Two-parallel modular multilevel power converter

Abstract

According to the present invention, a three-terminal leg circuit in which two two-terminal arm power circuits are connected in series is provided in each of u, v, and w phases. The three-terminal leg circuits have positive and negative terminals star-connected and connected to a DC power supply, and intermediate terminals connected to an AC power supply. When two MMC converters provided with an iron core reactor for suppressing a circulating current due to voltage unbalance between the three legs are connected in parallel to the AC power supply, the iron core reactors of the parallel MMC converters are magnetically coupled for each phase, and the arrangement, connection, and winding method of four coils per phase are optimally selected, thereby omitting six two-terminal reactors provided between the intermediate terminal and the AC power supply in order to suppress the AC circulating current between the parallel MMC converters.

Description

Two-parallel modular multilevel power converter

[0001] The present invention relates to a modular multilevel power converter. In particular, it relates to a modular multilevel power converter (hereinafter referred to as an "MMC converter" in the present invention) suitable for doubling the current capacity by connecting the AC sides of two MMC converters in parallel.

[0002] The circuit of the MMC converter consists of three three-terminal leg circuits provided for each phase. The three-terminal leg circuit has a positive terminal, a negative terminal, and an intermediate terminal. The three positive terminals are star-connected to the positive voltage terminal of the DC power supply, the three negative terminals are star-connected to the negative voltage terminal of the DC power supply, and the three intermediate terminals are connected to the terminals of the three-phase AC power supply.

[0003] The three-terminal leg circuit serially connects a two-terminal positive arm power circuit and a two-terminal positive inductive element between the positive terminal and the intermediate terminal, and serially connects a two-terminal negative arm power circuit and a two-terminal negative inductive element between the intermediate terminal and the negative terminal.

[0004] The two-terminal inductive elements on the positive and negative sides have a function of suppressing the circulating current flowing between the three-terminal leg circuits when the voltage between the three three-terminal leg circuits is unbalanced.

[0005] The positive and negative arm power circuits are formed by connecting K (K is a natural number of 2 or more) unit converters that generate the required voltage by controlling the modulation ratio of a PWM converter having an energy storage element with voltage source characteristics such as a capacitor as a voltage source in series.

[0006] The control device of the MMC converter requires capacitor voltage control to keep the 6×K capacitor voltages provided in the unit converter within a predetermined range. This is different from the control devices of conventional two-level or three-level converters. In the case of the MMC converter, as will be described later, the difference between the AC-side active power and the DC-side active power that can be independently controlled from each other becomes the charging and discharging power of the 6×K capacitors and acts on the average voltage of the 6×K capacitors.

[0007] In this invention, the current flowing from the positive terminal to the negative terminal of a three-terminal leg circuit is referred to as the through-current, the current passing through the intermediate terminal of the three-terminal leg circuit is referred to as the alternating current, and the current passing through the DC power supply is referred to as the direct current. Consequently, the direct current is equal to the zero-phase component of the through-current. These through-current, alternating current, and direct current are distinguished by the terminal of the three-terminal leg circuit they pass through, and none of them depend on their frequency components.

[0008] In conventional two-level or three-level converters, current passing through the DC power supply terminals always passes through the AC power supply terminals. In contrast, with MMC converters, AC current flows in and out only between the AC power supply terminals, and DC current flows in and out only between the DC power supply terminals. Therefore, the control device of an MMC converter can separate the control of DC current from the control of other current components. As a result, the active power on the AC and DC sides can be controlled independently of each other.

[0009] On the other hand, in the case of MMC converters, it is necessary to secure the inductive impedance values ​​required to suppress both AC current and through-current. The required impedance values ​​vary depending on the capacity of the AC system to which the MMC converter is connected, whether the connection end is part of a loop transmission network or the trailing end of a leading-end / trailing-end configuration, and the range of ground fault modes that undergo high-speed reclosing. Furthermore, it varies greatly depending on the capacitor capacity of the unit converter of the MMC converter itself.

[0010] Taking the above variations into account, the required inductive impedance for both AC and through-current circuits is approximately 10% to 20% of the impedance based on the rated voltage, current, and frequency of the MMC converter.

[0011] The inductive impedance value of an AC current circuit is the sum of the inductive impedance values ​​of the parallel circuit of the positive and negative two-terminal inductors of the MMC converter, and the transformer or generator motor that constitutes the three-phase AC power supply connected in series with this parallel circuit.

[0012] When designing transformers and generator motors based on economic rationality, the inductive impedance value should be approximately 7% to 15% of the rated value. Therefore, the inductive components of the positive and negative two-terminal inductors may be unnecessary. In this case, in order to supply reactive power from the MMC converter to the AC power source, it is necessary to compensate for the voltage drop due to the two-terminal inductors in the positive and negative arm power circuits, which has the disadvantage of leading to a larger MMC converter.

[0013] When the through-current is divided into positive-sequence, negative-sequence, and zero-sequence, the inductive impedance values ​​of the positive-sequence and negative-sequence through-current circuits become the combined impedance values ​​obtained by connecting the positive and negative two-terminal inductive elements of each phase in series and then wiring them in a three-phase configuration. Since the zero-sequence through-current is a DC current, the impedance of the DC power supply is added in series. However, the impedance of a DC power supply is generally small, and therefore cannot be expected to contribute to the inductive component impedance necessary for suppressing the through-current as described above.

[0014] From the above, it can be said that the essential function of the positive and negative two-terminal inductors constituting the MMC converter is the suppression of through-current. More specifically, the combined impedance value obtained by connecting the positive and negative two-terminal inductors of each phase in series and then connecting them in a three-phase Y-connection must have the required inductive impedance value.

[0015] The present invention relates to an inductive element for an MMC converter suitable for an MMC converter with two MMC converters connected in parallel on the AC power supply side.

[0016] When two MMC converters are connected in parallel on the AC power supply side, the AC current is divided into a current path that passes through the AC power supply terminals and a component that circulates between the intermediate terminals of each phase of the two parallel MMC converters. Similarly, when two MMC converters are connected in parallel on the DC power supply side, the DC current is divided into a current path that passes through the DC power supply terminals and a component that circulates between the positive terminals and the negative terminals of the two parallel MMC converters.

[0017] Hereinafter, in this invention, regardless of whether it is AC or DC current, the current passing through the power supply terminal will be referred to as the output current, and the current circulating between parallel converters will be referred to as the circulating current.

[0018] Based on the above, the current passing through the terminals of the two-parallel MMC converter can be divided into four categories. In the terminology of this invention, these categories are AC output current, AC circulating current, DC output current, and DC circulating current. In order to adjust the current of the two-parallel MMC converter, it is necessary to secure a unique inductive impedance value for each of the four current paths.

[0019] Patent Document 1 discloses a basic circuit configuration in which inductive elements such as reactors are provided between the positive terminal and the intermediate terminal and between the negative terminal and the intermediate terminal of the three-terminal leg circuit of an MMC converter in order to suppress the forward and reverse phase through-currents of the three-terminal leg circuit.

[0020] MMC converters are classified as voltage-type power converters that use self-extinguishing power semiconductor elements. Compared to conventional voltage-type power converters such as two-level and three-level converters, MMC converters have the advantage of halving power loss and eliminating the need for harmonic filters. On the other hand, they have the disadvantage of being larger due to the six air-core reactors housed in a magnetic shielding chamber to suppress through-current.

[0021] Patent Document 2 discloses a method for miniaturizing the six air-core reactors of Patent Document 1, in which a winding consisting of a coil connected to the positive arm power circuit and a coil connected to the negative arm power circuit is provided for every three phases, the two coils are magnetically coupled by winding the first to third winding cores concentrically for every three phases, and the zero-phase magnetic flux created by the magnetic flux flowing through the first to third winding cores is recirculated at the first and second side legs, and this three-phase five-legged reactor is provided in the MMC converter.

[0022] Patent Document 3 discloses a method for controlling the balance within an arm to maintain the balance between the capacitor voltages of the unit converters that constitute an MMC converter.

[0023] Patent Document 4 discloses a basic hierarchical configuration of a control system consisting of a PWM modulator and converter current control provided for each unit converter of an MMC converter. Specifically, it discloses a first layer provided for each unit converter, a second layer that adjusts the current of each part of the MMC converter, and a third layer consisting of a higher-level control device that outputs commands to the second layer in order to adjust the active power and reactive power output of the MMC converter. It also discloses a method for adding a second harmonic circulating current command to a fundamental wave current command.

[0024] Patent Document 5 discloses a higher-level control device for two MMC converters with their DC sides connected behind each other as frequency converters, and is particularly suitable for suppressing DC overcurrent and continuing operation in the event of a ground fault on the AC power supply side. When the higher-level control device of Patent Document 5 is combined with the reactor disclosed in Patent Document 2, the DC current excites the magnetic flux flowing through the first to third winding cores, so by suppressing DC overcurrent, magnetic saturation of the core can be prevented even in the event of a ground fault on the AC power supply side.

[0025] Patent Document 6 discloses a positive-negative balance control configuration for an MMC converter, particularly suitable for continuing operation in the event of a ground fault on the AC power supply side.

[0026] Japanese Patent Publication No. 5189105, Japanese Patent Publication No. 7312332, Japanese Patent Publication No. 4999930, Japanese Patent Publication No. 5197623, Japanese Patent Publication No. 7360559, International Publication No. 2024 / 150437, International Publication No. 2024 / 154308

[0027] Figure 9 shows the basic configuration of an MMC converter (801) using three sets of iron core reactors wound with two concentric coils as disclosed in Patent Document 2.

[0028] The circuit of the MMC converter (801) consists of three 3-terminal leg circuits (802u, 802v, 802w) provided for each phase. Each 3-terminal leg circuit has positive terminals (UP, VP, WP), negative terminals (UN, VN, WN), and intermediate terminals (UC, VC, WC). The three positive terminals (UP, VP, WP) are connected in a star configuration to the positive voltage terminal (P) of the DC power supply (804), the three negative terminals (UN, VN, WN) are connected in a star configuration to the negative voltage terminal (N) of the DC power supply (804), and the three intermediate terminals (UC, VC, WC) are connected to the 3-phase AC voltage terminals (U, V, W) of the AC power supply (803).

[0029] In the following, each phase (u, v, w) of a three-phase AC system will be represented by (x) or (X).

[0030] The 3-terminal leg circuit (802x) has a 2-terminal positive arm power circuit (821xp), a positive current transformer (823xp), and a positive 2-terminal inductor (822xp) connected in series between the positive terminal (XP) and the intermediate terminal (XC), and a negative 2-terminal inductor (822xn), a negative current transformer (823xn), and a 2-terminal negative arm power circuit (821xn) connected in series between the intermediate terminal (XC) and the negative terminal (XN).

