JP4677641B2 - Reactor loss measuring device and measuring method thereof - Google Patents

Description

  The present invention relates to a reactor loss measuring apparatus and a measuring method thereof, and more particularly to a reactor loss measuring apparatus and a measuring method thereof that devise a method for calculating the iron loss of a reactor in a filter circuit used for a PWM inverter or the like.

  In recent years, power converters such as PWM inverters used in hybrid cars, electric cars, trains, and the like have been developed with the aim of reducing size, increasing output, reducing weight, reducing loss, and reducing costs. Under such circumstances, it is particularly important to reduce loss in power conversion such as a PWM inverter. Conventionally, the magnetic loss in the reactor of the filter circuit of the PWM inverter, the DC chopper circuit, the smoothing reactor of the DC-DC converter, or the like is estimated from the iron loss data obtained by the excitation method using the sine wave voltage. It was a thing.

  Therefore, a method for calculating the iron loss in the reactor of the filter circuit of the PWM inverter has been proposed (see, for example, Non-Patent Document 1). The proposed method for calculating iron loss will be described below.

FIG. 11 shows a circuit example of a general single-phase inverter. In the figure, Q1 to Q4 are switching elements driven by PWM. Note that illustration of the PWM drive circuit is omitted. A filter circuit including a reactor L and a capacitor C is connected to the output side of the inverter, and power having a predetermined frequency is supplied to the load resistor R through the filter circuit. In the figure, i _L is the current flowing through the reactor L, V _L represents the voltage across the reactor L, respectively, i _R is the current flowing through the load resistor R, V _R is across the load resistor R Represents the voltage of each.

FIGS. 12 and 13 show operation waveforms of the output cycle T of each part when this single-phase inverter is operated by unipolar mode subharmonic modulation. 12A shows the waveform of the voltage V _PWM between the connection point of the switch elements Q1 and Q3 and the connection point of the switch elements Q2 and Q4, and FIG. 12B shows the current i _R flowing through the load resistor R. The waveform (c) shows the waveform of the voltage V _L across the reactor L, and (d) shows the current i _L flowing through the reactor L for one output cycle T.

FIG 13 (e), the change in the intensity H of the magnetic field which is calculated on the basis of the current _{i L} shown in FIG. 12 (d) is then, in FIG. 13 (f), the voltage V shown in FIG. 12 (c) The change of the magnetic flux density B calculated based on _L is shown for each output period T.

  Here, when a magnetic flux density-magnetic flux strength curve (BH curve) for one output period T is created based on the magnetic field strength H of FIG. 13E and the magnetic flux density B of FIG. As shown in FIG. FIG. 14 is a simulation diagram of an actual BH motion trajectory for one output period. Here, a portion indicated by an arrow represents one switching ripple (referred to as a dynamic minor loop) included in one output cycle T.

  Further, FIG. 15 shows a BH curve (dynamic minor loop of 2 ripple cycles) related to 2 switching ripples in 1 output cycle T regarding the filter reactor L. Here, based on the waveforms shown in each of the above figures, a method for calculating the iron loss of a conventional filter reactor will be described below.

As a calculation period for calculating the iron loss of the reactor of one output cycle T, the bottom peak value adjacent to the bottom peak value (minimum value in each switching ripple) of the current i _L flowing through the reactor shown in FIG. The interval time is Δt. A dynamic minor loop corresponding to the time Δt is extracted and a simplified BH curve is shown in FIG.

Here, if the reactor L is lossless, the dynamic minor loop draws the straight line PQ and the straight line QR, and therefore the area of the curved portion swelled from the straight line PQ and the straight line QR (hereinafter referred to as the pseudo minor loop area). referred to) is, was the iron loss a _r by the switching ripple. The iron loss due to switching ripple is the amount of switching ripple obtained by subtracting the contribution due to iron loss A _b (also referred to as the major loop area) due to the fundamental frequency of output current i _R or the frequency component near the fundamental frequency from reactor total iron loss. It means only iron loss. In addition, the time average of the bias magnetic field H at the time Δt that is the calculation section is set as the bias magnetic field H ₀ in the calculation section.

Similarly, with respect to the half output period T / 2 shown in FIG. 12, the pseudo minor loop area _Ar and the bias magnetic field are determined H ₀ for each ripple period, that is, for each calculation section, on the H ₀ -A _r plane. The one plotted in is called the loop area trajectory. An example of this loop area trajectory is shown in FIG. 17, the horizontal axis, a bias magnetic field H _0, and the vertical axis represents the pseudo-minor loop area A _r, respectively, the bias magnetic field H ₀ and the pseudo minor loop area A _r are obtained for each calculation interval Plotted with circles.

