Direct current smoothing reactor and direct current rectification system

Abstract

The application relates to a direct-current smoothing reactor and a direct-current rectification system. The direct-current smoothing reactor comprises two spaced-apart magnetic columns, a magnetic circuit frame formed by a first end magnetic yoke and a second end magnetic yoke connected to the two ends of the two magnetic columns respectively, at least four coils arranged on the two magnetic columns in two groups, and at least one sub-magnetic circuit yoke comprising a longitudinal connecting part and a transverse branch part. The coils are arranged only on the magnetic columns, and no coils are arranged on the longitudinal connecting part, the transverse branch part and the end magnetic yoke; only one coil is arranged in each sub-magnetic circuit, and the number of sub-magnetic circuits is equal to the number of coils; each coil is electrically independent and is respectively provided with an independent terminal. By positioning the sub-magnetic circuit yoke and the end magnetic yoke as pure magnetic flux channels and making the sub-magnetic circuits correspond to the coils one by one, a plurality of independently-operated filtering coils can be integrated in the same magnetic circuit frame, and the magnetic flux of each coil is independently closed in the corresponding sub-magnetic circuit, and each loop is electrically independent.

Description

Technical Field

[0001] This application relates to the field of reactors, and in particular to a DC smoothing reactor and a DC rectifier system. Background Technology

[0002] In high-power DC power supply systems, the output current of the rectifier after converting AC to DC contains ripple components. A DC smoothing reactor needs to be connected in series at the rectifier output to suppress ripple and smooth the DC current. When the rectifier system uses a multi-pulse rectifier topology, each rectifier bridge's DC output circuit needs to be configured with an independent DC smoothing reactor to ensure that the filtering performance of each DC output does not interfere with each other. For example, a 24-pulse rectifier system with four sets of thyristor full-bridge rectifiers requires four DC smoothing reactors.

[0003] Existing DC smoothing reactors typically employ independent magnetic circuit structures, with each reactor equipped with its own set of magnetic pillars and yokes. Multiple reactors can only be placed side by side, causing the overall installation volume to increase linearly with the number of reactors. Furthermore, the leakage flux between reactors lacks a guiding path, leading to coupling interference with the magnetic circuits of adjacent reactors. Maintaining installation spacing further increases the volume. If multiple coils are integrated into the same magnetic circuit frame to reduce volume, the shared yoke causes magnetic flux coupling within the shared magnetic circuit section. Current changes in one coil affect the inductance values ​​of adjacent coils through magnetic flux coupling, and the filtering performance of each circuit interferes with each other and cannot be independently controlled. Moreover, the superposition of DC bias fluxes from multiple coils within the shared magnetic circuit section can easily cause local saturation of the iron core, a sharp decrease in inductance value with increasing current, and a deterioration in the smoothing effect.

[0004] Even though existing multi-winding magnetic core schemes attempt to separate the magnetic flux channels of each coil within a shared magnetic circuit framework through specific structural arrangements, they are still constrained by specific technical objectives in terms of the positional arrangement of the working windings, the electrical connection relationship between the windings, and the correspondence between the number of sub-magnetic circuits and the number of coils. This makes it difficult to simultaneously achieve independent magnetic flux closure of multiple independently operating filter coils within the same magnetic circuit framework, electrical independence of each circuit, and differentiated optimization of anti-saturation capabilities at each position. Therefore, there has been a long-standing desire to propose a DC smoothing reactor that can integrate multiple coils within the same magnetic circuit framework, with independent magnetic flux closure for each coil, electrical independence of each circuit, and differentiated optimization of anti-saturation capabilities at each position. Summary of the Invention

[0005] In order to integrate multiple coils into the same magnetic circuit frame and enable each coil to close its magnetic flux independently, each circuit to be electrically independent, and each position to have differentiated anti-saturation capability, this application provides a DC smoothing reactor, a DC rectifier system, and an inductance calibration method.