[0031] The positive current transformers (823up, 823vp, 823wp) measure and output the current (I_up, I_vp, I_wp) of the positive arm power circuits (821up, 821vp, 821wp) to the power converter control device, and the negative current transformers (823un, 823vn, 823wn) measure and output the current (I_un, I_vn, I_wn) of the negative arm power circuits (821un, 821vn, 821wn) to the power converter control device.

[0032] The power converter control device adjusts the positive and negative arm currents (I_up, I_vp, I_wp, I_un, I_vn, I_wn) to the required current command values ​​according to the operating conditions. To this end, it outputs control commands to the positive arm power circuits (821up, 821vp, 821wp) and the negative arm power circuits (821un, 821vn, 821wn), which have control voltage source characteristics.

[0033] The current adjustment function of the power converter control device is configured on the premise that the AC power supply (803) can be approximated by a series circuit of a star-connected three-phase voltage source (831u, 831v, 831w) and an inductive element (832u, 832v, 832w). Furthermore, the DC power supply (804) is configured on the premise that it can be approximated by a series circuit of a DC voltage source (841) and a DC reactor (842).

[0034] The current adjustment function of the power converter control device can be configured by dividing the circuit configuration of the MMC converter shown in Figure 9 into five current paths with five degrees of freedom, and the circuit equation has five elements.

[0035] The first and second current paths are current paths that start from the star-shaped connection of the AC power supply (803) and divide the current equally between the star-shaped connection on the positive voltage terminal (P) side and the star-shaped connection on the negative voltage terminal (N) side.

[0036] The third, fourth, and fifth current paths are current paths in which the current (I_dc) of the DC power supply (804) is divided into three through-currents (I_cu, I_cv, I_cw) between the positive terminal (XP) and negative terminal (XN) of the three-terminal leg circuit (802x).

[0037] In the configuration shown in Figure 9, the following relationships hold between the AC power supply currents (I_u, I_v, I_w), the DC power supply current (I_dc), the positive and negative arm currents (I_up, I_vp, I_wp, I_un, I_vn, I_wn), and the through-currents of the three-terminal leg circuit (I_cu, I_cv, I_cw): Equations (1), (2), and (3).

[0038] To implement the current adjustment function, it is necessary to rearrange the seven equations (1), (2), and (3), which contain 13 current variables, into equations with 5 current variables corresponding to the current path with 5 degrees of freedom. For this purpose, the currents in each part are transformed and reversed. To avoid redundancy, the current variable (I), voltage variable (V), and flux linkage variable (λ) will be represented by the variable (f) below.

[0039] The variables (f_u, f_v, f_w) related to the AC power supply (803) are expressed by equation (4).

[0040] The variables related to the through-current (f_cu, f_cv, f_cw) are represented by Equation (5).

[0041] The variables related to the positive-side arm power circuits (821up, 821vp, 821wp) are represented by Equation (6) and Equation (7).

[0042] The variables related to the negative-side arm power circuits (821un, 821vn, 821wn) are represented by Equation (8) and Equation (9).

[0043] Obtain the circuit equations of the MMC converter shown in FIG. 9 with the five current variables (I_α, I_β, I_cα, I_cβ, I_dc) defined by Equation (4) and Equation (5) as state variables.

[0044] Here, in order to examine the circuit equations of the MMC converter shown in FIG. 9, the characteristics of the concentric two-winding reactor of Patent Document 2 are shown.

[0045] FIG. 10 shows the configuration of the concentric two-winding reactor (826x). Parts with the same numbers as in the previous FIG. 9 indicate the same parts. The description is omitted to avoid duplication. The positive-side two-terminal inductive element (822xp) and the negative-side two-terminal inductive element (822xn) constitute a concentric winding (825x) that winds around a winding core (824x) filled with a non-magnetic material in a concentric arrangement.

[0046] In FIG. 10, the positive-side two-terminal inductive element (822xp) is the inner coil in the clockwise (cw: right-handed winding) direction, and the negative-side two-terminal inductive element (822xn) is the outer coil in the counterclockwise (ccw: left-handed winding) direction.

[0047] With the above configuration, when the through-current (I_cx) of the two-terminal positive-side arm power circuit (821xp) and the two-terminal negative-side arm power circuit (821xn) flows through the concentric winding (825x), the two coils magnetize the winding core (824x) in the same direction. Hereinafter, in the present invention, the case where the concentric two coils magnetize the winding core in the same direction is called "concentric cooperative magnetization".

[0048] On the other hand, when the output current (I_x) passing through the intermediate terminal (XC) is divided into two equal parts and flows through the two-terminal positive arm power circuit (821xp) and the two-terminal negative arm power circuit (821xn) to the concentric winding (825x), the two coils excite the winding core (824x) in opposite directions. Hereinafter, in this invention, the case in which two concentric coils excite the winding core in opposite directions is referred to as "concentric differential excitation".

[0049] The left side of Figure 11 shows the magnetic flux distribution during concentric summative excitation due to the through-current (I_cx), and the right side shows the magnetic flux distribution during concentric differential excitation due to the output current (I_x). During concentric summative excitation, most of the magnetic flux flows through the winding core (824x). On the other hand, during concentric differential excitation, most of the magnetic flux flows through the cylindrical space surrounded by the inner and outer coils that make up the concentric winding (825x).

[0050] The characteristics of the reactor in Figure 10 are expressed by equation (10).

[0051] The reactor characteristics corresponding to the through-current (I_cx) during concentric summation excitation can be determined from the relationship of flux linkage that generates the electromotive force (XP-XN) between the XP terminal of the positive two-terminal inductor (822xp) and the XN terminal of the negative two-terminal inductor (822xn) due to the through-current (I_cx). The equivalent inductance per winding is [+(1 / 2) × (L_p + 2 × M_s + L_n)].

[0052] The reactor characteristics corresponding to the output current (I_x) during concentric differential excitation can be determined from the relationship of flux linkage that generates the sum voltage (XP-XN) of the terminal electromotive force (XN-XC) of the positive two-terminal inductor (822xp) and the terminal electromotive force (XN-XC) of the negative two-terminal inductor (822xn) due to the equal division of the output current (I_x) into two (I_x / 2). The equivalent inductance per winding is [+(1 / 2) × (L_p - 2 × M_s + L_n)].

[0053] The reactor characteristics of equation (10) can be expressed in terms of output current (I_x) and through-current (I_cx) as shown in equation (11).

[0054] In Figure 10, an example of a positive two-terminal inductor element (822xp) with a clockwise (CW: right-hand winding) inner coil is shown. However, even if the coil is counterclockwise (CCW: left-hand winding) or an outer coil is used, the operation will be the same as long as the coil arrangement and winding method are selected so that the through-current (I_cx) provides concentric summative excitation and the output current (I_x) provides concentric differential excitation, and this will not compromise generality.

[0055] Based on the above, the five-element circuit equation for the MMC converter shown in Figure 9 can be divided into three parts: two-element equations (12) and (13), and one-element equation (14).

[0056] The advantages and disadvantages of using the iron-core coupled reactor described in Patent Document 2 as the positive inductor (822xp) and negative inductor (822xn), and of using reactors that are magnetically independent of each other, will be explained below using the circuit equations shown in equations (12), (13), and (14).

[0057] In the case of two magnetically independent reactors, the mutual inductance (M_s) in equation (10) becomes (M_s = 0). Also, assuming that the positive and negative sides have the same characteristics, (L_p = L_n = L_0).

[0058] The ratio (k_i) of the inductance value [+(1 / 2) × (L_p - 2 × M_s + L_n)] for alternating current, based on the inductance value [+(1 / 2) × (L_p + 2 × M_s + L_n)] for through-current due to only the positive inductor (822xp) and the negative inductor (822xn), is (k_i = 1).

[0059] As a result, when considering the inductance due to the AC power supply (L_ac), the diagonal components of the inductance matrix on the left side of equations (12) and (13) are [L_ac + (+1 / 2) × L_0] for AC currents (I_α, I_β) and [(+1 / 2) × L_0] for through currents (I_cα, I_cβ). When the DC power supply inductance (L_dc) on the left side of equation (14) is negligibly small, the inductance on the left side becomes [(+2 / 3) × L_0]. The difference in the coefficients of the diagonal components on the left side of equations (12) and (13) and equation (14) between (1 / 2) and (1 / 3) is due to the difference in the number of parallel circuits.

[0060] In the MMC converter (801), the essential function required of inductive elements such as reactors is the suppression of through-currents (I_cα, I_cβ, I_dc). On the other hand, in the case of alternating currents (I_α, I_β), as described in paragraph 0012, depending on the value of the inductance (L_ac) of the transformer or generator motor windings that constitute the AC power supply (803), it is not necessary for inductive elements such as reactors to have an inductance value.

[0061] In other words, if the circuit inductance required for controlling and protecting the MMC converter (801) can be secured by the AC power supply inductance (L_ac), the disadvantage of having to enlarge the two-terminal positive and negative arm power circuits (821xp, 821xn) to compensate for the voltage drop due to the inductance (L_0) of the inductive element becomes greater.

[0062] On the other hand, in the case of an "ideal perfect coupling" where the characteristics of the positive inductor (822xp) and the negative inductor (822xn) are in equilibrium and perfectly coupled, the mutual inductance (M_s) in equation (10) becomes (M_s = L_p = L_n = L_0).

[0063] In this case, the diagonal components of the inductance matrices corresponding to (I_α, I_β) in equations (12) and (13) are (L_ac) from [L_ac + (1 / 4) × (L_p - 2 × M_s + L_n)].

[0064] On the other hand, the diagonal components of the inductance matrix corresponding to (I_cα, I_cβ) are [+(1 / 2) × (L_p + 2 × M_s + L_n)], which becomes (L_0).

[0065] Furthermore, the inductance corresponding to (I_dc) is obtained from [L_dc + (1 / 3) × (L_p + 2 × M_s + L_n)] to [L_dc + (4 / 3) × L_0].

[0066] From the above, it can be seen that in the case of an "ideal perfect coupling," inductive elements such as reactors contribute only to suppressing the through-current (I_cα, I_cβ, I_dc), which is an essential function, and do not act on the alternating current (I_α, I_β).

[0067] On the other hand, when using the core coupling described in Patent Document 2 for the positive inductive element (822xp) and the negative inductive element (822xn), even if the magnetic saturation of the core can be ignored, the effect of leakage flux generated by the two concentric coils winding a common core (824x) will appear.

[0068] The following explanation describes the case where the coil of the positive inductor (822xp) is the inner coil and the coil of the negative inductor (822xn) is the outer coil, as shown in Figure 10. The generality is not lost even when the assignment to the inner and outer coils is reversed.