This loop area trajectory represents the iron loss due to switching ripple per unit volume with respect to each bias magnetic field H ₀ in a circle for each ripple period, and the operation method of the iron core material constituting the reactor and the operation method of the inverter. It is known to draw different trajectories by changing, and it is known to be very useful in evaluating the relationship between the inverter operation control method and the iron loss of the reactor.

Journal of the Institute of Electrical Engineers of Japan, SPC-06-37, “Evaluation of Iron Loss and Loss Reduction Technique for Filter Reactor for Single-phase Voltage Type PWM Inverter Circuit”

However, the conventional iron loss calculation method has the following problems.
First, as a first problem, when the conventional calculation method is applied not only to the half output period T / 2 but also to all of the output periods, the iron of one output period T that is symmetrical around the time T / 2 is used. When trying to calculate the loss, as shown in FIG. 12, the time period for calculating the iron loss of the reactor of one ripple period with the time T / 2 as a boundary is determined from the time Δt indicating the adjacent bottom peak value from the bottom peak value. In the time spanning the time T / 2, the bottom peak value is changed to the top peak value (the maximum value in each switching ripple), and in the subsequent time, the top peak value is changed to the time Δt ′ indicating the adjacent top peak value. Must. Therefore, the calculation process has become very complicated.

Further, as a second problem, as shown in FIG. 16, if the reactor is lossless, the dynamic minor loop draws a straight line PQ and a straight line QR, and therefore swells from the straight line PQ and the straight line QR. area of the curved portion had the iron loss a _r by switching ripple, but this calculation method, the following reasons, were not able to calculate the accurate only the iron loss by the switching ripple.

A method for accurately calculating the iron loss due to the switching ripple will be described below. Figure 18 is the switching ripple of each shown in FIGS. 12 and 13, i.e., in one calculation period _t p ~t _r, reactor L to flow current _{i L,} the reactor L voltage _{V L} across the, reactor L It is the wave form diagram which extracted and displayed the waveform which concerns on the magnetic flux density B which arises in the bias magnetic field H and the reactor L which are applied to. 18 (a) is the same as FIG. 12 (c), FIG. 18 (b) is the same as FIG. 12 (d), FIG. 18 (c) is the same as FIG. 13 (e), and FIG. ) Respectively. Interval t _p ~t _r above, which corresponds to the interval Δt in FIG. 12 coincides with the ripple period of the switching elements of the inverter.

  FIG. 19 is a diagram in which the magnetic field H and the magnetic flux density B corresponding to FIG. 18 are drawn on the BH plane. The dynamic minor loop PQR related to one switching ripple in one output period T, and the switching ripple. The major loop P′R ′ based on the fundamental frequency component is shown.

1 iron loss _{A r} by switching ripple of the ripple period, originally in FIG. 20, a closed curve JPQRKJ (hereinafter, a closed curve C hereinafter) from the enclosed area A (hereinafter, referred to as dynamic minor loop area), low frequency It can be accurately obtained by reducing the area A _b (hereinafter referred to as the major loop area) surrounded by the closed curve J′P′R′K′J ′ (hereinafter referred to as the closed curve C _b ) corresponding to the component. .

Alternatively, in FIG. 18, for a switching ripple of one ripple period, the fundamental frequency current component I _{Lb of the} current i _L is obtained from the electric energy obtained by integrating the product of the current i _L flowing through the reactor and the voltage V _L at both ends of the reactor. When, if Genzure the amount of power obtained from the integral of the product of the fundamental frequency voltage component V _Lb of the voltage V _L, similarly, the iron loss a _r can be determined accurately.

The following describes the reason why the iron loss A _r can be determined accurately.
In FIG. 20, the H component and the B component related to each point are shown in parentheses shown in the vicinity of each point. In FIGS. 18C and 18D, symbols of the H component and the B component corresponding to each point are given. Here, as is generally well known, the area A of the closed curve C shown in FIG. 20 can be expressed by the following equation (1) using the magnetic field H and the magnetic flux density B.

Further, it is known that the current i _L flowing through the reactor is proportional to the magnetic field H, and the voltage V _L across the reactor is proportional to the time differentiation of the magnetic flux density B. Therefore, the above equation (1) can be rewritten into the following equation (2).

The above equations (1) and (2) show that the area A of the closed curve C shown in FIG. 20 is proportional to the amount of power generated in the reactor of one ripple period, that is, the total iron loss. Yes. The same is also true for the closed curve C _b corresponding to the basic frequency component. Area _{A b} of the closed curve _{C b} shown in FIG. 20, the fundamental frequency current component of the current _{i L} flowing through the reactor _{i Lb,} if the fundamental frequency voltage component of the voltage _{V L} of the reactor across the _{V Lb,} the following ( 3) Equation holds.

This is because the area A _b of the closed curve C _{b in} FIG. 20 corresponding to the fundamental frequency component corresponds to the electric energy of the fundamental frequency component generated in the reactor L in the switching ripple of one ripple period, that is, the iron loss related to the fundamental frequency component. It shows that it is proportional.