[0006] Firstly, this application provides a DC smoothing reactor, which adopts the following technical solution: A DC smoothing reactor, comprising: Two spaced magnetic pillars and a first end yoke and a second end yoke respectively connected to the two ends of the two magnetic pillars, the two magnetic pillars and the first end yoke and the second end yoke enclose to form a magnetic circuit frame. At least four coils are divided into two groups and wound on two magnetic pillars respectively, with at least two coils arranged on each magnetic pillar along the extension direction of the magnetic pillar; At least one sub-magnetic circuit yoke, the sub-magnetic circuit yoke including a longitudinal connecting part and a transverse branch part; the longitudinal connecting part is disposed between two magnetic pillars and connects the first end magnetic pillar and the second end magnetic pillar; the transverse branch part extends from the longitudinal connecting part to the magnetic pillar and is located between two adjacent coils on the same magnetic pillar; the corresponding sections of the transverse branch part, the longitudinal connecting part, the corresponding sections of the adjacent first end magnetic pillar or second end magnetic pillar, and the corresponding sections of the magnetic pillar enclose a sub-magnetic circuit, the sub-magnetic circuit yoke confining the magnetic flux generated by each coil within the corresponding sub-magnetic circuit; The coil is only wound on the magnetic column, and no coil is wound on the longitudinal connecting part, the transverse branch part, the first end yoke and the second end yoke; the number of sub-magnetic circuits is equal to the number of coils, and only one coil is wound in each sub-magnetic circuit, and each coil is only wound on the corresponding section of the magnetic column in one sub-magnetic circuit; Each coil is electrically independent of the others, and each coil has two independent terminals.

[0007] By adopting the above technical solution, the sub-magnetic circuit yoke and the end magnetic yoke are positioned as pure magnetic flux channels without coils, so that the magnetic flux generated by each coil is independently closed in the corresponding sub-magnetic circuit without spreading to adjacent sub-magnetic circuits; the sub-magnetic circuits correspond one-to-one with the coils, and each coil is electrically independent of each other, so that the current change of one coil will not affect the inductance value of other circuits through magnetic flux coupling or electrical connection. Multiple independently operating filter coils can be integrated in the same magnetic circuit frame, and the overall installation volume no longer increases linearly with the number of coils; the geometric position of the sub-magnetic circuit yoke and the end magnetic yoke does not need to take into account the coil installation space, and the cross-section, material and air gap configuration can be independently optimized according to the magnetic flux carrying characteristics of each position.

[0008] Optionally, the magnetic flux generated by each coil forms a lateral return path and a longitudinal crossing path at the yoke of the magnetic circuit; the lateral return path is the path through which the magnetic flux is conducted to the longitudinal connection through the lateral branch and returns to the coil through the sub-magnetic circuit; the longitudinal crossing path is the path through which the magnetic flux crosses the lateral branch along the magnetic column to reach the section where the adjacent coil is located; the magnetic reluctance of the lateral return path is lower than that of the longitudinal crossing path.

[0009] By adopting the above technical solution, the magnetic flux flows back naturally along the lateral return path with lower magnetic resistance without crossing into the adjacent sub-magnetic circuit. The sub-magnetic circuit yoke is upgraded from a geometric separation to a functional structure that actively guides the direction of magnetic flux by relying on the difference in magnetic resistance.

[0010] Optionally, each pair of coils, each wound around two magnetic pillars and adjacent to the same first end yoke or the same second end yoke, constitutes a coil pair; the winding directions of the two coils in each coil pair are opposite, such that the two coils generate magnetic flux in the same direction in the corresponding section of their respective magnetic pillars; the first end yoke and the second end yoke are each divided into two segments by the longitudinal connecting portion, and each segment of the first end yoke and each segment of the second end yoke carries only the magnetic flux of one coil; the magnetic flux of the two coils in the same coil pair flows through their respective segments of the first end yoke or the second end yoke to the longitudinal connecting portion, and converges in the same direction within the longitudinal connecting portion.

[0011] By adopting the above technical solution, after the end magnetic yoke is divided into two segments by the longitudinal connection, the magnetic flux of each of the two magnetic pillars flows independently in their respective end magnetic yoke segments and flows to the longitudinal connection. The two magnetic fluxes converge in the same direction in the longitudinal connection. The steady-state magnetic flux carried by the longitudinal connection is twice that of a single magnetic pillar, thus providing a geometric basis for subsequent position-based differential optimization of magnetic flux density.

[0012] Optionally, the steady-state magnetic flux carried by the longitudinal connecting portion is twice the steady-state magnetic flux carried by the corresponding section of a single magnetic column; air gaps are provided at at least two different positions among the magnetic column, the first end yoke, the second end yoke, and the longitudinal connecting portion, and the gap length and number of the air gaps at different positions are configured differently based on the steady-state magnetic flux density at each position; the total gap length of the air gaps provided on the longitudinal connecting portion is greater than the total gap length of the air gaps provided on a single magnetic column.

[0013] By adopting the above technical solution, the air gap length is configured differently according to the ratio of steady-state magnetic flux density at each position, so that the longitudinal connection and the magnetic column approach the saturation critical point at the same time under the rated current, thus avoiding the bottleneck of the overall machine's anti-saturation margin by saturating at a single position first.