[0069] The relationship between self-inductance (L_p, L_n) and mutual inductance (M_s) is L_p < M_s < L_n. As long as the design maintains cost-performance ratio rationality, the ratio k_i of the inductance value for AC current [+(1 / 2) × (L_p - 2 × M_s + L_n)] with respect to the inductance value for through-current [+(1 / 2) × (L_p + 2 × M_s + L_n)] is (k_i ≤ 0.1).

[0070] According to the configuration in Patent Document 2, characteristics closer to the ratio of "ideal perfect coupling" (k_i = 0) can be obtained than the ratio of "magnetically independent case" (k_i = 1). Furthermore, by concentrically arranging the coils of the positive inductor (822xp) and the negative inductor (822xn), even if an AC overcurrent occurs during a grid fault propagation, it does not affect the excitation of the iron core. This avoids a rapid decrease in inductance due to iron core saturation, and safely enables miniaturization by changing from an air core to an iron core. In addition, miniaturization and weight reduction can be achieved by consolidating the devices from six to three. In the case of a three-phase five-legged reactor configuration, it can be further consolidated into one unit.

[0071] Next, we will examine the inductance values ​​for each current when two MMC converters (801) are connected in parallel to an AC power supply (803) in order to double the current capacity of the MMC converter, which is the subject of the present invention.

[0072] The inductance (L_ac) of the AC power supply does not contribute to suppressing AC circulating current. If the inductance of the positive inductor (822xp) and negative inductor (822xn) of the MMC converter is insufficient, additional inductors must be added.

[0073] When suppressing through-current (I_cα, I_cβ), which is an essential function of inductive elements, based on a reasonable cost-performance ratio, an inductance value of (k_i≧0.3) is required to suppress the AC circulating current between the two parallel MMC converters (801).

[0074] When using six magnetically independent air-core reactors as the positive inductor (822xp) and negative inductor (822xn), as described in paragraph 0058 above (k_i = 1), the inductance value is sufficient to suppress AC circulating current. Therefore, no additional inductors are needed even when using two in parallel. The required number of devices is 12.

[0075] On the other hand, when the iron core coupling of Patent Document 2 is applied to the positive inductive element (822xp) and the negative inductive element (822xn), the case is as described in the previous paragraph 0069 (k_i ≤ 0.1).

[0076] Therefore, in order to ensure an inductance value of (k_i≧0.3) and suppress the AC circulating current between the two parallel MMC converters (801), it is necessary to provide six two-terminal inductors between the two sets of intermediate terminals (UC, VC, WC) and the three-phase AC voltage terminals (U, V, W).

[0077] As a result, the number of devices becomes 12 in total: 6 reactors equipped with iron cores (824x) plus 6 units for suppressing circulating current. This number is the same as when using 6 magnetically independent air-core reactors.

[0078] Therefore, when two MMC converters are connected in parallel, there is a problem in that the benefits of adopting the iron core coupling described in Patent Document 2 are largely offset.

[0079] The objective of the present invention is to solve the above problems and to realize the advantages of MMC converters using an iron core reactor wound with two concentric coils—miniaturization and improved operational continuity performance during grid fault propagation—by using a two-parallel MMC converter.

[0080] To solve the aforementioned problems, the objective is achieved by magnetically coupling a common iron core (824x) around which concentric coils of positive inductors (822xp) and negative inductors (822xn) provided for each of the two parallel MMC converters (801) are wound coaxially, and by optimally selecting the arrangement, connection, and winding method of the four coils per phase, thereby eliminating the need for six two-terminal reactors to be installed between the intermediate terminals and the AC power supply, which are necessary to suppress AC circulating current between the parallel MMC converters.

[0081] The following describes the means by which these functions are implemented.

[0082] Figure 12 shows the configuration of a reactor consisting of four coils, with two sets of concentric windings arranged coaxially on a common winding core. For the sake of explanation, the inner coil of the upper winding is referred to as the first coil, the outer coil as the second coil, the inner coil of the lower winding as the third coil, and the outer coil as the fourth coil.

[0083] When the currents in the first to fourth coils are (I_1, I_2, I_3, I_4) and the flux linkage is (λ_1, λ_2, λ_3, λ_4), the characteristics of the four-coil reactor can be expressed by equation (15) due to symmetry.

[0084] As explained using Figure 11, the upper and lower windings each have a concentric summative excitation mode and a concentric differential excitation mode. When the upper and lower windings are arranged coaxially with respect to the winding core, they can be divided into two modes: one in which they create magnetic flux in the same direction and another in which they create magnetic flux in opposite directions. In this invention, the former will be called "coaxial summative excitation" and the latter "coaxial differential excitation".

[0085] Based on the above, the magnetic flux distribution can be divided into four patterns: "two concentric patterns" x "two coaxial patterns" depending on the current direction of the four coils from the first to the fourth.

[0086] Figure 13 shows four patterns of coil current direction and magnetic flux distribution.

[0087] In all four patterns, the inductance values ​​of the four coils are equal, and these values ​​are determined by the magnetic flux distribution. The equivalent inductance value per coil (L_1) during concentric sum excitation and coaxial sum excitation is given by equation (16).

[0088] The equivalent inductance value (L_2) per coil during concentric summative excitation and coaxial differential excitation is given by equation (17).

[0089] The equivalent inductance value (L_3) per coil during concentric differential excitation and coaxial summative excitation is given by equation (18).

[0090] The equivalent inductance value (L_4) per coil during concentric differential excitation and coaxial differential excitation is given by equation (19).

[0091] In the case of reactors for MMC converters, a fundamental requirement is to design them so that the effect of magnetic saturation does not appear in the inductance value. For this purpose, a gap filled with a non-magnetic material is provided in the winding core. The ratio of the gap to the loop length of the magnetic flux is constrained by the rationality of the cost-performance ratio and the saturation magnetic flux density of the core material. As a result, the value of the ratio (ε_y) of the coil distance (h_y) between the upper and lower windings to the coil width (H_y) in the axial direction of the core, as shown in Figure 12, is limited to a range of selection due to the rationality of the cost-performance ratio. Similarly, the value of the ratio (ε_x) of the distance between the inner and outer coil centers (d_x) to the distance (D_x) between the core axis and the inner and outer coil centers is also limited to a range of selection due to the rationality of the cost-performance ratio and the insulation distance between the coils.

[0092] Figure 14 shows the change in the dimensionless equivalent inductance (L_1, L_2, L_3, L_4) with respect to the ratio (ε_y) of the coil distance (h_y) between the upper and lower windings to the coil width (H_y) in the axial direction of the iron core. The reference value for dimensionlessness is the (L_1) value when the upper and lower windings are in close contact (h_y = 0).

[0093] As the coil distance between the upper and lower windings increases and the value of (ε_y) increases, the equivalent inductance (L_2) increases during concentric summing and coaxial differential operation. The other equivalent inductance values ​​(L_1, L_3, L_4) remain approximately constant.

[0094] Figure 15 shows the change in the dimensionless equivalent inductance (L_1, L_2, L_3, L_4) with respect to the ratio (ε_x) of the distance between the centers of the inner and outer coils (d_x) to the distance (D_x) between the iron core axis and the centers of the inner and outer coils.

[0095] As the distance between the centers of the inner and outer coils increases and the value of (ε_x) increases, the equivalent inductance (L_3) for concentric differential and coaxial summation, and the equivalent inductance (L_4) for concentric differential and coaxial differential, increase. Also, the equivalent inductance (L_1) for concentric summation and coaxial summation gradually increases. On the other hand, the equivalent inductance (L_2) for concentric summation and coaxial differential remains almost constant.

[0096] When examining the relative magnitudes of the equivalent inductances (L_1, L_2, L_3, L_4), the equivalent inductance (L_1) is always maximized during concentric and coaxial summative excitation, regardless of the coil distance ratio between the upper and lower windings (ε_y) or the distance ratio between the centers of the inner and outer coils (ε_x).

[0097] The ratio of the equivalent inductance (L_2) during concentric summation and coaxial differential operation to the equivalent inductance (L_1) during concentric summation and coaxial summation can be adjusted between 0.2 and 0.4 by changing the coil distance ratio (ε_y).

[0098] The ratio of the equivalent inductances (L_3, L_4) during concentric differential operation to the equivalent inductance (L_1) during concentric summation and coaxial summation is 0.1 or less.

[0099] Based on the above, the equivalent inductances (L_1, L_2, L_3, L_4) are optimally allocated for current suppression purposes in AC output current circuits, AC circulating current circuits, through-current output current circuits, and through-current circulating current circuits.

[0100] The essential function required of a reactor is the suppression of through-output current. Prioritizing this, the equivalent inductance (L_1) for concentric and coaxial summation, which is always the maximum, is allocated accordingly.

[0101] Next, to suppress the AC circulating current, which requires a ratio (k_i≧0.3) to the through-current output current circuit inductance, the second largest equivalent inductance, the equivalent inductance during concentric summation and coaxial differential operation (L_2), is allocated.

[0102] The remaining equivalent inductances for concentric differential and coaxial summation (L_3) and concentric differential and coaxial differential (L_4) are then allocated to the inductance of the AC output current circuit and the DC circulating current circuit, respectively.

[0103] These equivalent inductances (L_3, L_4) all have a ratio of 0.1 or less to the equivalent inductance (L_1) during concentric and coaxial summation. As long as the equivalent inductance (L_1) is designed while maintaining cost-performance ratio rationality, the equivalent inductances (L_3, L_4) alone are insufficient as current suppression inductance values ​​for both AC output current circuits and DC circulating current circuits.

[0104] Therefore, in order to realize a 2-parallel MMC converter without additional reactors, it is a prerequisite to ensure a sufficiently large inductance value (L_ac) in the AC power supply to suppress the AC output current circuit.

[0105] On the other hand, as long as the equivalent inductance (L_1) is designed while maintaining a reasonable cost-performance ratio in order to suppress the through-current output current, it is difficult to secure an equivalent inductance (L_3) or (L_4) value sufficient to suppress the DC circulating current. Therefore, it is clear that two parallel DC power supply isolation is necessary to realize a two-parallel MMC converter without additional reactors. Even with power supply isolation, it is clear that additional reactors are necessary to stably suppress and control the DC circulating current even during transient phenomena such as the propagation of grid faults.

[0106] Based on the above, by securing the inductance value (L_ac) of the AC power supply and separating the DC power supply of the two parallel MMC converters, the equivalent inductances (L_3, L_4) can be arbitrarily assigned to different functions.

[0107] For the sake of explanation, we will classify the four coils per phase using the names shown in Figure 12, and examine the combinations of coil connection methods that result in concentric and coaxial summation when the through-current output current flows through the four coils. To avoid duplication, the function of the first coil (the inner coil of the upper winding) is fixed to the positive inductive element (22xp1) of the first converter in the two-parallel MMC converter.

[0108] The connections of the remaining three coils, excluding the first coil, are determined by the function assigned to each, resulting in 3 × 2 = 6 possible combinations. For the sake of explanation, these 6 connection combinations will be referred to as the first through sixth conditions.