Therefore, 1 iron loss A _r by switching ripple of the ripple period, as in the prior art, rather than calculated from the area of the bulging curved portion from the straight line PQ and the line QR shown in FIG. 16, surrounded by a closed curve C in FIG. 20 from the area a, it is possible to accurately determine if Genzure a closed curve C _b surrounded by the area a _b corresponding to the basic frequency component.

Alternatively, not only from the subtraction of the area described above, but also from the amount of power obtained by integrating the product of the current i _L flowing through the reactor having one ripple period and the voltage V _L at both ends of the reactor, the fundamental frequency current component i of i _L if Genzure the amount of power obtained from the integral of the product of the fundamental frequency voltage component V _Lb of _Lb and the voltage V _L, similarly, the iron loss a _r by switching ripple of 1 ripple cycle, also can be determined accurately I understand.

However, as in the iron loss calculation method described above, by subtracting the area A _b surrounded by the closed curve C _b corresponding to the fundamental frequency component from the area A surrounded by the closed curve C in FIG. , from the power amount determined from the integral of the product of the voltage V _L across the current i _L and a reactor flowing through the reactor of 1 ripple period, i _L fundamental frequency current component i _Lb and fundamental frequency voltage component V of the voltage V _L of by reducing the amount of power obtained from the integral of the product of the _Lb, 1 even iron loss a _r by switching ripple of the ripple period is accurately determined, the calculation method of the iron loss, the first problem described above Is not a solution.

  Therefore, in the present invention, when calculating the iron loss of the reactor of one ripple cycle over the entire output cycle T, the period for calculating the iron loss of the reactor of one ripple cycle is changed with the time T / 2 as a boundary. Even if not, the reactor loss measuring device can easily calculate the iron loss of one ripple period with symmetry around the time T / 2, and can accurately calculate only the iron loss due to the switching ripple. And it aims at providing the measuring method.

  In order to solve the above-described problems, in the reactor loss measuring apparatus according to the present invention, a storage device that stores measurement data related to a reactor current that changes due to PWM switching and a voltage across the reactor, and the reactor current that is stored in the storage means And extracting means for extracting a predetermined frequency current component and a predetermined frequency voltage component from the data of the voltage across the reactor, and calculating a magnetic field strength in the reactor by the reactor current and a magnetic field strength in the reactor by the predetermined frequency current component. , Calculating means for calculating the magnetic flux density in the reactor due to the voltage across the reactor and the magnetic flux density in the reactor due to the predetermined frequency voltage component, and the time point when the reactor current and the predetermined frequency current component intersect Adjacent odd or even number The interval between the crossover times is a loss calculation period, period calculation means for determining the loss calculation period over the output period related to the PWM switching, and for each of the calculated loss calculation periods, the calculation for the period is performed. Loss calculating means for calculating a loss of the reactor based on the strength of the magnetic field due to the reactor current and the predetermined frequency current component, and the magnetic flux density due to the voltage across the reactor and the predetermined frequency voltage component.

  In calculating the iron loss of the reactor, the intersection of the reactor current flowing through the reactor and the predetermined frequency current component is obtained for the calculation period for calculating the iron loss of the reactor of one ripple cycle, and the ripple cycle is satisfied. The interval between adjacent odd-numbered or even-numbered crossing points is set as the loss calculation period, and the calculation of the iron loss by switching ripple of one ripple period is based on the area of the dynamic minor loop in the corresponding loss calculation period. Subtract the area of the major loop due to the frequency component, or from the power amount calculated from the reactor current and the voltage at both ends of the reactor in the corresponding loss calculation period, the predetermined frequency current component related to the reactor current, and the reactor The iron loss is calculated by subtracting the amount of power calculated from the predetermined frequency voltage component of the voltage at both ends. It was so.

  As described above, according to the present invention, when calculating the iron loss of a reactor having one ripple period over the entire output period T in the PWM inverter, the iron of the reactor having one ripple period is defined with the time T / 2 as a boundary. Even without changing the period for calculating the loss, it is possible to easily calculate the iron loss of one ripple period with symmetry around the time T / 2, and accurately calculate only the iron loss due to switching ripple. can do.

  Next, an embodiment of a reactor loss measuring apparatus and measuring method according to the present invention will be described with reference to FIGS. First, the calculation method related to the iron loss of the reactor in the present embodiment will be described. Here, as a typical example of power conversion by PWM switching, the case of the single-phase inverter of unipolar mode subharmonic modulation shown in FIG. 1 will be described.