[0014] Optionally, the longitudinal connecting part is made of a magnetic material with a saturation magnetic flux density greater than that of the magnetic column; the cross-sectional area of ​​the longitudinal connecting part is greater than that of the magnetic column; and the cross-sectional area of ​​the first end yoke of each segment and the cross-sectional area of ​​the second end yoke of each segment are smaller than that of the magnetic column.

[0015] By adopting the above technical solution, the differentiating matching of each position is carried out from two dimensions: material saturation magnetic flux density and core cross-sectional area. This, together with the air gap differentiation, forms a three-dimensional synergistic optimization of material, cross-section, and air gap, further reducing the overall core weight and cost.

[0016] Optionally, the magnetic column, the first end yoke, the second end yoke, and the branch magnetic circuit yoke are made of laminated silicon steel sheets, with the lamination direction parallel to the main direction of the magnetic flux.

[0017] By adopting the above technical solution, the insulation layer between each laminated silicon steel sheet suppresses eddy current loss in the iron core, and the iron loss is minimized when the magnetic flux is transmitted along the laminated plane.

[0018] Optionally, the materials of the magnetic column, the first end yoke, the second end yoke, and the split magnetic circuit yoke are selected according to the ripple frequency of the application scenario. When the ripple frequency is low, oriented silicon steel sheet is used; when the ripple frequency is medium, amorphous alloy strip is used; and when the ripple frequency is high, ferrite material is used.

[0019] By adopting the above technical solution, magnetic materials are matched according to the spectral characteristics of the application scenario, so that the iron loss and saturation margin under various rectification systems can be matched.

[0020] Optionally, the gap length of the air gap provided on the transverse branch is smaller than the gap length of the air gap provided on the magnetic column in the same sub-magnetic circuit. The air gap on the transverse branch includes a shim with adjustable thickness, which is used to compensate for the magnetic reluctance deviation between sub-magnetic circuits caused by the core machining tolerance and assembly gap during the assembly stage.

[0021] By adopting the above technical solution, the air gap of the transverse branch is used as a tool for assembly tolerance calibration, so that the actual magnetic resistance of each sub-magnetic circuit is consistent after assembly, thus ensuring the engineering consistency of the inductance value of each coil.

[0022] Optionally, the number of coils wound on the two magnetic pillars is equal, and the two coils located in the same position on the two magnetic pillars are symmetrically arranged about the middle plane between the two magnetic pillars; the upper and lower sections of the longitudinal connecting part and the upper and lower coils are mirror symmetrical about the transverse plane where the transverse branch is located.

[0023] By adopting the above technical solution, the magnetic circuit frame achieves mirror symmetry in both the left and right and up and down directions, the geometric characteristics of the four sub-magnetic circuits are completely identical, and the consistency of the inductance value of each coil is guaranteed by the dual constraint of biaxial symmetry.

[0024] Optionally, the number of coils is four, with two coils wound on each magnetic post; the number of magnetic circuit yokes is one, and the magnetic circuit yokes are set at the same height position between the two coils on each magnetic post.

[0025] By adopting the above technical solution, a single magnetic circuit yoke can be divided into four independent sub-magnetic circuits within the same magnetic circuit framework, resulting in the most compact structure, which is suitable for four independent DC smoothing scenarios such as a 24-pulse rectifier system.

[0026] Optionally, a permanent magnet sheet is embedded in the air gap provided on the longitudinal connecting part, and the magnetization direction of the permanent magnet sheet is opposite to the steady-state direction of the magnetic flux on the longitudinal connecting part.

[0027] By adopting the above technical solution, the permanent magnet sheet provides reverse magnetization so that the operating point of the longitudinal connection part deviates from the saturation region, and the DC bias is supplemented by a passive magnetization source, further expanding the anti-saturation margin.

[0028] Optionally, the connection position between the longitudinal connecting part and the first end yoke and the second end yoke forms a region with extremely low magnetic flux density because the magnetic flux from the two coil pairs converges in opposite directions at this position. The magnetic flux monitoring device or tap is embedded in this position without affecting the main magnetic circuit.

[0029] By adopting the above technical solution, the low magnetic field zero point position of the magnetic circuit frame itself serves as a natural candidate position for sensor or tap installation, eliminating the need for additional space or magnetic flux shielding structure.

