mode is an operation mode in which there is a time period in which the current output from the inverter 1 is stopped.

The first terminal P1 and the second terminal P2 are connected to the positive electrode and the negative electrode of the DC power supply 3, respectively. The voltage of the DC power supply 3 is the voltage Vdc. The DC power supply 3 is, for example, a constant voltage source such as a battery or a control voltage source of a power converter capable of DC output. The power converter 100 may be configured such that the inverter 1 behaves as a control voltage source and a DC power source is established.

Inverter 1 includes a full bridge circuit FB and a current sensor 51. The full bridge circuit FB includes a first leg LG1 and a second leg LG2 connected in parallel between the first terminal P1 and the second terminal P2.

The first leg LG1 includes an upper arm Q1, a lower arm Q2, and freewheel diodes D1 and D2. The upper arm Q1 and the lower arm Q2 are connected in series between the first terminal P1 and the second terminal P2. The freewheel diode D1 is connected in antiparallel with the upper arm Q1. The freewheel diode D2 is connected in antiparallel with the lower arm Q2.

The second leg LG2 includes an upper arm Q3, a lower arm Q4, and freewheel diodes D3 and D4. The upper arm Q3 and the lower arm Q4 are connected in series between the first terminal P1 and the second terminal P2. The freewheel diode D3 is connected in antiparallel with the upper arm Q3. The freewheel diode D4 is connected in antiparallel with the lower arm Q4.

The full bridge circuit FB operates in the first switching mode or the second switching mode. In the first switching mode, the upper arm Q1 and the lower arm Q4 perform switching in synchronization with each other, and the lower arm Q2 and the upper arm Q3 are rendered non-conductive. In the second switching mode, the lower arm Q2 and the upper arm Q3 switch in synchronization with each other, and the upper arm Q1 and the lower arm Q4 are rendered non-conductive.

Examples of the upper arms Q1 and Q3 and the lower arms Q2 and Q4 include a self-extinguishing type semiconductor switching element represented by, for example, an IGBT (Insulated Gate Bipolar Transistor) or a MOSFET (Metal Oxide Semiconductor Field Effective Transistor). it can. When MOSFETs are used as the upper arms Q1 and Q3 and the lower arms Q2 and Q4, parasitic diodes of MOSFETs may be used instead of the freewheel diodes D1 to D4.

The current sensor 51 measures the reactor current Iinv flowing to the LC filter 2 from the first connection point N1 between the upper arm Q1 and the lower arm Q2, and outputs it to the low pass filter LPF1.

The low-pass filter LPF1 receives the value of the reactor current Iinv and outputs the current value FIinv to the current controller 6. The current value FIinv is a current value in which the high frequency component of the reactor current Iinv is suppressed. Note that the low-pass filter LPF1 can be omitted if it is possible to obtain information that can interpolate the difference between the sampling value and the average current including the current discontinuity in the current discontinuous mode. is there. Further, when the current of the capacitor C of the LC filter 2 is sufficiently small and the phase difference of the fundamental wave between the reactor current Iinv and the system current Igrid flowing through the AC power supply 4 is small, the current received by the low-pass filter LPF1 is changed to the reactor. The current Iinv can be changed to the system current Igrid. In the current discontinuous mode, there is a time zone in which the reactor current Iinv becomes 0.

The LC filter 2 includes a reactor L, a capacitor C, and a voltage sensor 52. The reactor L is connected between the first connection point N1 and the third terminal P3. The capacitor C is connected between the third terminal P3 and the fourth terminal P4. The fourth terminal is connected to a second connection point N2 between the upper arm Q3 and the lower arm Q4. The LC filter 2 smoothes the reactor current Iinv generated by the potential difference between the output voltage Vinv of the inverter 1 and the voltage Vc (C voltage), and outputs it to the AC power supply 4. A reactor current Iinv, which is an L current flowing through the reactor L, is measured by the current sensor 51. The voltage sensor 52 measures the voltage Vc of the capacitor C and outputs it to the low pass filter LPF2. The voltage Vinv is a voltage output from the inverter 1 to the LC filter 2, and is a voltage between the first connection point N1 and the second connection point N2.

The low-pass filter LPF2 receives the value of the voltage Vc and outputs the voltage value FVc to the non-linear compensator 701. The voltage value FVc is a voltage value in which the sampling error of the voltage Vc due to the switching noise of the full bridge circuit FB is reduced by the low pass filter LPF2. When the influence of the switching noise of the full bridge circuit FB is small, the low pass filter LPF2 may be omitted.

The AC power supply 4 is connected between the third terminal P3 and the fourth terminal P4. The AC power supply 4 is, for example, a single-phase AC system. The AC power supply 4 includes a system impedance Lo. Instead of the AC power source 4, a configuration may be used in which the inverter 1 controls the voltage Vc of the LC filter 2 via a controller of the reactor current Iinv during a self-sustaining operation in which a load is connected. The load can be, for example, a resistive load, an inductive load, a rectifier load, or a home appliance load.

The switching mode switch 5 receives the current command Iinv* (target current value) and outputs the polarity determination signal Ipole* to the nonlinear compensator 701 and the PWM controller 801. When the current command Iinv* has a positive polarity, the polarity determination signal Ipole* becomes 1. When the current command Iinv* has a negative polarity, the polarity determination signal Ipole* becomes 0. The full bridge circuit FB operates in the first switching mode when the polarity determination signal Ipole* is 1, and operates in the second switching mode when the polarity determination signal Ipole* is 0. Note that there is usually some delay until the polarity change of the current command Iinv* is reflected in the polarity of the reactor current Iinv through the current control process. In order to eliminate such a delay and make the polarity determination signal Ipole* and the timing of the polarity change of the reactor current Iinv coincide, the switching mode switching device 5 may include a correction term.

The current controller 6 uses the voltage command (both ends) of the reactor L so that the current value FIinv becomes the current command Iinv* based on the result of comparison between the current value FIinv from the low-pass filter LPF1 and the current command Iinv*. The target voltage value) VL* is calculated, and the reactor voltage command VL* is output to the non-linear compensator 701. The current controller 6 calculates the voltage command VL* based on the comparison result of the average value of the reactor current Iinv and the current command Iinv* in the period Tsw of the reference triangular wave used by the PWM controller 801 to generate the PWM signal. May be.

In the current discontinuous mode, the non-linear compensator 701 uses the reactor voltage command VL*, the conduction ratio command D* before the sampling time, and the voltage value FVc before the sampling time, and the conduction ratio at the current sampling time. The command D* is calculated and output to the PWM controller 801.