In the present embodiment, a time point at which the current i _L flowing through the reactor L and the predetermined frequency current component i _Lb cross each other is obtained, and between the time points at which adjacent odd-numbered or even-numbered times satisfy the ripple cycle. It is characterized by calculating the loss calculation period for calculating the iron loss of the reactor L of one ripple period from the time, and calculating only the loss due to switching ripple based on each loss calculation period obtained over the output period. Yes.

  Here, the crossing time point has already been obtained, and among the time points between the adjacent odd-numbered or even-numbered time points, the time of the even-numbered crossing time interval satisfies the ripple cycle. The measurement of loss according to the present embodiment will be described.

FIG. 1 shows operating waveforms of the output period T in each part when the single-phase inverter shown in FIG. 11 is operated by unipolar mode subharmonic modulation, as in FIG. 1A shows a waveform of the voltage V _PWM between the connection point of the switch elements Q1 and Q3 and the connection point of the switch elements Q2 and Q4, and FIG. 1B shows the current i _R flowing through the load resistor R. The waveform (c) shows the waveform of the voltage V _L across the reactor L, and (d) shows the current i _L flowing through the reactor L with respect to the output period T.

The waveform shown in FIG. 1 (d) includes numbers 1 to 1 in order over the entire output period T for convenience when the current i _L flowing through the reactor L intersects with the fundamental frequency current component i _Lb. 37 is attached. In this waveform example, the time between the odd-numbered crossover time obtained over the entire output cycle T is compared with the time between the even-numbered crossover time and the even-numbered crossover time. Is selected as satisfying the ripple period. In the figure, the calculation period Δt is typically shown as an interval between the even-numbered crossing points 6 and 8.

The FIG. 2 (e), the change in the intensity H of the calculated magnetic field based on the current _{i L} shown in FIG. 1 (d) is, and, in FIG. 2 (f), the voltage V of FIG. 1 (c) The change in the magnetic flux density B calculated based on _L is shown for the output period T, respectively. Here, the waveform diagrams according to the present embodiment shown in FIGS. 1 and 2 are different from those in FIGS. 12 and 13 in that, in FIGS. 12 and 13, the vertical broken line indicates the ripple of the switching element of the inverter. In the case of the present embodiment, in contrast to the period, the calculation period for calculating the iron loss of the reactor of one ripple period indicated by the vertical broken line is the calculation period between the even-numbered crossover points. Is Δt as shown in the figure.

  In such a calculation period, as easily understood from the waveform shown in FIG. 1, when calculating the ripple loss of one ripple period, that is, the reactor in each calculation period, over the entire output period T, As in the conventional case, without changing the period for calculating the iron loss of the reactor of one ripple period with the time T / 2 as a boundary, the iron loss of one ripple period that is symmetrical around the time T / 2 Can be calculated.

Further, in the present embodiment, the calculation of the iron loss by the switching ripple of one ripple period is performed by subtracting the area of the major loop from the area of the dynamic minor loop of the corresponding calculation period or the corresponding calculation period. From the amount of electric power calculated from the reactor current i _L flowing through the reactor _L and the reactor voltage V _{L at} both ends of the reactor _L , the fundamental frequency current component i _{Lb of the} reactor current i _L and the fundamental frequency voltage component V _{Lb of the} reactor voltage V _L. The calculation method is to subtract the amount of power calculated from.

5, the ripple period corresponding to the calculation period Δt shown in FIGS. 1 and 2, i.e., 1 Calculation Period t _d in ~t _e, current flows through the reactor L i _L, the voltage across the reactor L V _L , a bias magnetic field H applied to the reactor L, and a waveform relating to a magnetic flux density B generated in the reactor L are extracted. 15 (a) is the same as FIG. 1 (c), FIG. 15 (b) is the same as FIG. 1 (d), FIG. 15 (c) is the same as FIG. 2 (e), and FIG. ) Respectively. The above-described periods t _{d to} t _e correspond to the above-described calculation period Δt.

  3 and 4 depict a magnetic field H and a magnetic flux density B corresponding to FIG. 5 on the BH plane, showing a dynamic minor loop DQRE and a major loop D′ E ′ having one ripple period. ing. The dynamic minor loop and the major loop shown in FIGS. 3 and 4 are different from the shapes of the dynamic minor loop and the major loop in the conventional case shown in FIGS. 19 and 20 in that the reactor has one ripple period. This is because the method for determining the calculation period for calculating the iron loss is different from the conventional method.

In this way, even if the calculation period is determined differently, the same loss calculation method as described above can be adopted, and the iron loss α _r due to the switching ripple of one ripple period is as shown in FIG. The area α surrounded by the closed curve JDQREKJ (hereinafter referred to as the closed curve Γ) is surrounded by the closed curve J′D′E′K′J ′ (hereinafter referred to as the closed curve Γ _b ) corresponding to the fundamental frequency component. By reducing the area α _b , the iron loss α _r can be accurately obtained.