[0109] Figure 16 shows the function of allocating to the first to fourth coils under six conditions. (xp1) indicates the allocation to the positive side of the first converter, (xn1) to the negative side of the first converter, (xp2) to the positive side of the second converter, and (xn2) to the negative side of the second converter. It also shows whether the through-output current, AC circulating current, AC output current, and DC circulating current act as summation or differential excitation concentrically or as summation or differential excitation coaxially when connected under each condition.

[0110] Figure 16 shows that the AC circulating current acts as the desired concentric summing / coaxial differential (L_2) only under conditions 5 and 6. The other four conditions are unsuitable for realizing a 2-parallel MMC converter without additional reactors.

[0111] In Figure 16, the "Parallel MMC Balanced" column is indicated as "balanced" when either the positive or negative side of the first or second MMC converter is positioned on the inside and the other on the outside, and as "unbalanced" when both the positive and negative sides of the first or second MMC converter are positioned on the inside or outside. In the "balanced" case, the inductance values ​​for AC current of the first and second MMC converters are equal. In the "unbalanced" case, the inductance value for AC current of the converters positioned on the inside is approximately one-tenth of the inductance value for AC current of the converters positioned on the outside. As a result, even if the AC current of each converter can be adjusted, the risk of non-theoretical harmonics due to the unbalanced parallel converters exceeding harmonic regulations increases. For this reason, the conditions under which the "Parallel MMC Balanced" column becomes "unbalanced" are not suitable for practical use.

[0112] Of the two conditions (Condition 5 and Condition 6) in which the AC circulating current acts as the desired concentric summing and coaxial differential excitation, (Condition 5) is "unbalanced" and therefore should be excluded from consideration. Based on the above, it was found that (Condition 6) is the most suitable connection condition and functional allocation for realizing a 2-parallel MMC converter without additional reactors.

[0113] Figure 17 shows the connection and coil arrangement for (condition 6) in Figure 16. As described in the previous paragraph 0107, in Figure 16, to avoid duplication, the function of the first coil (upper / inner coil) is fixed to the positive inductive element (22xp1) of the first converter of the two parallel MMC converters.

[0114] When assigning the function of the first coil (the inner coil of the upper winding) to the negative inductor element (22xn1) of the first transducer, the connection and function are as shown in Figure 18. When assigning the function of the first coil (the inner coil of the upper winding) to the positive inductor element (22xp2) of the second transducer, the connection and function are as shown in Figure 19. When assigning the function of the first coil (the inner coil of the upper winding) to the negative inductor element (22xn2) of the second transducer, the connection and function are as shown in Figure 20. The same functionality as in Figure 17 can be achieved using Figures 18, 19, and 20 described above.

[0115] In both cases, the objectives are to perform concentric summation and coaxial summation excitation operation with through-current output current and concentric summation and coaxial differential excitation operation with AC circulating current. In both cases, two inductive elements of the same polarity in the first and second MMC converters are allocated to two windings and assigned to the same inner coil or the same outer coil, and the positive and negative inductive elements are allocated to the inner and outer coils for each of the first and second MMC converters.

[0116] The above configuration proved to be suitable for solving the problem.

[0117] The two-parallel MMC converter according to the present invention can achieve both miniaturization of the device and assurance of continued operation performance during grid fault propagation.

[0118] Figure 1 shows the configuration of a two-parallel MMC converter according to the present invention. Figure 2 shows the configuration of a two-terminal arm power circuit according to the present invention. Figure 3 shows the configuration of a power converter control device according to the present invention. Figure 4 shows the configuration of a four-coil reactor according to the present invention. Figure 5 shows another configuration of a four-coil reactor according to the present invention. Figure 6 shows the configuration of a four-coil reactor according to Embodiment 2 of the present invention. Figure 7 shows the configuration of a two-parallel MMC converter according to Embodiment 3 of the present invention. Figure 8 shows another configuration of a two-parallel MMC converter according to Embodiment 3 of the present invention. Figure 9 shows the configuration of a conventional one-parallel MMC converter. Figure 10 shows the configuration of a concentric two-coil reactor. Figure 11 shows the current direction and magnetic flux distribution of a concentric two-coil reactor. Figure 12 shows the names and representative dimensions of the windings and coils constituting the four-coil reactor. Figure 13 shows the current direction and magnetic flux distribution of a four-coil reactor. Figure 14 shows the change in inductance with respect to the winding core axial direction and the coil distance between the upper and lower windings. Figure 15 shows the change in inductance value with respect to the radial distance between the centers of the concentric inner and outer coils. Figure 16 shows the functional assignments to the four coils constituting the reactor and the excitation modes in each current mode. Figure 17 shows the coil arrangement and connection when the inner coil of the upper winding is assigned to the positive inductance element of the first MMC converter in the reactor configuration realizing the present invention. Figure 18 shows the coil arrangement and connection when the inner coil of the upper winding is assigned to the negative inductance element of the first MMC converter in the reactor configuration realizing the present invention. Figure 19 shows the coil arrangement and connection when the inner coil of the upper winding is assigned to the positive inductance element of the second MMC converter in the reactor configuration realizing the present invention. Figure 20 shows the coil arrangement and connection when the inner coil of the upper winding is assigned to the negative inductance element of the second MMC converter in the reactor configuration realizing the present invention. Figure 21 shows the change in the sum and difference of the inductance values ​​of the inner and outer coils with respect to the radial distance between the centers of the concentric inner and outer coils. Figure 22 shows the change in the sum and difference of the inductance values ​​of the inner and outer coils with respect to the coil distance between the upper and lower windings in the axial direction of the winding core.

[0119] An embodiment of the two-parallel MMC converter according to the present invention will be described in detail below with reference to the drawings. However, this embodiment does not limit the present invention.

[0120] Figure 1 shows the configuration of the two-parallel MMC converter (1) according to the present invention.

[0121] The circuit of the 2-parallel MMC converter (1) consists of three 3-terminal leg circuits (2u1, 2v1, 2w1) provided for each phase constituting the first MMC converter and three 3-terminal leg circuits (2u2, 2v2, 2w2) provided for each phase constituting the second MMC converter.

[0122] The three-terminal leg circuits (2u1, 2v1, 2w1) constituting the first MMC converter are equipped with positive terminals (UP1, VP1, WP1), negative terminals (UN1, VN1, WN1), and intermediate terminals (UC1, VC1, WC1). The three positive terminals (UP1, VP1, WP1) are connected in a star configuration to the positive first voltage terminal (P1) of the DC power supply (26), and the three negative terminals (UN1, VN1, WN1) are connected in a star configuration to the negative first voltage terminal (N1) of the DC power supply (26).

[0123] The three-terminal leg circuits (2u2, 2v2, 2w2) constituting the second MMC converter are equipped with positive terminals (UP2, VP2, WP2), negative terminals (UN2, VN2, WN2), and intermediate terminals (UC2, VC2, WC2). The three positive terminals (UP2, VP2, WP2) are connected in a star configuration to the positive second voltage terminal (P2) of the DC power supply (26), and the three negative terminals (UN2, VN2, WN2) are connected in a star configuration to the negative second voltage terminal (N2) of the DC power supply (26).

[0124] The currents (I_u1, I_v1, I_w1) passing through the intermediate terminals (UC1, VC1, WC1) of the first MMC converter and the currents (I_u2, I_v2, I_w2) passing through the intermediate terminals (UC2, VC2, WC2) of the second MMC converter are divided into AC circulating currents (I_du, I_dv, I_dw) flowing from the first MMC converter towards the second MMC converter and AC output currents (I_u, I_v, I_w) to the three-phase AC voltage terminals (U, V, W) of the AC power supply (27).

[0125] The AC output current (I_u, I_v, I_w) and the AC circulating current (I_du, I_dv, I_dw) can be expressed by equations (20) and (21).

[0126] The characteristics of the AC power supply (27) are that of a series circuit of a star-connected three-phase voltage source (271u, 271v, 271w) and an inductive element (272u, 272v, 272w) with an inductance value (L_ac).

[0127] The characteristics of the DC power supply (26) are a series circuit of a DC voltage source (261) connected to the first MMC converter and a DC reactor (262) having an inductance value (L_dc), and a series circuit of a DC voltage source (263) connected to the second MMC converter and an inductive element (264) having an inductance value (L_dc).

[0128] The paths of the current (I_dc1) flowing through the DC voltage source (261) and the current (I_dc2) flowing through the DC voltage source (263) are separate. However, for the purpose of examining the control characteristics, a virtual DC output current (I_dcs) and a DC circulating current (I_dcd) are defined by equation (22).

[0129] When a frequency converter is configured using a two-parallel MMC converter (hereinafter referred to as the "background MMC converter") having the same configuration as the two-parallel MMC converter (1), and the positive terminals (UP1, VP1, WP1) and negative terminals (UN1, VN1, WN1) of the first converter constituting the background MMC converter and the positive terminals (UP2, VP2, WP2) and negative terminals (UN2, VN2, WN2) of the second converter are connected one-to-one with the corresponding terminals of the two-parallel MMC converter (1), the implementation configuration of the DC voltage source (261) and DC voltage source (263) is the sum of the zero-phase voltage output of the positive-arm power converter and the zero-phase voltage output of the negative-arm power converter constituting the background MMC converter.

[0130] Furthermore, the DC reactors (262) and (264) are implemented as inductive elements (22) that constitute the background MMC converter, and their inductance values ​​are proportional to the concentric summation and coaxial summation inductances. The inductance value (L_dc) is given by equation (37) described later.

[0131] In the following, each phase of the three-phase AC (u, v, w) will be represented by the subscript (x) or (X). Also, the classification of the first MMC converter or the second MMC converter (1, 2) will be represented by the subscript (j).

[0132] The 3-terminal leg circuit (2xj) has a 2-terminal positive arm power circuit (21xpj), a positive current transformer (23xpj), and a positive 2-terminal inductor (22xpj) connected in series between the positive terminal (XPj) and the intermediate terminal (XCj), and a negative 2-terminal inductor (22xnj), a negative current transformer (23xnj), and a 2-terminal negative arm power circuit (21xnj) connected in series between the intermediate terminal (XCj) and the negative terminal (XNj).

[0133] The positive current transformers (23upj, 23vpj, 23wpj) measure and output the current (I_upj, I_vpj, I_wpj) of the positive arm power circuits (21upj, 21vpj, 21wpj) to the power converter control device (4) described later, and the negative current transformers (23unj, 23vnj, 23wnj) measure and output the current (I_unj, I_vnj, I_wnj) of the negative arm power circuits (21unj, 21vnj, 21wnj) to the power converter control device (4) described later.