The PWM controller 801 generates PWM signals Sa and Sb using the conduction ratio command D*, the reference triangular wave Carrier and the polarity determination signal Ipole*. The cycle of the reference triangular wave Carrier is the carrier cycle Tsw. The conduction ratio of the PWM signal Sa is D*. The PWM signal Sa is output to the upper arm Q1 and the lower arm Q4. The duty ratio of the PWM signal Sb is (1-D*). The PWM signal Sb is output to the lower arm Q2 and the upper arm Q3. When the polarity determination signal Ipole* is 1 (the current command Iinv* has a positive polarity), the PWM signal Sb becomes 0, and the full bridge circuit FB operates in the first switching mode. When the polarity determination signal Ipole* is 0 (the current command Iinv* is negative), the PWM signal Sa becomes 0 and the full bridge circuit FB operates in the second switching mode.

Next, the current control performed in power conversion device 100 will be described separately for the first switching mode and the second switching mode. First, the current control in the first switching mode will be described with reference to FIGS. The current control in the second switching mode will be described with reference to FIGS. 9 to 15.

FIG. 2 is a time chart of the reactor current Iinv in the first switching mode. FIG. 2 shows changes in the reactor current Iinv in the range of the carrier cycle Tsw. The conduction ratio command corresponding to the time zone shown in FIG. 2 is D*, and the non-conduction ratio is Dn. As shown in FIG. 2, the reactor current Iinv, which is 0 at the time tm1, increases until the time tm2, and becomes the current ipeak (>0) at the time tm2. After that, the reactor current Iinv decreases until time tm3 and becomes 0 at time tm3. After that, the reactor current Iinv is 0 until time tm4. The time intervals tm1 to tm2 are represented as D*·Tsw. The time intervals tm2 to tm3 are represented by (1-D*-Dn)·Tsw. The time intervals tm3 to tm4 are represented as Dn·Tsw.

FIG. 3 is a diagram showing a conduction state of the full bridge circuit FB in a time zone (time zones tm1 to tm2 in FIG. 2) in which the reactor current Iinv increases. As shown in FIG. 3, in the time zones tm1 to tm2, the upper arm Q1 and the lower arm Q4 conduct. In the first switching mode, the lower arm Q2 and the upper arm Q3 are non-conductive. The voltage Vinv output from the inverter 1 becomes the voltage Vdc of the DC power supply 3. The current flows from the positive electrode of the DC power supply 3 to the third terminal P3 via the upper arm Q1 and the reactor L, and also flows from the fourth terminal P4 to the negative electrode of the DC power supply 3 via the lower arm Q4. As a result, the value of the reactor current Iinv increases from 0 to ipeak.

FIG. 4 is a diagram showing a conduction state of the full bridge circuit FB in a time zone (time zones tm2 to tm3 in FIG. 2) in which the reactor current Iinv decreases. As shown in FIG. 4, in the time zones tm3 to tm4, the upper arm Q1, the lower arm Q2, the upper arm Q3, and the lower arm Q4 are all non-conductive. The voltage Vinv is the reverse voltage (-Vdc) of the DC power supply 3. The current from the reactor L is returned to the positive electrode of the DC power supply 3 via the third terminal P3, the fourth terminal P4, and the freewheel diode D3. As a result, the value of the reactor current Iinv decreases from ipeak to 0.

FIG. 5: is a figure which shows the conduction|electrical_connection state of the full bridge circuit FB in the time zone (time zone tm3-tm4 of FIG. 2) in which the electric current does not flow into the reactor L. As shown in FIG. 5, in the time zones tm3 to tm4, the upper arm Q1, the lower arm Q2, the upper arm Q3, and the lower arm Q4 are all non-conductive. Since the reactor current Iinv output from the inverter 1 is 0, the voltage Vinv becomes the voltage Vc.

The voltage Vinv in the time zones tm1 to tm2 (D*·Tsw) is Vdc. The voltage Vinv in the time zones tm2 to tm3 ((1-D*-Dn).Tsw) is -Vdc. The voltage Vinv in the time zone tm3 to tm4 (Dn·Tsw) is Vc. Therefore, the voltage Vinv in one carrier cycle Tsw can be expressed by the following formula (1) as an average value of the voltage values in each time zone.

FIG. 6 is a block diagram showing the transmission from the voltage Vinv to the voltage Vc in the circuit model of the power conversion device 100 when operating in the first switching mode. As shown in FIG. 6, the reactor voltage VL is obtained from the difference between the voltage Vinv and the voltage Vc obtained by the equation (1). The reactor current Iinv is obtained from the reactor voltage VL and the impedance of the reactor L. The system current Igrid is obtained from the voltage Vc, the voltage Vgrid of the AC power supply 4, and the system impedance Lo. The voltage Vc is obtained from the difference value between the reactor current Iinv and the system current Igrid and the impedance of the capacitor C.

Expression (2) represents the reactor voltage VL (=Vinv-Vc) obtained from the conduction ratio command D* and the voltage Vc shown in FIG.

The non-conduction rate Dn can be expressed as in Expression (3) using the current average value iave in the carrier period Tsw in FIG. 2 and the current change width ipeak in the times tm1 to tm2.

The current change width ipeak can be expressed as in Expression (4).

By substituting the equation (4) into the equation (3), the non-flow rate Dn is expressed by the equation (5).

By substituting the equation (5) into the equation (2), the reactor voltage VL is expressed by the equation (6).

In the current discontinuous mode, the initial value of the reactor voltage VL is 0. By substituting 0 for the reactor voltage VL of the equation (6), the initial value of the current average value iave is expressed by the equation (7).

The voltage change width sL·iave in the carrier period Tsw of the reactor voltage VL can be approximated as in Expression (8).

By substituting the equation (8) into the equation (6), the equation (9) is derived.

The initial value of the conduction ratio command D* is D0, the minute change amount of the conduction ratio command D* is ΔD, the initial value of the voltage Vc is Vc0, the minute change amount of the voltage Vc is ΔVc, and the initial value of the current average value iave is When iave0 and the minute change amount of the current average value iave are Δiave, the equation (9) is expressed by the equation (10). Expression (10) represents the minute characteristics of the conduction ratio command D*, the voltage Vc, and the average current value iave of Expression (9).

When the differential term with respect to the constant (excluding the minute change amounts ΔD, ΔVc, and Δiave) of the equation (10) and the square term of the minute change amounts ΔD, ΔVc, and Δiave are approximated to 0, the equation (10) becomes 11).