Alternatively, as shown in FIG. 5, the fundamental frequency current component i of the reactor current i _L is obtained from the amount of power obtained by integrating the product of the current i _L flowing through the reactor having one ripple period and the voltage V _L at both ends of the reactor. and _Lb, if Genzure the amount of power obtained from the integral of the product of the fundamental frequency voltage component V _Lb of the reactor voltage across V _L, likewise, be obtained iron loss alpha _r according to reactor 1 ripple cycle accurately it can. In FIG. 4, symbols in parentheses shown in the vicinity of each point represent the H component and B component of each point. Also in the waveform diagrams shown in FIGS. 5C and 5D, the symbols of the H component and the B component are added as in the illustration of FIG.

  Here, the area α of the closed curve Γ shown in FIG. 4 can be expressed by the equation (4), similarly to the above equation (1), using the magnetic field H and the magnetic flux density B.

Similarly to the above equation (2), the current i _L flowing through the reactor is proportional to the magnetic field H, and the voltage V _L across the reactor is proportional to the time derivative of the magnetic flux density B. Equation (4) can be rewritten as the following equation (5).

From the above equations (4) and (5), it is shown that the area α of the closed curve Γ shown in FIG. 4 is proportional to the amount of power generated in the reactor of one ripple period, that is, the total iron loss. Yes. Similarly, this is also true for the closed curve Γ _b corresponding to the fundamental frequency component. The area α _b of the closed curve Γ _b shown in FIG. 4 is expressed as follows, assuming that the fundamental frequency current component of the current i _L flowing through the reactor is i _Lb and the fundamental frequency voltage component of the voltage V _L across the reactor is V _Lb. Similar to equation (3), the following equation (6) holds.

According to the above equation (6), the area α _b of the closed curve Γ _b shown in FIG. 4 corresponding to the fundamental frequency component is related to the electric energy of the fundamental frequency component generated in the reactor of one ripple period, that is, the fundamental frequency component. This shows that it is proportional to iron loss.

Therefore, even by the reactor loss calculation method according to the present embodiment, the iron loss α _r due to the switching ripple of one ripple period is the closed curve corresponding to the fundamental frequency component from the area α surrounded by the closed curve Γ shown in FIG. It can be accurately obtained by a calculation method of reducing the area α _b surrounded by Γ _b .

Or, as apparent from the above formulas (5) and (6), not only from the subtraction of the area, but the product of the current i _L flowing through the reactor of one ripple period and the voltage _VL at both ends of the reactor. from the power amount obtained from integration, it is a method of reducing the amount of power obtained from the integral of the product of the low frequency component V _Lb of the low-frequency component i _Lb and V _L of the i _L, similarly, accurately iron loss alpha _r It shows that it can be obtained.

  Therefore, according to the present embodiment, when calculating the iron loss of the reactor of one ripple cycle over the entire output cycle T, the iron loss of the reactor of one ripple cycle with the time T / 2 as a boundary as in the conventional case. Even without changing the period for calculating the time, it is possible to easily calculate the iron loss of one ripple period with symmetry around the time T / 2, and to accurately calculate only the iron loss due to switching ripple Can do.

  The reactor loss calculation method according to this embodiment has been described above. Next, a reactor loss measurement apparatus using the reactor loss calculation method according to this embodiment will be described. An outline of the system configuration of the reactor loss measuring apparatus is shown in FIG.

The reactor loss measuring device of this embodiment includes a control device 1, an input device 2, a display device 3, a storage device 4, and a measuring device 5. The control device 1 has a central processing unit CPU, and executes various arithmetic processes according to a stored program by an input operation by an operator from the input device 2. The display device 3 displays an operation waveform and the like as a result of arithmetic processing by the control device 1. Information measured by the measuring device 5 is input to the control device 1 and stored in the storage device 4. As shown in FIG. 11, the measuring device 5 includes a reactor current detection unit 51 that detects a current i _L flowing through the reactor L, and a reactor voltage detection unit 52 that detects a voltage V _L at both ends of the reactor L. It is done.

  The control device 1 includes a fundamental frequency component extraction unit 11, a magnetic flux density calculation unit 12, a bias magnetic field calculation unit 13, a period calculation unit 14, a loss calculation unit 15, and a loop area locus drawing unit 16. Each of these means is for realizing the reactor loss calculation method of the present embodiment, and processing is executed according to a program.

  FIG. 7 shows specific operation waveforms of each part in the single-phase inverter circuit shown in FIG. In the filter circuit connected to this single-phase inverter circuit, the reactor L having a ferrite core and an inductance L = 1 [mH] was used. Other circuit constants were a load resistance R = 100 [Ω], a capacitor C = 10 [μF], and a power supply voltage E = 50 [V]. Then, when the single-phase inverter is operated by unipolar mode subharmonic modulation with an output frequency of 50 [Hz], a switching frequency of 20 [kHz], and a modulation degree of 70 [%], the output cycle of each part T = 20 [msec] The operation waveform is shown.