[0134] Figure 2 shows the configuration of a two-terminal arm power circuit (21). The positive arm power circuit (21xpj) and the negative arm power circuit (21xnj) are identical in configuration except for their connections. The positive arm power circuit (21xpj) has terminal (b) connected to the positive terminal (XPj), and the negative arm power circuit (21xnj) has terminal (a) connected to the negative terminal (XNj).

[0135] The two-terminal arm power circuit (21) has a configuration in which K unit converters (3) (where K is a natural number of 2 or more) are connected in series between the positive terminal (b) and the negative terminal (a). Note that in Figure 2, the circuit configuration is omitted except for the unit converter (3) labeled "No. i".

[0136] The unit converter (3) has two terminals, a positive terminal (y) and a negative terminal (x), and connects the self-extinguishing elements (31H) and (31L) and the antiparallel diodes (32H) and (32L) that constitute a bidirectional chopper circuit to the capacitor (33).

[0137] The gate drive units (GDU) (34H) and (34L) receive gate commands (GH, GL) from the power converter control device (4) via terminals (c) and (g), output firing and extinguishing commands to the self-extinguishing elements (31H) and (31L), and perform PWM control to adjust the voltage between the (xy) terminals to the target voltage.

[0138] The voltage detector (35) outputs the voltage measurement signal (Vc) of the capacitor (33) to the power converter control device (4) via terminal (d) through the signal converter (CONV) (36).

[0139] Figure 3 shows the configuration of the power converter control device (4).

[0140] The power converter control device (4) receives K capacitor voltage signals from the terminals (d) of the K unit converters of the positive arm power circuit (21up1) that constitutes the first MMC converter of the two parallel MMC converters (1), and receives K capacitor voltage signals from the terminals (d) of the K unit converters of the negative arm power circuit (21un1).

[0141] Similarly, 4 × K capacitor voltage signals are input from the positive and negative arm power circuits (21vp1, 21vn1, 21wp1, 21wn1).

[0142] Similarly, the power converter control device (4) receives 6 × K capacitor voltage signals from the positive and negative arm power circuits (21up2, 21un2, 21vp2, 21vn2, 21wp2, 21wn2) of the second MMC converter.

[0143] The power converter control device (4) receives positive arm currents (I_up1, I_vp1, I_wp1) from the positive current transformers (23up1, 23vp1, 23wp1) that constitute the first MMC converter of the two parallel MMC converters (1), and negative arm currents (I_un1, I_vn1, I_wn1) from the negative current transformers (23un1, 23vn1, 23wn1).

[0144] Similarly, the power converter control device (4) receives positive arm currents (I_up2, I_vp2, I_wp2) from the positive current transformers (23up2, 23vp2, 23wp2) that constitute the second MMC converter of the two parallel MMC converters (1), and negative arm currents (I_un2, I_vn2, I_wn2) from the negative current transformers (23un2, 23vn2, 23wn2).

[0145] The power converter control device (4) receives the voltage (V_dc1) from the DC voltage source (261) connected to the first MMC converter and the voltage (V_dc2) from the DC voltage source (263) connected to the second MMC converter.

[0146] Figure 3 shows the case where a frequency converter is configured by connecting to a rear MMC converter. Voltages (V_dc1) and (V_dc2) are input from the power converter control device (4_1) of the rear MMC converter, and the zero-phase through-current voltages (V_dcp1 + V_dcn1) and (V_dcp2 + V_dcn2), calculated from the voltages of the positive arm power circuit (21xpj) and the negative arm power circuit (21xnj), are output to the power converter control device (4_1) of the rear MMC converter.

[0147] The values ​​of the zero-phase through-voltages (V_dcp1 + V_dcn1) and (V_dcp2 + V_dcn2) correspond to equation (34) described later.

[0148] The power converter control device (4) includes a coordinate converter that takes 12 positive and negative arm currents as input and calculates and outputs 10 currents (I_α, I_β, I_dα, I_dβ, I_cα, I_cβ, I_cdα, I_cdβ, I_dcs, I_dcd).

[0149] The power converter control device (4) includes a current regulator that calculates voltage commands so that the 10 coordinate-transformed current values ​​match the 10 current command values ​​calculated from the operating conditions.

[0150] The power converter control device (4) receives a voltage command from the current regulator and includes a coordinate inverse converter that calculates and outputs arm voltage commands to the 12 positive and negative arm power circuits (21up1, 21un1, 21vp1, 21vn1, 21wp1, 21wn1, 21up2, 21un2, 21vp2, 21vn2, 21wp2, 21wn2) that constitute the two parallel MMC converters (1).

[0151] The power converter control device (4) includes an interstage voltage regulator that outputs voltage correction commands to self-extinguishing elements (31H) and (31L) for each of the K unit converters (3) that make up each arm power circuit (21), to each of the K unit converters (3).

[0152] The power converter control device (4) is equipped with a PWM modulator for each unit converter (3). For each unit converter (3), it inputs a voltage correction command from the interstage voltage regulator to the arm voltage command from the coordinate inverse converter, biases it, and inputs it to the PWM modulator. It also outputs firing and arc extinguishing commands to the self-extinguishing elements (31H) and (31L) to the gate drive units (GDU) (34H) and (34L) via the terminal (g) of the unit converter (3).

[0153] The configuration of the coordinate converter and inverse coordinate converter of the power converter control device (4) will be described below.

[0154] To avoid duplication, the current variable (I), voltage variable (V), and flux linkage variable (λ) will be represented by the variable (f) below.

[0155] In the following, each phase of the three-phase AC (u, v, w) will be represented by the subscript (x) or (X). Also, the classification of the first MMC converter or the second MMC converter (1, 2) will be represented by the subscript (j).

[0156] In the following, the variables relating to the AC current, through-current, positive arm power circuit, and negative arm power circuit inside the first or second MMC converter are the same as those in equations (1) through (9) above, except for the subscript (j) at the end, and explanations are omitted to avoid duplication.

[0157] In the coordinate transformation and inverse transformation of the output current and circulating current by the first or second MMC converter constituting the two-parallel MMC converter (1), the positive and negative through-current variables are given by equation (23).

[0158] The AC output variable and the AC circulation variable are given by equations (24) and (25).

[0159] The through-flow output variable and the through-flow circulation variable are given by equations (26) to (28).

[0160] The characteristics of the four-coil reactor in Figure 17, which correspond to the connection and coil arrangement in (Condition 6) of Figure 16, can be expressed by equation (29).

[0161] Based on the above coordinate transformation, the 10 current variables and 10 circuit equations that define the configuration of the current regulator of the power converter control device (4) can be separated into α phase, β phase and zero phase. The 10 circuit equations can be further divided into five sets: two sets of α phase current variables (I_α, I_cα) and (I_dα, I_cdα), two sets of β phase current variables (I_β, I_cβ) and (I_dβ, I_cdβ), and two zero phase current variables (I_dcs, I_dcd).

[0162] The two sets of binary circuit equations for the α phase are given by equations (30) and (31).

[0163] The two sets of binary circuit equations for the β phase are given by equations (32) and (33).

[0164] The zero-phase binary circuit equation is given by equation (34).

[0165] Here, the inductance values ​​appearing in the diagonal elements of the left-hand inductance matrix of each circuit equation shown in equations (30), (31), (32), (33), and (34) above are given by equations (35), (36), (37), and (38).

[0166] From the first line of equations (30) and (32) above, it can be seen that the inductance (L_acs) of the AC output current (I_α, I_β) circuit is proportional to the equivalent inductance (L_3) during concentric differential and coaxial summative excitation, as can be seen by comparing it with equation (18).

[0167] From the first line of equations (31) and (33) above, it can be seen that the inductance (L_acd) of an AC circulating current (I_dα, I_dβ) circuit is proportional to the equivalent inductance (L_2) during concentric summation and coaxial differential operation, as can be seen by comparing it with equation (17).

[0168] From the first line of equation (34) above, it can be seen that the inductance (L_dcs) of the DC output current (I_dcs) circuit is proportional to the equivalent inductance (L_1) during concentric summation and coaxial summation, as can be seen by comparing it with equation (16).

[0169] From the second line of equation (34) above, it can be seen that the inductance (L_dcd) of the DC circulating current (I_dcd) circuit is proportional to the equivalent inductance (L_4) in the concentric differential and coaxial differential circuits, by comparison with equation (19).

[0170] The circuit inductance values ​​in equations (30), (31), (32), (33), and (34) above are all equivalent inductance values ​​intended in the sixth condition (Figure 16).

[0171] The above is the circuit equation when the connections and functions shown in Figure 17 are assigned in accordance with the sixth condition in Figure 16. The circuit inductance values ​​when the connections and functions shown in Figures 18, 19, and 20 are assigned are the same as in Figure 17, and the same function can be ensured.

[0172] The inductance values ​​appearing in the off-diagonal elements (other than the diagonal elements) of the inductance matrix on the left-hand side of each circuit equation shown in equations (30), (31), (32), and (33) above represent interference terms with respect to the diagonal components of the current being regulated, as viewed from the current regulator. It is desirable that these inductance values ​​be as small as possible for current regulation purposes.

[0173] The inductance values ​​(L_sp) and (L_sn) of the off-diagonal elements are defined by equations (39) and (40).

[0174] Figure 21 shows the changes in the dimensionless sum of the inner and outer inductance values ​​(L_sp) and the difference between the inner and outer inductance values ​​(L_sn) with respect to the ratio (ε_x) of the distance between the inner and outer coil centers (d_x) to the distance (D_x) between the iron core axis and the inner and outer coil centers (D_x) in Figure 12.

[0175] Figure 22 shows the changes in the dimensionless sum of the internal and external inductance values ​​(L_sp) and the difference between the internal and external inductance values ​​(L_sn) with respect to the ratio (ε_y) of the coil distance between the upper and lower windings to the coil width (H_y) in the axial direction of the iron core, as shown in Figure 12.

[0176] In both Figures 21 and 22, the dimensionless reference value is the equivalent inductance value (L_dcs) corresponding to concentric summative and coaxial summative excitation when the upper and lower winding coils are in close contact (h_y = 0).

[0177] In Figures 21 and 22, both the sum of the internal and external inductance values ​​(L_sp) and the difference between the internal and external inductance values ​​(L_sn) have negative signs. This is the case when the coil connection and arrangement are selected as in Figures 17 and 19. When the coil connection and arrangement are selected as in Figures 18 and 20, the sign becomes positive, but the change in its absolute value is independent of the coil connection and arrangement in Figures 17 through 20.

[0178] The following explains the absolute value changes of the sum of the internal and external inductance values ​​(L_sp) and the difference between the internal and external inductance values ​​(L_sn).