By substituting the equation (7) representing the initial value of the average current value iave into the equation (11), the equation (12) can be transformed. In Expression (12), D*, iave, and Vc in Expression (7) are replaced with D0, iave0, and Vc0 that are initial value functions, respectively.

It is difficult to control the average current value iave above the frequency of the reference triangular wave Carrier. Since the time constants of the numerator and the denominator of the equation (12) are equal to or more than the carrier period Tsw, the first-order term of the s function in the equation (12) can be almost ignored regarding the control of the current average value iave. Therefore, the average current value iave can be controlled based on the equation (13) in which only the zero-order term of the s function in the equation (12) is left.

Here, the voltage change width sL·iave per carrier period Tsw of the reactor voltage VL shown in the equation (8) can be replaced by the equation (14).

By substituting the equation (14) into the equation (13), the equation (15) is derived.

From the equation (15), the block diagram shown in FIG. 6 can be replaced with the block diagram shown in FIG. 7. The small change model is reflected in the block diagram shown in FIG. 7. As shown in FIG. 7, the reactor voltage VL includes a non-linear disturbance change 811 with a small change amount ΔVc and a non-linear gain change 812 with a small change amount ΔD. It is difficult to realize the desired reactor voltage VL by the conduction ratio command D* calculated in the control model that does not consider these. In order to realize the desired reactor voltage VL, a minute change amount ΔD is calculated according to the following equation (16) so that the nonlinear disturbance change 811 and the nonlinear gain change 812 are canceled, and the conduction ratio command D* is calculated. Need to decide. The expression (16-1) shows the first correction value CV11, and the expression (16-2) shows the second correction value CV12.

FIG. 8 is a control block diagram corresponding to the minute change model shown in FIG. Equation (16) is shown in the first nonlinear compensation mode 71 of FIG. The first non-linear compensation mode 71 includes a first correction term 711 and a second correction term 712. In order to cancel the nonlinear disturbance change 811 of the minute change amount ΔVc, the first correction term 711 is multiplied by the minute change amount ΔVc one sampling time before, and the first correction value CV11 is calculated. .. The first correction value CV11 is added to the reactor voltage command VL*, and the second correction value CV12 is calculated. In order to cancel the nonlinear gain change 812 of the minute change amount ΔD, the second correction term 712 is multiplied by the second correction value CV12 to calculate the minute change amount ΔD. In the first nonlinear compensation mode 71 of FIG. 8, the initial value Vc0 and the minute change amount ΔVc of the equation (15) are replaced with the initial value FVc0 and the minute change amount ΔFVc of the detected value of the voltage Vc, respectively. The reactor voltage command VL* output from the current controller 6 is controlled by the first non-linear compensation mode 71 so as to substantially match the reactor voltage VL.

Next, current control in the second switching mode will be described with reference to FIGS. 9 to 15.
FIG. 9 is a time chart of the reactor current Iinv in the second switching mode. FIG. 9 shows changes in the reactor current Iinv in the range of the carrier cycle Tsw. The conduction ratio corresponding to the time zone shown in FIG. 9 is set as the conduction ratio command D*, and the non-conduction ratio is defined as Dn. As shown in FIG. 9, the reactor current Iinv, which is 0 at time tm5, decreases until time tm6, and becomes ipeak (<0) at time tm6. Reactor current Iinv increases until time tm7 and becomes 0 at time tm7. Reactor current Iinv is 0 until time tm8. The time intervals tm5 to tm6 are represented by (1-D*)·Tsw. The time intervals tm6 to tm7 are represented by (D*-Dn)·Tsw. The time intervals tm7 to tm8 are represented as Dn·Tsw.

FIG. 10: is a figure which shows the conduction state of the full bridge circuit FB in the time zone (time zone tm5-tm6 of FIG. 9) when reactor current Iinv reduces. As shown in FIG. 10, in the time zones tm5 to tm6, the lower arm Q2 and the upper arm Q3 conduct. In the second switching mode, the upper arm Q1 and the lower arm Q4 are non-conductive. The voltage Vinv output from the inverter 1 becomes the reverse voltage (-Vdc) of the DC power supply 3. The current flows from the positive electrode of the DC power supply 3 to the fourth terminal P4 via the upper arm Q3, and also flows from the third terminal P3 to the negative electrode of the DC power supply 3 via the reactor L and the lower arm Q2. As a result, the value of the reactor current Iinv decreases from 0 to ipeak.

FIG. 11 is a diagram showing the conduction state of the full bridge circuit FB in the time zone (time zones tm6 to tm7 in FIG. 9) in which the reactor current Iinv increases. As shown in FIG. 11, in the time zones tm6 to tm7, the upper arm Q1, the lower arm Q2, the upper arm Q3, and the lower arm Q4 are all non-conductive. The voltage Vinv becomes the voltage Vdc of the DC power supply 3. The current from the reactor L is returned to the positive electrode of the DC power supply 3 via the freewheel diode D1. As a result, the value of the reactor current Iinv increases from ipeak to 0.

FIG. 12 is a diagram showing a conduction state of the full bridge circuit FB in a time zone (time zones tm7 to tm8 in FIG. 9) in which no current flows in the reactor L. As shown in FIG. 12, in the time zones tm7 to tm8, the upper arm Q1, the lower arm Q2, the upper arm Q3, and the lower arm Q4 are all non-conductive. Since the reactor current Iinv output from the inverter 1 is 0, the voltage Vinv becomes the voltage Vc.

The voltage Vinv in the time zones tm5 to tm6 ((1-D*)·Tsw) is −Vdc. The voltage Vinv in the time zone tm6 to tm7 ((D*-Dn)·Tsw) is Vdc. The voltage Vinv in the time zone tm7 to tm8 (Dn·Tsw) is Vc. Therefore, in the second switching mode, the voltage Vinv in one carrier cycle Tsw can be expressed by the following equation (17) as an average value of the voltage values in each time zone.

FIG. 13 is a block diagram showing the transfer from voltage Vinv to voltage Vc in the circuit model of power conversion device 100 when operating in the second switching mode. As shown in FIG. 13, the reactor voltage VL is obtained from the difference between the voltage Vinv and the voltage Vc obtained by the equation (17). The reactor current Iinv is obtained from the reactor voltage VL and the impedance of the reactor L. The system current Igrid is obtained from the voltage Vc, the voltage Vgrid of the AC power supply 4, and the system impedance Lo. The voltage Vc is obtained from the difference value between the reactor current Iinv and the system current Igrid and the impedance of the capacitor C.

Expression (18) represents the reactor voltage VL (=Vinv-Vc) obtained from the conduction ratio command D* and the voltage Vc shown in FIG.