A current IL flowing through the reactor _L is measured by a current probe (not shown) which is a reactor current detection means 51, and a voltage V _L across the reactor is a ferrite core of the reactor L which is a reactor voltage detection means 52. The voltage induced in the secondary winding wound around is measured. For example, when the voltage across the reactor is directly measured with a differential probe, etc., if the loss calculation is performed using the measurement result, not only the reactor iron loss but also the loss including the copper loss is measured. Will do. Therefore, in the present embodiment, the voltage across the reactor indicates the voltage V _L induced in the secondary winding wound around the ferrite core of the reactor L.

In the basic-frequency component extraction unit 11, from the input storage device stored in the 4 reactor current i _L and reactor voltage V _L, the respective fundamental frequency current components i _Lb and fundamental frequency voltage components V _Lb, the current i _L and the voltage Each V _L is extracted by applying a generally well-known Fourier transform. Threshold frequency of the fundamental frequency component used for extraction at that time, was 50 [Hz] of the fundamental frequency of the output current i _R. In addition, it is assumed that the measurement phase error caused by the measurement system of the current i _L and the voltage V _L is canceled before the measurement.

Next, the bias magnetic field calculation means 13 calculates the bias magnetic field H applied to the reactor L and the fundamental frequency magnetic field component _Hb by the following well-known formula (7), and the magnetic flux density calculation means: 12, the magnetic flux density B generated in the reactor L and its fundamental frequency density component B _b are obtained from the following equation (8) that is generally well known.

Figure 8 shows an enlarged waveform excerpted part of the reactor current i _L in FIG. In FIG. 8, among the points indicated by circles, the point between the two circles indicates a calculation period for calculating the iron loss of the reactor of one ripple period according to the present embodiment, and is adjacent to the ripple period. It shows an intersection between the odd-numbered current i _L and the basic frequency current component i _Lb fit.

The crossing point of this circle is obtained by the period calculation means 14, and the absolute value of the difference between the current i _L and the fundamental frequency current component i _Lb is minimized over the entire output period T = 20 [msec]. Is obtained every half ripple period. Of the obtained odd-numbered or even-numbered crossing time intervals adjacent to each other, the one that satisfies the ripple cycle is selected as the calculation period.

  Based on the iron loss of the switching ripple calculated by the loss calculating means 15 using the odd-numbered adjacent crossing time intervals satisfying the ripple period thus determined, that is, the calculation period in the present embodiment, the loop area is determined. The drawing means 16 creates loop area trajectory data, and the created loop area trajectory data is drawn on the screen of the display device 3. Examples of the drawn loop area trajectory are shown in FIGS.

The iron loss of the switching ripple of the loop area locus shown in FIG. 10 is obtained by a method of subtracting the area of the major loop from the area of the dynamic minor loop in each calculation period of the present embodiment. Similarly, the iron loss due to the switching ripple of the loop area trajectory is determined based on the electric energy obtained from the integration of the product of the current i _L flowing through the reactor and the voltage V _L at both ends of the reactor in each calculation period of the present embodiment. _L and fundamental frequency current component i _Lb of, those obtained by the method of reducing the amount of power obtained from the integral of the product of the fundamental frequency voltage component V _Lb of the voltage V _L.

The bias magnetic field H ₀ shown in FIG. 8 and FIG. 9 is a time average of the magnetic field H in each calculation period, and thus was obtained by the bias magnetic field calculation means 13 by the following equation (9).

Iron loss P _ra by switching ripple per unit volume in Figure 10, from the area of the dynamic minor loops at each calculation period, with losses alpha _r calculated by subtracting the area of the major loop, the loss calculation unit 15, It calculated | required by following (10) Formula.

Further, the iron loss P _rp due to the switching ripple per unit volume in FIG. 9 is obtained from the electric energy obtained by integrating the product of the current i _L flowing through the reactor and the voltage V _L at both ends of the reactor in each calculation period. _In the loss calculation means 15, the following is calculated using the W _r calculated by subtracting the electric energy obtained from the integration of the product of the fundamental frequency current component i _Lb of _{L and} the fundamental frequency voltage component V _Lb of the voltage _VL : It was calculated by equation (11).

As is clear from a comparison of the loop area trajectories in FIGS. 9 and 10, the same values are obtained for the iron losses P _ra and P _rp due to the switching ripple per unit volume for the same bias magnetic field H ₀ .

  Therefore, according to the present embodiment, when calculating the iron loss of the reactor of one ripple period over the entire output period T = 20 [msec] and drawing the loop area trajectory, the time T / 2 is conventionally used. Without changing the period for calculating the iron loss of a reactor with one ripple cycle as a boundary, it is possible to easily calculate the iron loss of one ripple cycle with symmetry around time T / 2, Only iron loss due to switching ripple can be accurately calculated.