[0179] In Figure 21, the absolute values ​​of the sum of the internal and external inductance values ​​(L_sp) and the difference between the internal and external inductance values ​​(L_sn) increase with increasing ratio (ε_x) of the distance between the centers of the internal and external coils (d_x). In particular, the sum of the internal and external inductance values ​​(L_sp) is twice the difference between the internal and external inductance values ​​(L_sn). In equations (30), (31), (32), and (33), which are the circuit equations for the α-phase and β-phase, the ratio of the off-diagonal elements to the diagonal elements of the inductance matrix on the left side indicates the magnitude of the interference. The first ratio is the ratio (L_sn / L_acd) of the first row in equations (31) and (33). This ratio indicates the interference of through-circulating currents (I_cdα, I_cdβ) with respect to the control of AC circulating currents (I_dα, I_dβ).

[0180] The equivalent inductance (L_acd) during concentric summation and coaxial differential operation, which is necessary for controlling AC circulating currents (I_dα, I_dβ), is proportional to (L_2), and is therefore hardly affected by the ratio (ε_x) of the distance between the centers of the inner and outer coils (d_x), as shown in Figure 15.

[0181] From the above, it can be seen that, for the reactor application of the present invention for a two-parallel MMC converter, it is desirable to set the distance between the centers of the inner and outer coils as small as possible, and the ratio (ε_x) of the distance between the centers of the inner and outer coils (d_x), as far as the mechanical structural constraints and the constraints of the minimum insulation distance between the inner and outer coils allow.

[0182] The equivalent inductance (L_2) required for concentric summation and coaxial differential winding control of AC circulating currents (I_dα, I_dβ) must be such that, according to paragraph 0101, the ratio to the equivalent inductance during concentric summation and coaxial summation (k_i≧0.3) must be ensured. From Figure 14, the ratio (ε_y) of the coil distance between the upper and lower windings (h_y) to the coil width (H_y) in the axial direction of the iron core should be approximately (ε_y≧0.5).

[0183] As shown in Figure 22, the sum of the internal and external inductance values ​​(L_sp) gradually increases in absolute value as the ratio (ε_y) of the coil distance between the upper and lower windings (h_y) to the coil width (H_y) in the axial direction of the iron core increases. However, the rate of increase becomes smaller than that of the equivalent inductance (L_acd), and the ratio (L_sn / L_acd) in the first row reaches a maximum value.

[0184] Increasing the coil distance ratio (ε_y) leads to an increase in the axial winding core length, resulting in a larger reactor, thus limiting the practical range of the coil distance ratio (ε_y). In the practical range of coil distance ratio (1≧ε_y≧0.5), (L_sn / L_acd≦0.07) is observed. This ratio is sufficiently small, and it can be seen that interference with AC circulating current (I_dα, I_dβ) can be suppressed to a range that does not cause problems in practical operation by controlling the through-circulating current (I_cdα, I_cdβ) with current control, including during transient phenomena such as the propagation of system faults.

[0185] As mentioned above, in equations (30), (31), (32), and (33), which are the circuit equations for the α-phase and β-phase, the ratio of the off-diagonal elements to the diagonal elements of the inductance matrix on the left side indicates the magnitude of the interference. The second ratio is the ratio of the second row of equations (31) and (33) (L_sn / L_dcd / 4).

[0186] The ratio in the second row shows the interference of the alternating current (I_dα, I_dβ) with the control of the through-circulating current (I_cdα, I_cdβ).

[0187] The equivalent inductance (L_dcd) required for controlling the through-circulating current (I_cdα, I_cdβ) during concentric differential and coaxial differential operation is proportional to (L_2) and increases with the ratio (ε_x) of the distance between the centers of the inner and outer coils (d_x), as shown in Figure 15.

[0188] As shown in Figure 21, the difference in inductance values ​​between the inner and outer coils (L_sn) also increases with the ratio (ε_x) of the distance between the centers of the inner and outer coils (d_x).

[0189] As a result, the ratio of the second row (L_sn / L_dcd / 4) has a minimum value of approximately 0.12 within the practical constraint range (ε_x = 0.3 to 0.4). From the above, it can be seen that by setting the ratio of the second row to a sufficiently small value and suppressing the AC circulating current (I_dα, I_dβ) by current control, including during transient phenomena such as when a system fault propagates, interference with the through-circulating current (I_cdα, I_cdβ) can be suppressed to a range that does not cause problems in practical operation.

[0190] Figure 4 shows the arrangement and connection of the four coils constituting the reactor according to the present invention to the winding core. Components with the same reference numerals as those described in Figure 1 represent the same components. Explanations are omitted to avoid duplication. In addition, each phase (u, v, w) of the three-phase AC is represented by (x) or (X).

[0191] The three-terminal leg circuit (2x1) of the first MMC converter and the three-terminal leg circuit (2x2) of the second MMC converter are magnetically coupled via a common wound core (24x) in a reactor consisting of the upper winding (425x1), which is the first winding, and the lower winding (425x2), which is the second winding.

[0192] The wound core (24x) forms a closed magnetic circuit through the upper yoke (28) and the lower yoke (29).

[0193] The current (I_xp1) circuit from the positive arm power circuit (21xp1) of the first MMC converter is connected to the upper yoke (28) side terminal of the inner coil (422xp1) of the upper winding (425x1), and the opposite terminal is connected to the outer coil (422xn1) terminal of the opposing lower winding (425x2). It is also branched to the intermediate terminal (XC1). The lower yoke (29) side terminal of the outer coil (422xn1) of the lower winding (425x2) is connected to the current (I_xn1) circuit from the negative arm power circuit (21xn1) of the first MMC converter.

[0194] The inner coil (422xp1) of the upper winding (425x1) and the outer coil (422xn1) of the lower winding (425x2) are wound in the same direction. In the example in Figure 4, both are wound clockwise (right-hand screw), and the through-current (I_cx1) has a coaxial summation excitation effect.

[0195] The current (I_xp2) circuit from the positive arm power circuit (21xp2) of the second MMC converter is connected to the terminals on the opposite side of the upper and lower windings of the inner coil (422xp2) of the lower winding (425x2), and the terminal on the opposite side of the lower yoke (29) is connected to the terminal on the upper yoke (28) side of the outer coil (422xn2) of the upper winding (425x1) via a connecting wire (430) which is arranged coaxially and parallel to the lower winding (425x2) and the upper winding (425x1). It is also branched to the intermediate terminal (XC2). The terminals on the opposite side of the upper and lower windings of the outer coil (422xn2) of the upper winding (425x1) are connected to the current (I_xn2) circuit from the negative arm power circuit (21xn2) of the second MMC converter.

[0196] The inner coil (422xp2) of the lower winding (425x2) and the outer coil (422xn2) of the upper winding (425x1) are wound in the same direction as the inner coil (422xp1) of the upper winding (425x1) and the outer coil (422xn1) of the lower winding (425x2). In the example in Figure 4, both are wound clockwise (right-hand screw) and the through-current (I_cx2) has a coaxial summation excitation effect.

[0197] With the above connection method, coil arrangement, and coil winding method, the through-current output current (I_cx) exhibits concentric summative and coaxial summative excitation effects, and the AC circulating current (I_dx) exhibits concentric summative and coaxial differential excitation effects, thereby realizing the intended function as a reactor for a 2-parallel MMC converter.

[0198] According to the embodiment shown in Figure 4, there is an advantage in being able to use the same inner and outer coils for the upper winding (425x1) and the lower winding (425x2). In addition, since the four coils are wound in the same direction, there is an advantage in simplifying the setup of the winding jig.

[0199] Figure 5 shows an alternative arrangement and connection of the four coils constituting the reactor according to the present invention to the winding core. Components with the same reference numerals as those described in Figures 1 and 4 represent the same components. To avoid duplication, their descriptions are omitted.

[0200] The current (I_xp1) circuit from the positive arm power circuit (21xp1) of the first MMC converter is connected to the upper yoke (28) side terminal of the inner coil (522xp1) of the upper winding (525x1), which is the first winding. The terminals on the opposite side of the upper and lower windings are connected to the lower yoke (29) side terminal of the outer coil (522xn1) of the lower winding (525x2) via a connecting wire (530) that is arranged coaxially and parallel to the lower winding (525x2), which is the second winding. It is also branched and connected to the intermediate terminal (XC1).

[0201] The terminals on the opposite sides of the upper and lower windings of the outer coil (522xn1) are connected to the current (I_xn1) circuit from the negative arm power circuit (21xn1) of the first MMC converter.

[0202] The inner coil (522xp1) of the upper winding (525x1) and the outer coil (522xn1) of the lower winding (525x2) are wound in opposite directions, causing the through-current (I_cx1) to have a coaxial summative excitation effect. In the example in Figure 5, the former is wound clockwise (right-hand screw) and the latter is wound counterclockwise (left-hand screw), and the through-current (I_cx1) has a coaxial summative excitation effect.

[0203] The current (I_xp2) circuit from the positive arm power circuit (21xp2) of the second MMC converter is connected to the terminals on the opposite side of the upper and lower windings of the inner coil (522xp2) of the lower winding (525x2), and the terminal on the opposite side of the lower yoke (29) is branched to the intermediate terminal (XC2). It is also connected to the terminals on the opposite side of the upper and lower windings of the outer coil (522xn2) of the upper winding (525x1) via a connecting wire (531) arranged coaxially and parallel to the lower winding (525x2). The terminal on the upper yoke (28) side of the outer coil (522xn2) of the upper winding (525x1) is connected to the current (I_xn2) circuit from the negative arm power circuit (21xn2) of the second MMC converter.

[0204] The inner coil (522xp2) of the lower winding (525x2) and the outer coil (522xn2) of the upper winding (525x1) are wound in opposite directions. This causes the through-current (I_cx2) to have a coaxial summation excitation effect. In the example in Figure 5, the former is wound clockwise (right-hand screw) and the latter is wound counterclockwise (left-hand screw).

[0205] In the example shown in Figure 5, the inner coils (522xp1, 522xp2) of the upper winding (525x1) and lower winding (525x2) are both wound clockwise (right-hand thread), while the outer coils are wound counterclockwise (left-hand thread).

[0206] With the above connection method, coil arrangement, and coil winding method, the through-current output current (I_cx) exhibits concentric summative and coaxial summative excitation effects, and the AC circulating current (I_dx) exhibits concentric summative and coaxial differential excitation effects, thereby realizing the intended function as a reactor for a 2-parallel MMC converter.

[0207] According to the embodiment shown in Figure 5, there is an effect of making the inner and outer coils of the upper winding (525x1) and lower winding (525x2) common. In addition, the connecting wires (530) and (531) are distributed between the first MMC converter and the second MMC converter, and since both can be made to the same length, there is an effect of suppressing inductance imbalance between parallel converters.

[0208] Figure 6 shows the arrangement and connection of the four coils constituting the reactor according to Embodiment 2 of the present invention to the winding core. Components with the same reference numerals as those described in Figures 1 and 4 represent the same components. To avoid duplication, their descriptions are omitted.