The non-conduction rate Dn can be expressed as in Expression (3) as in the first switching mode by using the current average value iave in the carrier period Tsw and the current change width ipeek in the times tm5 to tm6 in FIG. it can. The current change width ipeak in the second switching mode is expressed as in Expression (19).

By substituting the equation (19) into the equation (3), the equation (20) is derived.

Substituting equation (20) into equation (18) leads to equation (21).

In the current discontinuous mode, the initial value of the reactor voltage VL is 0. By substituting 0 for the reactor voltage VL of the equation (21), the initial value of the current average value iave is expressed by the equation (22).

By substituting the equation (8), which means the voltage change width of the reactor voltage VL per carrier period Tsw, into the equation (22), the equation (23) is derived.

The initial value of the conduction ratio command D* is D0, the minute change amount of the conduction ratio command D* is ΔD, the initial value of the voltage Vc is Vc0, the minute change amount of the voltage Vc is ΔVc, and the initial value of the average current value iave is When iave0 and the minute change amount of the current average value iave are Δiave, the equation (23) is expressed by the equation (24). Expression (24) represents minute characteristics of the conduction ratio command D*, voltage Vc, and current average value iave of Expression (23).

When the differential term with respect to the constant (excluding the minute change amounts ΔD, ΔVc, and Δiave) of the equation (24) and the square term of the minute change amounts ΔD, ΔVc, and Δiave are approximated to 0, the equation (24) becomes the expression (24). 25).

By substituting the equation (22) representing the initial value of the average current value iave into the equation (25), the equation (26) is derived. In Expression (26), D*, iave, and Vc in Expression (22) are replaced with D0, iave0, and Vc0 that are initial value functions.

It is difficult to control the average current value iave above the frequency of the reference triangular wave Carrier. Since the time constants of the numerator and the denominator of the equation (26) are equal to or more than the carrier period Tsw, the first-order term of the s function in the equation (26) can be almost ignored for controlling the average current value iave. Therefore, the average current value iave can be controlled based on the equation (27) in which only the zero-order term of the s function in the equation (26) is left.

By substituting the equation (14), which means the voltage change width of the reactor voltage VL per carrier period Tsw, into the equation (27), the equation (28) is derived.

From the equation (28), the block diagram shown in FIG. 13 can be replaced with the block diagram shown in FIG. The block diagram shown in FIG. 14 reflects the small change model. As shown in FIG. 14, the reactor voltage VL includes a non-linear disturbance change 821 of a minute change amount (output voltage change amount) ΔVc and a non-linear gain change 822 of a minute change amount (commutation rate change amount) ΔD. Be done. It is difficult to realize the desired reactor voltage VL by the conduction ratio command D* calculated in the control model that does not consider these. In order to realize the desired reactor voltage VL, a minute change amount ΔD is calculated according to the following equation (29) so that the nonlinear disturbance change 821 and the nonlinear gain change 822 are canceled, and the conduction ratio command D* is calculated. Need to decide. The equation (29-1) shows the first correction value CV21, and the equation (29-2) shows the second correction value CV22.

FIG. 15 is a control block diagram corresponding to the minute change model shown in FIG. Equation (29) is shown in the second non-linear compensation mode 72 of FIG. The second non-linear compensation mode 72 includes a first correction term 721 and a second correction term 722. In order to cancel the nonlinear disturbance change 821, the first correction term 721 is multiplied by the minute change amount ΔVc one sampling time before, and the first correction value CV21 is calculated. The first correction value CV21 is added to the reactor voltage command VL*, and the second correction value CV22 is calculated. In order to cancel the nonlinear gain change 822 of the minute change amount ΔD, the second correction term 722 is multiplied by the second correction value CV22, and the minute change amount ΔD is calculated. In the second non-linear compensation mode 72, the initial value Vc0 and the minute change amount ΔVc in equation (29) are replaced with the initial value FVc0 of the detected value of the voltage Vc and the minute change amount ΔFVc, respectively. The reactor voltage command VL* output from the current controller 6 is controlled by the second non-linear compensation mode 72 so as to substantially match the reactor voltage VL.

The first correction term 711 and the second correction term 712 of the first switching mode shown in FIG. 8 are different from the first correction term 721 and the second correction term 722 of the second switching mode shown in FIG. 15, respectively. Therefore, it is necessary to switch the processing performed by the non-linear compensator 701 between the first switching mode and the second switching mode.

As shown in FIG. 2, in the first switching mode, the reactor current Iinv has a positive polarity. On the other hand, as shown in FIG. 9, in the second switching mode, the reactor current Iinv has a negative polarity. Therefore, the power converter 100 switches between the first switching mode and the second switching mode based on the polarity determination signal Ipole*. When the switching mode is switched from the first switching mode to the second switching mode at the current zero cross, it is necessary to switch the reference (initial value) of the flow rate command D* from 0 to 1. On the contrary, when the second switching mode is switched to the first switching mode at the current zero cross, it is necessary to switch the reference of the conduction ratio command D* from 1 to 0.

FIG. 16 is a functional block diagram showing a functional configuration of the non-linear compensator 701. As shown in FIG. 16, the non-linear compensator 701 includes a first non-linear compensation mode 71 and a second non-linear compensation mode 72. When the polarity determination signal Ipole* is 1, the first nonlinear compensation mode 71 is selected, and the initial value D0 of the flow rate command D* is set to 0. When the polarity determination signal Ipole* is 0, the second non-linear compensation mode 72 is selected and the initial value D0 is set to 1.

FIG. 17 is a diagram additionally showing time charts of the conduction coefficient D*, the reference triangular wave Carrier, the polarity determination signal Ipole*, and the PWM signals Sa and Sb. In FIG. 17, a time chart from time tm9 to tm12 is shown. In FIG. 17, the reference triangular wave Carrier is simply drawn for the sake of easy viewing of the drawing. The reference triangular wave Carrier changes between 0 and 1 with the carrier cycle Tsw.

As shown in FIG. 17, at time tm9 to tm10 and time tm11 to tm12, the PWM signal Sa changes between On and Off, and the PWM signal Sb is Off. The polarity determination signal Ipole* is 1 at times tm9 to tm10 and times tm11 to tm12. At times tm9 to tm10 and times tm11 to tm12, the full bridge circuit FB is operating in the first switching mode.

At times tm10 to tm11, the PWM signal Sb changes between On and Off, and the PWM signal Sa is Off. At times tm10 to tm11, the polarity determination signal Ipole* is 0. At times tm10 to tm11, the full bridge circuit FB is operating in the second switching mode.