  In the description so far, the operation and control method of the PWM inverter has been described in a state where the PWM inverter is operated in unipolar mode / subharmonic modulation using a single-phase inverter in order to simplify the description. The present invention is not limited to this operation and control method, and is also applied to the case of a power converter based on other PWM control.

As can be easily understood from the above description, of the intersection of the current i _L flowing through the reactor and its fundamental frequency current component i _Lb , which is a calculation period for calculating the iron loss of the reactor of one ripple period according to the present invention, the ripple The time of the adjacent odd-numbered crossing time intervals or the adjacent even-numbered crossing time intervals satisfying the cycle can be calculated if the ripple cycle is known in advance and there is no change with time. .

  Therefore, the present invention can also be applied to the operation and control method of the well-known single-phase bipolar mode subharmonic modulation or three-phase PWM modulation, for example, in which the ripple period is known in advance and does not change with time. .

In the example of this embodiment described above, the threshold frequency of the predetermined frequency component, has been that the fundamental frequency 50 [Hz] in the output current i _R, i _R is always output current, the fundamental frequency Since it is not composed of only a single frequency component and often becomes a distorted wave, it can be judged by looking at the result of applying Fourier transform to the current i _L flowing through the reactor L, and the output current i _R The vicinity of the fundamental frequency may be used as the threshold frequency of the low frequency component.

Furthermore, if you want to find the iron loss of the reactor above the arbitrary frequency and the iron loss below the arbitrary frequency, look at the result of Fourier transform of the current i _L flowing through the reactor _L and the voltage across the reactor V _L If the arbitrary frequency is set as the threshold frequency of the predetermined frequency component of the present invention, it can be obtained in the same manner as the iron loss due to the switching ripple of the present invention described above.

It is an operation | movement waveform of each part of an inverter explaining the calculation period of this invention. It is a waveform which showed the change of the magnetic field and magnetic flux density relevant to the operation | movement waveform of FIG. It is a BH curve of one ripple period explaining iron loss calculation by switching ripple from the area of the present invention. It is a BH curve of one ripple period explaining iron loss calculation by switching ripple from the area of the present invention. It is an operation waveform of each part of an inverter of 1 ripple period explaining iron loss calculation by switching ripple from the amount of electric power of the present invention. It is a system configuration figure concerning the reactor loss measuring device of the present invention. It is an operation | movement waveform of each part of an inverter explaining the Example of this invention. It is an enlarged view of the reactor current waveform explaining the calculation period of the present invention. It is the loop area locus which calculated the iron loss by switching ripple from the electric energy of this invention. It is a loop area locus which calculated the iron loss by switching ripple from the area of the present invention. It is a circuit diagram of a single phase inverter. It is an operation | movement waveform of each part of an inverter explaining the conventional calculation period. It is a waveform which showed the change of the magnetic field and magnetic flux density relevant to the operation | movement waveform of FIG. It is a simulation diagram of an actual BH operation locus for one output cycle. It is a BH curve of the 2 ripple period explaining the iron loss calculation by the switching ripple from the conventional area. It is a BH curve of 1 ripple period explaining the iron loss calculation by the switching ripple from the conventional area. It is the loop area locus which calculated the iron loss by switching ripple from the conventional area. It is an operation waveform of each part of an inverter of one ripple period in the conventional calculation period. It is a BH curve of one ripple period explaining iron loss calculation by switching ripple from the area of the present invention in the conventional calculation period. It is a BH curve of one ripple period explaining iron loss calculation by switching ripple from the area of the present invention in the conventional calculation period.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Control apparatus 11 Fundamental frequency component extraction means 12 Magnetic flux density calculation means 13 Bias magnetic field calculation means 14 Period calculation means 15 Loss calculation means 16 Loop area locus drawing means 2 Input device 3 Display device 4 Storage device 5 Measurement device 51 Reactor current detection means 52 Reactor voltage detection means

Claims (12)