[0209] The current (I_xp1) circuit from the positive arm power circuit (21xp1) of the first MMC converter is connected to the upper yoke (28) side terminal of the inner coil (622xp1) of the upper winding (625x1), which is the first winding, and is connected to the outer coil (622xn1) terminal of the lower winding (625x2), which is the second winding, between the terminals on opposite sides of the upper and lower windings. It is also branched and connected to the intermediate terminal (XC1).

[0210] The lower yoke (29) side terminal of the outer coil (622xn1) is connected to the current (I_xn1) circuit from the negative arm power circuit (21xn1) of the first MMC converter.

[0211] The inner coil (622xp1) of the upper winding (625x1) and the outer coil (622xn1) of the lower winding (625x2) are wound in the same direction, thereby causing the through-current (I_cx1) to have a coaxial summation excitation effect. Figure 6 shows an example of clockwise (right-hand screw) winding.

[0212] The current (I_xp2) circuit from the positive arm power circuit (21xp2) of the second MMC converter is connected to the lower yoke (29) side terminal of the inner coil (622xp2) of the lower winding (625x2), and is connected to the outer coil (622xn2) terminal of the upper winding (625x1) between the terminals on opposite sides of the upper and lower windings. It is also branched and connected to the intermediate terminal (XC2).

[0213] The upper yoke (28) side terminal of the outer coil (622xn2) of the upper winding (625x1) is connected to the current (I_xn2) circuit from the negative arm power circuit (21xn2) of the second MMC converter.

[0214] The inner coil (622xp2) of the lower winding (625x2) and the outer coil (622xn2) of the upper winding (625x1) are wound in the same direction. As a result, the through-current (I_cx2) has a coaxial summation excitation effect. In the example in Figure 6, the winding is counterclockwise (left-hand screw).

[0215] In the example shown in Figure 6, the coils on the first MMC converter side (622xp1, 622xn1) and the coils on the second MMC converter side (622xp2, 622xn2) are wound in opposite directions.

[0216] With the above connection method, coil arrangement, and coil winding method, the through-current output current (I_cx) exhibits concentric summative and coaxial summative excitation effects, and the AC circulating current (I_dx) exhibits concentric summative and coaxial differential excitation effects, thereby realizing the intended function as a reactor for a 2-parallel MMC converter.

[0217] According to the embodiment shown in Figure 6, the connecting wire between the upper winding (625x1) and the lower winding (625x2) is the shortest distance connection between the terminals on opposite sides of the upper and lower windings, thus suppressing the need to increase the size of the device. Furthermore, since there is no need to bring the connecting wire outside the iron core, leakage of magnetic flux to the outside is suppressed, and the risk of inductive overheating of surrounding equipment is reduced.

[0218] Figure 7 shows the configuration of the two-parallel MMC converter (7) according to Embodiment 3 of the present invention.

[0219] Parts with the same numbers as those explained in Figure 1 above represent the same parts. To avoid duplication, the explanation is omitted.

[0220] The first DC reactor (71) is a two-terminal reactor installed between the positive terminals (UP1, VP1, WP1) of the first MMC converter and the positive first voltage terminal (P1) of the DC power supply (26).

[0221] The second DC reactor (72) is a two-terminal reactor installed between the positive terminals (UP2, VP2, WP2) of the second MMC converter and the positive second voltage terminal (P2) of the DC power supply (26).

[0222] The first DC reactor (71) and the second DC reactor (72) are magnetically coupled by a common iron core (73), and the coil arrangement and winding direction are selected such that differential excitation occurs when the current (I_dc1) flowing through the DC voltage source (261) and the current (I_dc2) flowing through the DC voltage source (263) have the same inflow and outflow directions to the positive first voltage terminal (P1) and positive second voltage terminal (P2), respectively, and summation excitation occurs when they are in opposite directions.

[0223] If the self-inductance of the first DC reactor (71) is (L_dx1), the self-inductance of the second DC reactor (72) is (L_dx2), and the mutual inductance is (M_dx>0), then the relationship between the flux linkage (λ_dx1, λ_dx2) and the current (I_dc1, I_dc2) is given by equation (41).

[0224] Using the previous equation (28), equation (41) can be converted into DC output and DC circulation components to obtain equation (42).

[0225] Based on the above, when the first and second DC reactors (71, 72) shown in Figure 7 are installed, the zero-phase binary circuit equation shown in equation (34) above becomes equation (43).

[0226] On the other hand, even if the first and second DC reactors (71, 72) shown in Figure 7 are installed, the α and β phases are not affected, and the circuit equations (30), (31), (32), and (33) remain unchanged.

[0227] The first and second DC reactors (71, 72) have the effect of adding the summation excitation inductance {(1 / 2)・(L_dx1 + 2・M_dx + L_dx2)} to equation (34) for the DC circulating current, and have the effect of stably suppressing DC current imbalance between parallel converters by controlling the DC circulating current even during transient phenomena such as system ground faults.

[0228] Here, the first DC reactor (71) and the second DC reactor (72) are constructed as two concentric coils wound around a common iron core (73). The ratio of the inductance of the DC output current (I_dcs) to the inductance of the DC circulating current (I_dcd) is equal to (k_i) in the previous paragraph 0069, so (k_i ≤ 0.1). Therefore, the effect of the overcurrent value of the DC output current on the magnetic saturation of the common iron core (73) can be said to be negligible at the (k_i) level.

[0229] If the peak value of the DC output current (I_dcs) during a grid fault propagation is suppressed to twice the rated value or less, and the DC circulating current (I_dcd) is suppressed to 0.1 times the DC output current or less, the peak value of the DC circulating current (I_dcd) will be 0.2 times the rated value.

[0230] As a result, the volume of the common iron core (73) required to suppress magnetic saturation is proportional to the stored magnetic energy, which allows for a significantly smaller and lighter design compared to a DC reactor designed for rated current.

[0231] Figure 8 shows the configuration of another two-parallel MMC converter (8) according to Embodiment 3 of the present invention.

[0232] Parts with the same reference numerals as those explained in Figures 1 and 7 represent the same parts. Explanations are omitted to avoid duplication.

[0233] The positive DC reactor (881p) consists of two concentric coils winding around a wound core (882p). The positive inner coil (883p) is connected between the positive terminals (UP1, VP1, WP1) of the first MMC converter and the positive first voltage terminal (P1) of the DC power supply (26), and the positive outer coil (884p) is connected between the positive terminals (UP2, VP2, WP2) of the second MMC converter and the positive second voltage terminal (P2) of the DC power supply (26). When the current (I_dc1) of the positive inner coil (883p) and the current (I_dc2) of the positive outer coil (884p) are of the same polarity, these concentric coils perform differential excitation, creating a magnetic flux in opposite directions on the wound core (882p).

[0234] The negative DC reactor (881n) consists of two concentric coils winding around a wound core (882n). The negative outer coil (884n) is connected between the negative terminals (UN1, VN1, WN1) of the first MMC converter and the negative first voltage terminal (N1) of the DC power supply (26), and the negative inner coil (883n) is connected between the negative terminals (UN2, VN2, WN2) of the second MMC converter and the negative second voltage terminal (N2) of the DC power supply (26). When the current (I_dc1) of the negative outer coil (884n) and the current (I_dc2) of the negative inner coil (883n) are of the same polarity, these concentric coils perform differential excitation, creating a magnetic flux in opposite directions on the wound core (882n).

[0235] In Figure 8, the inner coil is wound clockwise (right-hand screw direction) and the outer coil is wound counterclockwise (left-hand screw direction). However, as long as differential excitation is performed with currents of the same polarity, the coil winding direction can be selected to optimize the connection wiring according to the equipment layout.

[0236] In Figure 8, the coil connected to the positive terminal of the first MMC converter is selected as the positive inner coil (883p), and the coil connected to the negative terminal is selected as the negative outer coil (884n). The coil connected to the positive terminal of the second MMC converter is selected as the positive outer coil (884p), and the coil connected to the negative terminal is selected as the negative inner coil (883n). This makes it possible to match the inductance of the DC current (I_dc1) circuit of the first MMC converter and the DC current (I_dc2) circuit of the second MMC converter.

[0237] Here, the characteristics of the positive DC reactor (881p) are given by equation (44), and the characteristics of the negative DC reactor (881n) are given by equation (45). The inductance values ​​of the inner coil (L_dxi) and the outer coil (L_dxo) are given by equation (46).

[0238] The zero-phase binary circuit equation for the two-parallel MMC converter (8) using the positive DC reactor (881p) and the negative DC reactor (881n) with the above characteristics is given by equations (43) to (47) of the two-parallel MMC converter (7) in Figure 7.

[0239] The off-diagonal components of the inductance matrix on the left side of equation (47) are zero, indicating that there is no interference between the DC output current (I_dcs) and the DC circulating current (I_dcd).

[0240] Based on the above, by distributing two DC reactors (881p) and (881n), each consisting of concentric coils with the same characteristics, to the positive and negative sides, and by assigning the positive and negative coils of each parallel converter to the inner and outer coils, sufficient inductance values ​​can be secured to stably suppress DC circulating current even during a grid fault, and interference between DC output current and DC circulating current control can be suppressed.

[0241] 1, 7, 8 2-parallel modular multilevel power converters 2u1, 2v1, 2w1, 2u2, 2v2, 2w2, 802u, 802v, 802w 3 terminal leg circuits 3 unit converters 4 power converter control devices 21 arm power circuits 21up1, 21vp1, 21wp1, 21up2, 21vp2, 21wp2, 821up, 821vp, 821wp positive arm power circuits 21un1, 21vn1, 21wn1, 21un2, 21vn2, 21wn2, 821un, 821vn, 821wn negative arm power circuits 22up1, 22vp1, 22wp1, 22up2, 22vp2, 22wp2, 822up, 822vp, 822wp: Positive two-terminal inductors 22un1, 22vn1, 22wn1, 22un2, 22vn2, 22wn2, 822un, 822vn, 822wn: Negative two-terminal inductors 23up1, 23vp1, 23wp1, 23up2, 23vp2, 23wp2, 823up, 823vp, 823wp: Positive current transformers 23un1, 23vn1, 23wn1, 23un2, 23vn2, 23wn2, 823un, 823vn, 823wn: Negative current transformers 24u, 24v, 24w, 824u, 824v, 824w, 882p, 882n Wound core 26, 804 DC power supply 27, 803 AC power supply 28 Upper yoke 29 Lower yoke 31H, 31L Self-extinguishing power semiconductor element 32H, 32L Antiparallel diode 33 Energy storage unit 34H, 34L Gate drive unit 35 Voltage detector 36 Signal converter 71, 72, 262, 264, 842, 881p, 881n DC reactor 73 Common core 883p Positive inner coil 884p Positive outer coil 883n Negative inner coil 884n Negative outer coil 261, 263, 841 DC voltage source 272u, 272v, 272w, 832u, 832v, 832w Induction elements 271u, 271v, 271w, 831u, 831v, 831w Three-phase voltage sources422 up1, 422vp1, 422wp1, 422 up2, 422vp2, 422wp2, 522 up1, 522vp1, 522wp1, 522 up2, 522vp2, 522wp2, 622 up1, 622vp1, 622wp1, 622 up2, 622vp2, 622wp2 inner coil Outer coils: 422un1, 422vn1, 422wn1, 422un2, 422vn2, 422wn2, 522un1, 522vn1, 522wn1, 522un2, 522vn2, 522wn2, 622un1, 622vn1, 622wn1, 622un2, 622vn2, 622wn2 Outer coils: 425u1, 425v1, 425w1, 525u1, 525v1, 525w1, 625u1, 625v1, 625w1 Upper windings: 425u2, 425v2, 425w2, 525u2, 525v2, 525w2, 625u2, 625v2, 625w2 Lower windings: 430, 530, 531; Crossover wire: 801; Modular multilevel power converter (MMC converter): 825u, 825v, 825w; Concentric windings: 826u, 826v, 826w; Concentric two-winding reactor