The switching mode of the full bridge circuit FB changes from the first switching mode to the second switching mode at time tm10, and changes from the second switching mode to the first switching mode at time tm11. The conduction ratio command D* is set to 1 which is the initial value of the second switching mode at time tm10. The conduction ratio command D* is set to 0 which is the initial value of the first switching mode at time tm11.

FIG. 18 is a diagram collectively showing respective time charts of the duty ratio command D*, the reactor current Iinv, the current value FIinv, the current command Iinv*, the voltage Vc, and the system current Igrid. Also in FIG. 18, as in FIG. 17, a time chart from time tm9 to tm12 is shown. The time chart of the conduction ratio command D* of FIG. 18 is the same as the time chart of the conduction ratio command D* of FIG. In FIG. 18, the reactor current Iinv shows only the envelope for the sake of easy viewing of the drawing.

As shown in FIG. 18, the current command Iinv* and the current value FIinv substantially match, so that the current control is stably performed in the current discontinuous mode. Note that the waveform of the system current Igrid of FIG. 18 is distorted at time tm10 and time tm11 because of the difference between the timing at which the value of the polarity determination signal Ipole* in FIG. 17 is switched and the zero-cross timing of the reactor current Iinv. Is. In order to adjust the zero-cross timing of the reactor current Iinv and improve the waveform distortion of the system current Igrid, in addition to the current command Iinv*, the waveform distortion of the reactor current Iinv and the alternating waveform waveform are input to the switching mode switch 5. The polarity determination signal Ipole* may be adjusted by adding at least one of the power per cycle.

As described above, according to the power conversion device according to the first embodiment, current control can be stabilized in the current discontinuous mode. As a result, power can be stably supplied in the current discontinuous mode.

Embodiment 2.
In the first embodiment, the case where the power conversion device operates in the current discontinuous mode has been described. In the second embodiment, a case will be described in which the power conversion device can selectively operate in the current discontinuous mode and the current continuous mode. The difference between the first embodiment and the second embodiment is a non-linear compensator. The configuration other than the non-linear compensator is the same, and therefore the description will not be repeated.

FIG. 19 is a functional block diagram of the non-linear compensator 702 of the power conversion device according to the second embodiment. The configuration of the non-linear compensator 702 is a configuration in which a normal mode 73 corresponding to the continuous current mode and a conduction ratio selector 74 are added to the configuration of the non-linear compensator 701 of FIG. Other than these, it is the same, and therefore the description will not be repeated.

The normal mode 73 receives the reactor voltage command VL* and the voltage value FVc and outputs the conduction ratio command Dccm*. The conduction ratio selector 74 receives the conduction ratio command Dccm*, the conduction ratio command Ddcm* output from the first non-linear compensation mode 71 or the second non-linear compensation mode 72, and the polarity determination signal Ipole*, Outputs the duty ratio command D*.

FIG. 20 is a functional block diagram showing a functional configuration of the conduction ratio selector 74. The conduction ratio selector 74 includes a comparator 741 and an exclusive OR (XOR) circuit 742. The conduction ratio commands Dccm* and Ddcm* are input to the non-inverting input terminal and the inverting input terminal of the comparator 741. The signal CompD and the polarity determination signal Ipole* output from the comparator 741 are input to the XOR circuit 742. When the signal SlctD is 1, the conduction ratio command Dccm* is output as the conduction ratio command D*. When the signal SlctD is 0, the conduction ratio command Ddcm* is output as the conduction ratio command D*.

FIG. 21 is a diagram additionally showing time charts of the flow rate commands Dccm*, Ddcm*, D*, and the polarity determination signal Ipole*. In FIG. 21, a time chart of times tm20 to tm27 is shown.

As shown in FIG. 21, at times tm20 to tm21, tm22 to tm23, tm24 to tm25, and tm26 to tm27, the flow rate command Dccm* is larger than Ddcm*, and thus the signal CompD is 1.

At times tm20 to tm21, tm22 to tm23, and tm26 to tm27, the polarity determination signal Ipole* is 1, so the input of the XOR circuit 742 is the signal CompD=1 and the polarity determination signal Ipole*=1, and the signal SlctD is It becomes 0. At times tm20 to tm21, tm22 to tm23, and tm26 to tm27, the flow rate command Ddcm* is selected as the flow rate command D*.

Since the polarity determination signal Ipole* is 0 from time tm24 to tm25, the input of the XOR circuit 742 is the signal CompD=1 and the polarity determination signal Ipole*=0, and the signal SlctD is 1. From time tm24 to tm25, the conduction ratio command Dccm* is selected as the conduction ratio command D*.

At times tm21 to tm22, tm23 to tm24, and tm25 to tm26, the signal CompD is 0 because the flow rate command Ddcm* is larger than Dccm*.

Since the polarity determination signal Ipole* is 1 at times tm21 to tm22, the input of the XOR circuit 742 is the signal CompD=0 and the polarity determination signal Ipole*=1, and the signal SlctD is 1. At times tm21 to tm22, the flow rate command Dccm* is selected as the flow rate command D*.

At times tm23 to tm24 and tm25 to tm26, the polarity determination signal Ipole* is 0, so the input of the XOR circuit 742 is the signal CompD=0 and the polarity determination signal Ipole*=0, and the signal SlctD is 0. At times tm23 to tm24 and tm25 to tm26, the conduction ratio command Ddcm* is selected as the conduction ratio command D*.

FIG. 22 is a diagram also showing the respective time charts of the duty ratio command D*, the reference triangular wave Carrier, the polarity determination signal Ipole*, and the PWM signals Sa and Sb. Also in FIG. 22, a time chart from time tm20 to tm27 is shown as in FIG. Also in FIG. 22, the reference triangular wave Carrier is simply drawn for the sake of easy viewing of the drawing. The reference triangular wave Carrier changes between 0 and 1 with the carrier cycle Tsw.

As shown in FIG. 22, at times tm20 to tm23 and tm26 to tm27, the PWM signal Sa changes between On and Off, and the PWM signal Sb is Off. At times tm20 to tm23 and tm26 to tm27, the polarity determination signal Ipole* is 1. At times tm20 to tm23 and tm26 to tm27, the full bridge circuit FB is operating in the first switching mode.

From time tm23 to tm26, the PWM signal Sb changes between On and Off, and the PWM signal Sa is Off. From time tm23 to tm26, the polarity determination signal Ipole* is 0. From time tm23 to tm26, the full bridge circuit FB is operating in the second switching mode.