[0030] Secondly, the DC rectifier system provided in this application adopts the following technical solution: A DC rectifier system, comprising: Rectifier transformer; Multiple rectifier bridges are connected to the output side of the rectifier transformer; As described above, in a DC smoothing reactor, each coil of the DC smoothing reactor is connected in series in the DC output circuit of a rectifier bridge.

[0031] By adopting the above technical solution, a single DC smoothing reactor replaces the multiple independent smoothing reactors originally required in the multi-pulse rectifier system. Each coil is responsible for suppressing the ripple output of one rectifier bridge. Since each coil is independently closed in the magnetic circuit and electrically isolated from each other, the output current change of one rectifier bridge will not affect the inductance value of the circuit where other rectifier bridges are located through the reactor. The overall installation volume, weight and cost of the multi-pulse rectifier system are significantly reduced, and the filtering performance of each circuit can be adjusted independently.

[0032] Optionally, the two coils in the same coil pair can be connected to the DC output circuits of two rectifier bridges with complementary phases.

[0033] By adopting the above technical solution, the common-mode noise between the two rectifier bridges is naturally suppressed in the end yoke due to reverse cancellation, and differential suppression capability is obtained without additional filtering circuit.

[0034] Thirdly, the inductance calibration method provided in this application adopts the following technical solution: A method for calibrating the inductance of the aforementioned DC smoothing reactor, wherein an air gap is provided on the transverse branch of the DC smoothing reactor, and the air gap on the transverse branch includes a shim with adjustable thickness; the inductance calibration method includes the following steps: S1. Apply AC test signals of the same frequency and amplitude to the two coils in each coil pair, and measure the inductance of the two coils respectively.

[0035] S2. Calculate the difference in inductance between the two coils in the same coil pair, and use it as the inductance difference of the coil pair.

[0036] S3. In response to the inductance difference exceeding the preset threshold, adjust the thickness of the pads on the transverse branches of the two sub-magnetic circuits where the coil pair is located, so that the inductance difference is reduced to below the preset threshold.

[0037] S4. Perform S1 to S3 sequentially on each coil pair until all coil pairs reach a state where the inductance difference is less than the preset threshold.

[0038] After calibration using the inductance calibration method, the sum of the total gap lengths of the air gaps set in each sub-magnetic circuit is equal between each sub-magnetic circuit.

[0039] By adopting the above technical solution, the reverse symmetry of two coils in the same coil center is used as the differential inductance measurement reference. The adjustable shims on the transverse branch are used as a special calibration handle. The core machining tolerance and assembly gap, which cannot be precisely controlled during the assembly stage, are converged to within the preset threshold through a closed-loop iterative method. The entire calibration process uses only the working winding itself as a sensor and does not introduce external measuring equipment or special calibration fixtures. The total length of the air gap of each sub-magnetic circuit of the calibration target is equal, which becomes a quantifiable criterion for the engineering consistency of the inductance value of each coil.

[0040] Optionally, the AC test signal in S1 is a small-signal sinusoidal voltage signal or a current signal, and the current amplitude generated in the coil by the AC test signal is less than the rated DC current of the coil; when S1 is executed, the terminals of the coil other than the coil being measured are kept open.

[0041] By adopting the above technical solution, the test signal is limited to a small-signal AC signal to avoid the iron core entering the saturation region and affecting the inductance measurement accuracy. The open circuit of the coil terminals other than the coil under test avoids the mutual inductance of adjacent coils from interfering with the inductance measurement results.

[0042] Optionally, during the operation of the DC smoothing reactor after it has been calibrated by the inductance calibration method, the inductance ratio of the two coils in each coil pair is continuously monitored. In response to the inductance ratio deviating from the preset normal operating range, an alarm signal indicating an abnormal state in the coil pair is triggered.

[0043] By adopting the above technical solution, the differential inductance measurement framework in the assembly stage is extended to the operation period. The working winding itself serves as a monitoring sensor during operation. An alarm is triggered when the inductance ratio deviates from the preset range, thus realizing online monitoring of the winding integrity without adding any auxiliary windings.

[0044] In summary, this application includes at least one of the following beneficial technical effects: 1. This application integrates multiple independently operating filter coils into the same magnetic circuit frame by positioning the sub-magnetic circuit yoke and the end magnetic yoke as pure magnetic flux channels without coils, with each sub-magnetic circuit corresponding to a coil and each coil electrically independent. This allows the magnetic flux of each coil to close independently in the corresponding sub-magnetic circuit, and each circuit to be electrically independent. The overall installation volume no longer increases linearly with the number of coils.