Claims

1. A two-parallel modular multilevel power converter (1) connected between three-phase AC voltage terminals (U terminal, V terminal, W terminal) and the positive voltage terminal (P1 terminal) and negative voltage terminal (N1 terminal) of a first DC power supply and the positive voltage terminal (P2 terminal) and negative voltage terminal (N2 terminal) of a second DC power supply, wherein the two-parallel modular multilevel power converter (1) comprises a power converter control device (4) and two sets of three three-terminal leg circuits (2u1, 2v1, 2w1), (2u2, 2v2, 2w2), The aforementioned three-terminal leg circuit (2xj), in which x is a representative representation of three phases (u, v, w) and j is a representative representation of two parallel connections (1, 2), is a three-terminal leg circuit (2xj) in which a two-terminal positive arm power circuit (21xpj), a positive current transformer (23xpj), a positive two-terminal inductor element (22xpj), and an intermediate terminal (xCj) are connected in series starting from the positive terminal (xPj), and a negative two-terminal inductor element (22xnj), a negative current transformer (23xnj), a two-terminal negative arm power circuit (21xnj), and a negative terminal (xNj) are connected in series starting from this intermediate terminal (xCj), The positive current transformers (23upj, 23vpj, 23wpj) measure and output the current (I_upj, I_vpj, I_wpj) of the positive arm power circuits (21upj, 21vpj, 21wpj) to the power converter control device (4), and the negative current transformers (23unj, 23vnj, 23wnj) measure and output the current (I_unj, I_vnj, I_wnj) of the negative arm power circuits (21unj, 21vnj, 21wnj) to the power converter control device (4). The positive arm power circuit (21xpj) and the negative arm power circuit (21xnj) are series circuits consisting of K (K is a natural number of 2 or more) unit converters (3) each having one energy storer (33) and at least two self-extinguishing power semiconductor elements (31H, 31L), wherein the unit converters (3) receive PWM-modulated gate commands (GH, GL) to the self-extinguishing power semiconductor elements (31H, 31L) output from the power converter control device (4), output a required terminal voltage with zero as the lower limit and the voltage of the energy storer (33) as the upper limit, and output a voltage measurement signal (Vc) of the energy storer (33) to the power converter control device (4),The two-terminal inductive elements (22upj, 22unj, 22vpj, 22vnj, 22wpj, 22wnj) are equipped with a function to suppress circulating current flowing between the three-terminal leg circuits (2uj, 2vj, 2wj) when the voltage ratio between the three-terminal leg circuits (2uj, 2vj, 2wj) is unbalanced, and the power converter control device (4) is a power converter control device (4) that includes a function to suppress circulating current flowing between the three-terminal leg circuits (2uj, 2vj, 2wj), The positive terminals (UP1, VP1, WP1) of the first set of three-terminal leg circuits (2u1, 2v1, 2w1) are connected in a star configuration to the positive voltage terminal (P1 terminal) of the first DC power supply, the negative terminals (UN1, VN1, WN1) are connected in a star configuration to the negative voltage terminal (N1 terminal) of the first DC power supply, and the intermediate terminals (UC1, VC1, WC1) of the first set of three-terminal leg circuits (2u1, 2v1, 2w1) are connected to the three-phase AC voltage terminals (U terminal, V terminal, W terminal). In a two-parallel modular multilevel power converter (1), the positive terminals (UP2, VP2, WP2) of the second set of three-terminal leg circuits (2u2, 2v2, 2w2) are connected in a star configuration to the positive voltage terminal (P2 terminal) of the second DC power supply, the negative terminals (UN2, VN2, WN2) are connected in a star configuration to the negative voltage terminal (N2 terminal) of the second DC power supply, and the intermediate terminals (UC2, VC2, WC2) of the second set of three-terminal leg circuits (2u2, 2v2, 2w2) are connected to the three-phase AC voltage terminals (U terminal, V terminal, W terminal), and the four coils of the two-terminal inductive elements (22xp1, 22xn1, 22xp2, 22xn2) are magnetically coupled by winding them around a wound iron core (24x) provided for each of the three phases (u, v, w). The two coils of the two-terminal inductor element (22xp1) and the two-terminal inductor element (22xn1) are wound in a direction that adds up the excitation of the wound core (24x) with the current (I_cx1) passing through the first set of three-terminal leg circuits (2x1) (hereinafter referred to as "additive excitation"), and the two coils of the two-terminal inductor element (22xp2) and the two-terminal inductor element (22xn2) are wound in a direction that adds up the excitation of the wound core (24x) with the current (I_cx2) passing through the second set of three-terminal leg circuits (2x2),If the current (I_cx1) passing through the first set of three-terminal leg circuits (2x1) and the current (I_cx2) passing through the second set of three-terminal leg circuits (2x2) are of the same polarity, the four coils of the two-terminal inductors (22xp1, 22xn1, 22xp2, 22xn2) are wound around the winding core (24x) in a direction that causes summation excitation, the two-terminal inductors (22xp1) and the two-terminal inductors (22xn2) are arranged concentrically to form the first winding (25x1), and the two-terminal inductors (22xn1) and the two-terminal inductors (22xp2) are arranged concentrically to form the second winding (25x2), The first winding (25x1) and the second winding (25x2) are arranged coaxially on the winding core (24x), the two coils of the two-terminal inductor (22xp1) and the two-terminal inductor (22xp2) on the coaxial axis are arranged concentrically inside, and the two coils of the two-terminal inductor (22xn1) and the two-terminal inductor (22xn2) on the coaxial axis are arranged concentrically outside, or the two coils of the two-terminal inductor (22xp1) and the two-terminal inductor (22xp2) on the coaxial axis are arranged concentrically outside, and the two coils of the two-terminal inductor (22xn1) and the two-terminal inductor (22xn2) on the coaxial axis are arranged concentrically inside, The through-currents (I_cu, I_cv, I_cw) between the intermediate terminal (xC1) and the intermediate terminal (xC2) act on the winding core (24x) in an additive excitation direction by the two coils constituting the first winding (25x1), and act on the winding core (24x) in an additive excitation direction by the two coils constituting the second winding (25x2), and the first winding (25x1) and the second winding (25x2), which are coaxially arranged on the winding core (24x), act on each other in a differential excitation direction which is opposite to the additive excitation direction on the same axis, characterized in that the two parallel modular multilevel power converter (1) is configured such that the through-currents (I_cu, I_cv, I_cw) between the intermediate terminal (xC1) and the intermediate terminal (xC2) act on the winding core (24x) in an additive excitation direction.

2. The two-parallel modular multilevel power converter (1) according to claim 1, characterized in that one of the two concentric coils of the two-terminal inductor element (22xp1) and the two-terminal inductor element (22xn2) constituting the first winding (25x1) is right-handed (clockwise) and the other is left-handed (counterclockwise), one of the two concentric coils of the two-terminal inductor element (22xn1) and the two-terminal inductor element (22xp2) constituting the second winding (25x2) is right-handed (clockwise) and the other is left-handed (counterclockwise), and the coils of the two-terminal inductor element (22xp1) and the two-terminal inductor element (22xp2) are wound in the same direction.

3. The two-parallel modular multilevel power converter (1) according to claim 1, wherein a first DC reactor (71) is provided between the star-shaped connection terminals of the positive terminals (UP1, VP1, WP1) of the first set of three-terminal leg circuits (2u1, 2v1, 2w1) and the positive voltage terminal (P1 terminal) of the first DC power supply, a second DC reactor (72) is provided between the star-shaped connection terminals of the positive terminals (UP2, VP2, WP2) of the second set of three-terminal leg circuits (2u2, 2v2, 2w2) and the positive voltage terminal (P2 terminal) of the second DC power supply, and a common iron core (73) is provided to magnetically couple the first DC reactor (71) and the second DC reactor (72), A two-parallel modular multilevel power converter (1) is characterized in that, when the direction of current flowing through the positive voltage terminal (P1 terminal) of the first DC power supply and the positive voltage terminal (P2 terminal) of the second DC power supply are the same, the first DC reactor (71) and the second DC reactor (72) act in differential excitation mode, and when the current directions are opposite, they act in additive excitation mode.

4. A two-parallel modular multilevel power converter (1) according to claim 1, comprising: a positive side DC reactor (881p) having a positive side inner coil (883p) provided between the star-shaped connection terminals (UP1, VP1, WP1) of the first set of three-terminal leg circuits (2u1, 2v1, 2w1) and the positive side voltage terminal (P1 terminal) of the first DC power supply; and a positive side outer coil (884p) provided between the star-shaped connection terminals (UP2, VP2, WP2) of the second set of three-terminal leg circuits (2u2, 2v2, 2w2) and the positive side voltage terminal (P2 terminal) of the second DC power supply; A negative DC reactor (881n) comprises a negative inner coil (883n) provided between the star-shaped connection terminals (UN2, VN2, WN2) of the negative terminals (UN2, VN2, WN2) of the second set of three-terminal leg circuits (2u2, 2v2, 2w2) and the negative voltage terminal (N2 terminal) of the second DC power supply, and a negative outer coil (884n) provided between the star-shaped connection terminals (UN1, VN1, WN1) of the first set of three-terminal leg circuits (2u1, 2v1, 2w1) and the negative voltage terminal (N1 terminal) of the first DC power supply, A two-parallel modular multilevel power converter (1) is characterized in that the positive inner coil (883p) and the positive outer coil (884p), and the negative inner coil (883n) and the negative outer coil (884n) are configured to act as differential excitation when the direction of current flowing to the positive voltage terminal (P1 terminal) of the first DC power supply and the positive voltage terminal (P2 terminal) of the second DC power supply are the same, and as additive excitation when the direction of current is opposite.

Images