The switching mode of the full-bridge circuit FB changes from the first switching mode to the second switching mode at time tm23, and changes from the second switching mode to the first switching mode at time tm26. At time tm23, the conduction ratio command D* is set to 1 which is the initial value of the second switching mode. The conduction ratio command D* is set to 0 which is the initial value of the first switching mode at time tm26.

FIG. 23 is a diagram additionally showing time charts of the conduction ratio command D*, the reactor current Iinv, the current value FIinv, the current command Iinv*, the voltage Vc, and the system current Igrid. Also in FIG. 23, a time chart from time tm20 to tm27 is shown as in FIG. In FIG. 22, as in FIG. 18, only the envelope of the reactor current Iinv is displayed for ease of viewing the drawing.

As shown in FIG. 23, at times tm20 to tm21, tm22 to tm24, and tm25 to tm27, there is a time zone in which reactor current Iinv becomes 0, so the power converter operates in the current discontinuous mode. There is. At times tm21 to tm22 and tm24 to tm25, there is no time zone in which reactor current Iinv becomes 0, so the power conversion device operates in the current continuous mode. At times tm21 and tm24, the operation mode of the power conversion device switches from the current discontinuous mode to the current continuous mode. At times tm22 and tm25, the operation mode of the power conversion device switches from the current continuous mode to the current discontinuous mode.

Since the current command Iinv* and the current value FIinv substantially match, the current control is stably performed even in the situation where the operation mode is selectively switched between the current discontinuous mode and the current continuous mode. There is. At times tm21 and tm24 when the operation mode of the power conversion device switches from the current discontinuous mode to the current continuous mode, the waveform of the grid current Igrid is hardly distorted. The waveform distortion of the grid current Igrid at times tm22 and tm25 when the operation mode of the power conversion device switches from the current continuous mode to the current discontinuous mode is more than the distortion of the waveform of the grid current Igrid at times tm23 and tm26 when the switching mode switches. It is suppressed.

The waveform of system current Igrid in FIG. 23 is distorted at time tm23 and time tm26 for the same reason as the waveform of system current Igrid is distorted in FIG. 18 of the first embodiment. Similar to the first embodiment, in order to adjust the zero-cross timing of the reactor current Iinv to improve the waveform distortion of the system current Igrid, in addition to the current command Iinv*, the reactor current Iinv* is input as an input of the switching mode switch 5. The polarity determination signal Ipole* may be adjusted by adding at least one of the waveform distortion and the power per AC waveform cycle.

As described above, according to the power conversion device of the second embodiment, current control can be stabilized even in a situation in which the operation mode is selectively switched between the current discontinuous mode and the current continuous mode. As a result, electric power can be stably supplied regardless of the operation mode of the power converter.

Embodiment 3.
In the first and second embodiments, the case where the flow rate command D* changes in the range of 0 to 1 has been described. In the third embodiment, a case where the flow rate command D* changes in the range of -1 to 1 will be described.

FIG. 24 is a functional block diagram showing a functional configuration of the power conversion device 300 according to the third embodiment. The configuration of the power conversion device 300 is a configuration in which the non-linear compensator 701 and the PWM controller 801 of FIG. 1 are replaced by the non-linear compensator 703 and the PWM controller 801, and the switching mode switch 5 is removed. The other configuration is the same, and therefore the description will not be repeated.

As shown in FIG. 24, the non-linear compensator 703 receives the voltage value FVc and the reactor voltage command VL*, calculates the flow rate command D*, and outputs the flow rate command D* to the PWM controller 803. Output to. The conduction ratio command D* is 0 to 1 in the first switching mode and -1 to 0 in the second switching mode. In power conversion device 300, the difference in switching mode appears as the polarity of conduction ratio command D*, and thus switching mode switch 5 and polarity determination signal Ipole* used in the first embodiment are unnecessary.

The PWM controller 803 includes comparators 831 and 832. The conduction ratio command D* is input to the non-inverting input terminal of the comparator 831. The reference triangular wave Carrier 1 (0 to 1) is input to the inverting input terminal of the comparator 831. The comparator 831 outputs the PWM signal Sa to the upper arm Q1 and the lower arm Q4. In the second switching mode, since the conduction ratio command D* is smaller than 0, the conduction ratio command D* becomes smaller than the reference triangular wave Carrier1. As a result, the PWM signal Sa that is the output of the comparator 831 becomes 0 in the second switching mode.

The non-inverting input terminal of the comparator 832 receives a value obtained by inverting the polarity of the conduction ratio command D*. The reference triangular wave Carrier 2 (0 to 1) is input to the inverting input terminal of the comparator 832. The comparator 832 outputs the PWM signal Sb to the lower arm Q2 and the upper arm Q3. In the first switching mode, since the conduction ratio command D* is larger than 0, the value obtained by reversing the polarity of the conduction ratio command D* is smaller than 0. In the first switching mode, the value obtained by reversing the polarity of the conduction ratio command D* is smaller than the reference triangular wave Carrier2. As a result, the PWM signal Sb output from the comparator 832 becomes 0 in the first switching mode.

Note that the reference triangular wave Carrier2 used for generating the PWM signal Sb may be changed in the range of -1 to 0 instead of using the value obtained by inverting the polarity of the flow rate command D*. Further, the phases of the reference triangular waves Carrier1 and Carrier2 do not have to be common to each other.

FIG. 25 is a functional block diagram showing the functional configuration of the non-linear compensator 703. As shown in FIG. 25, the non-linear compensator 703 includes a first non-linear compensation mode 710, a second non-linear compensation mode 720, a normal mode 730, and a conduction ratio selector 740.

As shown in FIG. 25, when the automatic switching signal Dpole* is 1, the first non-linear compensation mode 710 is selected, and the current-flow rate command Ddcm* calculated by the first non-linear compensation mode 710 is the current-flow rate selector. It is output to 740. When the automatic switching signal Dpole* is 0, the second nonlinear compensation mode 720 is selected, and the conduction ratio command Ddcm* calculated by the second nonlinear compensation mode 720 is output to the conduction ratio selector 740.

In the first non-linear compensation mode 710 and the second non-linear compensation mode 720, unlike the first non-linear compensation mode 71 and the second non-linear compensation mode 72 shown in FIG. 16, even if the switching mode is switched, the flow rate command D There is no need to change the * standard. The difference between the first non-linear compensation mode 710 and the first non-linear compensation mode 71 is that the reference of the flow rate command D* is not changed. The difference between the second non-linear compensation mode 720 and the second non-linear compensation mode 72 is that "1-D*" in the equations (17) to (29) is added in addition to the fact that the reference of the flow rate command D* is not changed. It is replaced with "-D*", and "1-D0" in the block diagrams shown in FIGS. 14 and 15 is replaced with "-D0".