A storage device for storing measurement data relating to the reactor current and the voltage across the reactor, which are changed by PWM switching;
Extraction means for extracting a predetermined frequency current component and a predetermined frequency voltage component from the data of the reactor current and the voltage across the reactor stored in the storage means;
The magnetic field strength in the reactor by the reactor current and the magnetic field strength in the reactor by the predetermined frequency current component are calculated, and the magnetic flux density in the reactor by the voltage across the reactor and the magnetic flux density in the reactor by the predetermined frequency voltage component are calculated. Computing means for computing;
A time point at which the reactor current and the predetermined frequency current component cross each other is obtained, and an interval between the obtained odd-numbered and even-numbered crossing time points adjacent to each other is set as a loss calculation period. A period calculation means for determining a loss calculation period;
For each calculated loss calculation period, based on the calculated magnetic field strength due to the reactor current and the predetermined frequency current component calculated for the period, and the magnetic flux density due to the voltage across the reactor and the predetermined frequency voltage component. A reactor loss measuring device comprising: a loss calculating means for calculating the loss of the reactor.   The extraction means extracts the fundamental frequency current component and the fundamental frequency voltage component obtained from the reactor current and the voltage across the reactor as the predetermined frequency current component and the predetermined frequency voltage component, or is in the vicinity of the fundamental frequency. 2. The reactor loss measuring apparatus according to claim 1, wherein a frequency current component and a frequency voltage component according to frequency are extracted.   The time period calculating means determines a time point at which an absolute value of a difference between the reactor current and the predetermined frequency current component is minimum as the crossing time point, and determines the loss calculation time period based on the crossing time point. Item 3. The reactor loss measuring apparatus according to Item 1 or 2.   The loss calculation means, from the area of the magnetic flux density-magnetic field strength curve by the strength of the magnetic field calculated from the reactor current corresponding to the loss calculation period and the magnetic flux density calculated from the voltage across the reactor, The reactor loss in the loss calculation period is obtained by subtracting the area of the magnetic flux density-magnetic field strength curve drawn by the magnetic field strength calculated from the predetermined frequency current component and the magnetic flux density calculated from the predetermined frequency voltage component. The reactor loss measuring apparatus according to any one of claims 1 to 3.   The loss calculating means calculates the amount of power calculated based on the predetermined frequency current component and the predetermined frequency voltage component from the amount of power calculated based on the reactor current corresponding to the loss calculation period and the voltage across the reactor. The reactor loss measuring device according to any one of claims 1 to 3, wherein the reactor loss in the loss calculation period is obtained by subtracting.   Based on the strength of the magnetic field calculated for each loss calculation period by the calculation means and the loss calculated for each loss calculation period by the loss calculation means, drawing data for drawing a loop area locus is created. The reactor loss measuring apparatus according to claim 1, further comprising a drawing unit. Storing measurement data relating to the reactor current and the voltage across the reactor that change due to PWM switching;
Extracting a predetermined frequency current component and a predetermined frequency voltage component from the stored data of the reactor current and the voltage across the reactor; and
The magnetic field strength in the reactor by the reactor current and the magnetic field strength in the reactor by the predetermined frequency current component are calculated, and the magnetic flux density in the reactor by the voltage across the reactor and the magnetic flux density in the reactor by the predetermined frequency voltage component are calculated. A step of calculating;
A time point at which the reactor current and the predetermined frequency current component cross each other is obtained, and an interval between the obtained odd-numbered and even-numbered crossing time points adjacent to each other is set as a loss calculation period. Determining the loss calculation period;
For each calculated loss calculation period, based on the calculated magnetic field strength due to the reactor current and the predetermined frequency current component calculated for the period, and the magnetic flux density due to the voltage across the reactor and the predetermined frequency voltage component. And a step of calculating the loss of the reactor.   In the extraction step, as the predetermined frequency current component and the predetermined frequency voltage component, the fundamental frequency current component and the fundamental frequency voltage component obtained from the reactor current and the voltage across the reactor are extracted, or in the vicinity of the fundamental frequency 8. The reactor loss measuring method according to claim 7, wherein a frequency current component and a frequency voltage component according to frequency are extracted.   The time period calculating step is characterized in that a time point at which an absolute value of a difference between the reactor current and the predetermined frequency current component is minimum is set as the crossing time point, and the loss calculation period is obtained based on the crossing time point. Item 9. The reactor loss measurement method according to Item 7 or 8.   In the loss calculation step, from the area of the magnetic flux density-magnetic field strength curve by the magnetic field strength calculated from the reactor current corresponding to the loss calculation period and the magnetic flux density calculated from the voltage across the reactor, The reactor loss in the loss calculation period is obtained by subtracting the area of the magnetic flux density-magnetic field strength curve drawn by the magnetic field strength calculated from the predetermined frequency current component and the magnetic flux density calculated from the predetermined frequency voltage component. The reactor loss measuring method according to claim 7, wherein the reactor loss is measured.   In the loss calculation step, the electric energy calculated based on the predetermined frequency current component and the predetermined frequency voltage component from the electric energy calculated based on the reactor current and the voltage across the reactor corresponding to the loss calculation period. The reactor loss measurement method according to any one of claims 7 to 9, wherein the reactor loss in the loss calculation period is obtained by subtracting.   Based on the magnetic field strength calculated for each loss calculation period in the calculation step and the loss calculated for each loss calculation period in the loss calculation step, drawing data for drawing a loop area trajectory is created. The reactor loss measuring method according to claim 7, further comprising a step.