The automatic switching signal Dpole* is 1 when the flow rate command Ddcm* is larger than 0 (positive polarity), and is 0 when the flow rate command Ddcm* is smaller than 0 (negative polarity).

Normal mode 730 receives reactor voltage command VL*, voltage value FVc, and automatic switching signal Dpole*, and outputs conduction ratio command Dccm*. When the automatic switching signal Dpole* is 1 (the conduction ratio command Ddcm* is positive), the normal mode 730 outputs a conduction ratio command Ddcm* similar to the normal mode 73 of FIG. When the automatic switching signal Dpole* is 0 (the conduction ratio command Ddcm* is negative), the normal mode 730 subtracts 1 from the conduction ratio command Dccm* and then outputs the conduction ratio command Dccm*.

The conduction ratio selector 740 receives the conduction ratio command Dccm*, the conduction ratio command Ddcm* output from the first nonlinear compensation mode 710 or the second nonlinear compensation mode 720, and the automatic switching signal Dpole*, Outputs the duty ratio command D*.

FIG. 26 is a functional block diagram showing a functional configuration of the conduction ratio selector 740. The conduction ratio selector 740 includes a comparator 751. The absolute value ABS1 of the conduction ratio command Ddcm* and the absolute value ABS2 of the Dccm* are input to the non-inverting input terminal and the inverting input terminal of the comparator 751. When the absolute value ABS1 is larger than ABS2, 1 is output from the comparator 751 and the flow rate command Dccm* is output as the flow rate command D*. When the absolute value ABS2 is larger than ABS1, the conduction ratio command Ddcm* is output as the conduction ratio command D*. That is, the conduction ratio selector 740 selects the conduction ratio command Dccm* or Ddcm* having the smaller absolute value as the conduction ratio command D*.

FIG. 27 is a diagram collectively showing time charts of the conduction ratio commands Dccm*, Ddcm*, and D* in the current discontinuous mode. FIG. 27 shows a time chart from time tm30 to time tm33.

As shown in FIG. 27, from time tm30 to tm33, the absolute value of the flow rate command Ddcm* is smaller than the absolute value of the flow rate command Dccm*, so the flow rate command Ddcm* is the flow rate command D. Selected as * The flow rate command D* changes in the range of 0 to 1 at times tm30 to tm31 and tm32 to tm33, and is positive. The conduction ratio command D* changes in the range of -1 to 0 at the times tm31 to tm32 and has a negative polarity.

FIG. 28 is a diagram additionally showing time charts of the duty ratio command D*, the reference triangular waves Carrier1 and Carrier2, the automatic switching signal Dpole*, and the PWM signals Sa and Sb. Similarly to FIG. 27, FIG. 28 also shows a time chart from time tm30 to tm33. Also in FIG. 28, the reference triangular waves Carrier1 and Carrier2 are simply drawn for the sake of easy viewing of the drawing. Each of the reference triangular waves Carrier1 and Carrier2 changes between 0 and 1 with the carrier cycle Tsw.

As shown in FIG. 28, at times tm30 to tm31 and tm32 to tm33, the PWM signal Sa changes between On and Off, and the PWM signal Sb is Off. At times tm30 to tm31 and tm32 to tm33, the automatic switching signal Dpole* is 1. At times tm30 to tm31 and tm32 to tm33, the conduction ratio command D* has a positive polarity, and the full bridge circuit FB operates in the first switching mode.

At times tm31 to tm32, the PWM signal Sb changes between On and Off, and the PWM signal Sa is Off. At times tm31 to tm32, the automatic switching signal Dpole* is 0. At times tm31 to tm32, the conduction ratio command D* has a negative polarity, and the full bridge circuit FB is operating in the second switching mode.

FIG. 29 is a diagram additionally showing time charts of the duty ratio command D*, the reactor current Iinv, the current value FIinv, the current command Iinv*, the voltage Vc, and the system current Igrid. Similarly to FIG. 28, FIG. 29 also shows a time chart from time tm30 to time tm33. In FIG. 29 as well, as in FIG. 18, the reactor current Iinv displays only the envelope for ease of viewing the figure.

As shown in FIG. 29, current command Iinv* and current value FIinv substantially match, so that current control is stably performed in the current discontinuous mode. The waveform distortion of the system current Igrid at the times tm31 and tm32 at which the switching modes are switched is suppressed more than the distortion of the waveform of the system current Igrid at the times tm10 and tm11 in FIG.

FIG. 30 is a diagram also showing respective time charts of the conduction ratio commands Dccm*, Ddcm*, and D* when the current discontinuous mode and the current continuous mode are selectively switched. In FIG. 30, a time chart of times tm40 to tm47 is shown.

As shown in FIG. 30, at times tm40 to tm41, tm42 to tm44, and tm45 to tm47, the absolute value of the flow rate command Ddcm* is smaller than the absolute value of the flow rate command Dccm*. The conduction ratio command Ddcm* is selected as the conduction ratio command D*. At times tm41 to tm42 and tm44 to tm45, the absolute value of the flow rate command Dccm* is smaller than the absolute value of the flow rate command Ddcm*, so the flow rate command Dccm* is the flow rate command Dccm*. Selected as *

FIG. 31 is a diagram also showing time charts of the duty ratio command D*, the reference triangular waves Carrier1 and Carrier2, the automatic switching signal Dpole*, and the PWM signals Sa and Sb. Similar to FIG. 30, FIG. 31 also shows a time chart from time tm40 to tm47. Also in FIG. 31, the reference triangular waves Carrier1 and Carrier2 are simply drawn for the sake of easy viewing of the drawing. Each of the reference triangular waves Carrier1 and Carrier2 changes between 0 and 1 with the carrier cycle Tsw.

As shown in FIG. 31, the PWM signal Sa changes between On and Off and the PWM signal Sb is Off at times tm40 to tm43 and tm46 to tm47. At times tm40 to tm43 and tm46 to tm47, the automatic switching signal Dpole* is 1. At times tm40 to tm43 and tm46 to tm47, the conduction ratio command D* has a positive polarity, and the full bridge circuit FB operates in the first switching mode.

From time tm43 to tm46, the PWM signal Sb changes between On and Off, and the PWM signal Sa is Off. From time tm43 to tm46, the automatic switching signal Dpole* is 0. From time tm43 to tm46, the conduction ratio command D* has a negative polarity, and the full bridge circuit FB is operating in the second switching mode.

FIG. 32 is a diagram additionally showing time charts of the duty ratio command D*, the reactor current Iinv, the current value FIinv, the current command Iinv*, the voltage Vc, and the system current Igrid. Similarly to FIG. 31, FIG. 32 also shows a time chart at times tm40 to tm47. Also in FIG. 32, the reactor current Iinv displays only the envelope for the sake of easy viewing of the drawing.

As shown in FIG. 32, at times tm40 to tm41, tm42 to tm44, and tm45 to tm47, there is a time zone in which reactor current Iinv is 0, so power conversion device 300 operates in the current discontinuous mode. ing. At times tm41 to tm42 and tm44 to tm45, since there is no time zone in which reactor current Iinv becomes 0, power conversion device 300 operates in the continuous current mode. At times tm41 and tm44, the operation mode of power conversion device 300 switches from the current discontinuous mode to the current continuous mode. At times tm42 and tm45, the operation mode of power conversion device 300 switches from the current continuous mode to the current discontinuous mode.

Since the current command Iinv* and the current value FIinv substantially match, the current control is stably performed even in the situation where the operation mode is selectively switched between the current discontinuous mode and the current continuous mode. There is. The distortion of the waveform of the system current Igrid at the times tm43 and tm46 when the switching mode is switched is suppressed more than the distortion of the waveform of the system current Igrid at the times tm23 and tm26 of FIG.

As described above, according to the power conversion device according to the third embodiment, current control can be stabilized even in a situation where the operation mode is selectively switched between the current discontinuous mode and the current continuous mode. As a result, electric power can be stably supplied regardless of the operation mode of the power converter. Further, according to the power converter of the third embodiment, distortion of the system current at the current zero cross (time when the switching mode is switched) can be suppressed more than in the first and second embodiments.

The embodiments disclosed this time are also planned to be implemented in an appropriate combination as long as there is no contradiction. The embodiments disclosed this time are to be considered as illustrative in all points and not restrictive. The scope of the present invention is shown not by the above description but by the scope of the claims, and is intended to include meanings equivalent to the scope of the claims and all modifications within the scope.

1 inverter, 2 filter, 3 DC power supply, 4 AC power supply, 5 switching mode switcher, 6 current controller, 7 non-linear compensator, 701, 702, 703 non-linear compensator, 51 current sensor, 52 voltage sensor, 71, 710 1st non-linear compensation mode, 72,720 2nd non-linear compensation mode, 73,730 normal mode, 74,740 conduction ratio selector, 100,300 power converter, 741,751,831,832 comparator, 742 XOR circuit, 800, 801, 803 PWM controller, C capacitor, D1 to D4 free wheel diode, FB full bridge circuit, L reactor, LG1 first leg, LG2 second leg, LPF1, LPF2 low pass filter, Lo system impedance, P1 first Terminal, P2 2nd terminal, P3 3rd terminal, P4 4th terminal, Q1, Q3 upper arm, Q2, Q4 arm.

Claims (8)

In a current discontinuous mode, a power conversion device for converting a DC voltage from a DC power supply into an AC voltage,
First and second terminals respectively connected to the positive electrode and the negative electrode of the DC power supply;
Third and fourth terminals for outputting the alternating voltage,
A full-bridge circuit including first and second legs connected in parallel between the first terminal and the second terminal;
A first connection point between the upper arm and the lower arm of the first leg, and a reactor connected between the third terminal,
A current controller configured to calculate a target voltage value across the reactor so that a current value flowing through the reactor becomes a target current value,
A non-linear compensator configured to receive the target voltage value and the output voltage value output between the third terminal and the fourth terminal and calculate a target conduction ratio;
A PWM controller configured to output a PWM signal to the full bridge circuit in response to a target conduction ratio from the non-linear compensator to operate the full bridge circuit in a first or second switching mode. ,
The fourth terminal is connected to a second connection point between the upper arm and the lower arm of the second leg,
In the current discontinuous mode, there is a time period during which no current flows in the reactor,
In the first switching mode, the upper arm of the first leg and the lower arm of the second leg are switched in synchronization with each other, and the lower arm of the first leg and the upper arm of the second leg are Is considered non-conducting,
In the second switching mode, the lower arm of the first leg and the upper arm of the second leg are switched in synchronization with each other, and the upper arm of the first leg and the lower arm of the second leg are separated from each other. Is considered non-conducting,
The non-linear compensator, in the current discontinuous mode, uses the target voltage value, the target conduction ratio before one sampling time, and the output voltage value before one sampling time to determine the target conduction ratio for the current sampling time. A power converter configured to calculate. In the current discontinuous mode, the non-linear compensator calculates an output voltage change amount one sampling time ago, multiplies the output voltage change amount by a first correction term, and calculates a first correction value. The correction value is added to the target voltage value to calculate the second correction value, and the second correction value is multiplied by the second correction term to calculate the change rate of the flow rate, and the target flow rate of the previous sampling time is calculated. The power conversion device according to claim 1, wherein the power conversion device is configured to calculate the target flow rate of the current sampling time by adding the change rate of the flow rate to. The first correction term is different between the first switching mode and the second switching mode,
The power conversion device according to claim 2, wherein the second correction term is different between the first switching mode and the second switching mode. The power conversion device is operable by selectively switching between the current discontinuous mode and the current continuous mode in which current continuously flows through the reactor,
In the current continuous mode, the non-linear compensator is configured to use the target voltage value and the output voltage value one sampling time before to calculate a target conduction ratio for the current sampling time. Item 5. The power conversion device according to any one of items 1 to 3. The PWM controller is
When the polarity of the target current value is positive, the full bridge circuit is operated in the first switching mode,
The power converter according to any one of claims 1 to 4, which is configured to operate the full bridge circuit in the second switching mode when the polarity of the target current value has a negative polarity. The PWM controller is
When the polarity of the target conduction ratio is positive, the full bridge circuit is operated in the first switching mode,
The power converter according to any one of claims 1 to 4, wherein the full bridge circuit is configured to operate in the second switching mode when the polarity of the target conduction ratio is negative. .. Further comprising a low-pass filter for receiving a current value flowing through the reactor,
The current controller is configured to calculate the target voltage value based on a comparison result between the current value that has passed through the low-pass filter and the target current value. The power converter according to item 1. The current controller, based on the result of comparison between the average value of the current value flowing through the reactor and the target current value in the cycle of the reference triangular wave used by the PWM controller to generate the PWM signal, the target voltage value. The power conversion device according to claim 1, wherein the power conversion device is configured to calculate.