[0045] 2. This application uses a magnetic flux organization mode in which the coil is wound in opposite directions so that the two magnetic fluxes flow independently through the end yokes of each segment separated by the longitudinal connection and converge in the same direction in the longitudinal connection. Combined with the three-dimensional synergistic optimization of air gap, cross section and material according to the differential configuration of steady-state magnetic flux density at each position, the longitudinal connection and the magnetic column approach the saturation critical point at the same time under the rated current, and the overall anti-saturation margin is significantly improved.

[0046] 3. This application provides an inductance calibration method and an operation protection method based on coil-to-differential inductance. It uses only the working winding itself as a sensor and the adjustable shims on the transverse branch as calibration grippers to achieve tolerance closed-loop convergence during the assembly stage and online monitoring of winding integrity during operation without introducing any external measuring equipment or auxiliary windings. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the structure of a DC smoothing reactor in one embodiment of this application.

[0048] Figure 2 This is a schematic diagram of the topology of a DC rectifier system in one embodiment of this application.

[0049] Figure 3 This is a schematic diagram of the magnetic flux path within a single sub-magnetic circuit in one embodiment of this application.

[0050] Figure 4 This is a schematic diagram of the magnetic flux direction of the coil pair in the end yoke and longitudinal connecting portion in one embodiment of this application.

[0051] Explanation of reference numerals in the attached figures: 10. Rectifier transformer; 11. Magnetic column; 21. First end yoke; 22. Second end yoke; 30. Magnetic circuit yoke; 31. Longitudinal connection; 32. Lateral branch; 41. First coil; 42. Second coil; 43. Third coil; 44. Fourth coil; 50. Sub-magnetic circuit; 60. Gasket. Detailed Implementation

[0052] The present application will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of the application and are not intended to limit the scope of the application.

[0053] To facilitate understanding of the technical solution of this application, the following terms are explained.

[0054] Sub-magnetic circuit 50 refers to the closed magnetic flux path formed by the sub-magnetic circuit yoke 30, the end magnetic yoke, and the magnetic column 11. The magnetic flux generated by each coil is confined within the corresponding sub-magnetic circuit 50 and does not diffuse into adjacent sub-magnetic circuits 50. A coil pair refers to a combination of two coils adjacent to the same first end magnetic yoke 21 or the same second end magnetic yoke 22. The two coils in the same coil pair are wound at the same height position on the two magnetic columns 11. Steady-state magnetic flux density refers to the magnetic flux density in the core of the DC smoothing reactor at various positions under rated DC current, relative to the magnetic flux density at the moment of transient change of the ripple component. The total air gap length of the sub-magnetic circuit 50 refers to the sum of the air gap lengths provided at various locations on the magnetic column 11 section, the end magnetic yoke section, the longitudinal connecting part 31 section, and the transverse branch part 32 section constituting the same sub-magnetic circuit 50.

[0055] See attached document Figure 1 As shown, the DC smoothing reactor provided in this embodiment includes a magnetic circuit frame, four coils, and a sub-magnetic circuit yoke 30. Magnetic posts 11, a first end yoke 21, and a second end yoke 22 enclose and form the magnetic circuit frame. The four coils are wound in two groups around two magnetic posts 11, with two coils arranged along the extension direction of each magnetic post 11. The sub-magnetic circuit yoke 30 is disposed within the magnetic circuit frame, dividing the magnetic circuit frame into four independent sub-magnetic circuits 50.

[0056] In some embodiments, the number of coils wound on each magnetic post 11 can be more than two, and correspondingly the number of magnetic circuit yokes 30 can be increased to multiple, with multiple magnetic circuit yokes 30 respectively disposed between each pair of adjacent coils on the same magnetic post 11. This extended method also achieves the purpose of integrating multiple coils within the same magnetic circuit frame.

[0057] See attached document Figure 1 As shown, the magnetic circuit frame is formed by two spaced-apart magnetic pillars 11 and a first end yoke 21 and a second end yoke 22 connecting the two ends of the two magnetic pillars 11, respectively. The first end yoke 21 and the second end yoke 22 extend horizontally and connect to the same end of the two magnetic pillars 11, while the two magnetic pillars 11 extend vertically and form the two sidewalls of the magnetic circuit frame. The first end yoke 21, the second end yoke 22, and the two magnetic pillars 11 are smoothly joined together by the iron core mating surface, and a stacking pressure is applied to the mating surface to ensure that the magnetic circuit frame remains stable after assembly.

[0058] The magnetic column 11, the first end yoke 21, the second end yoke 22, and the branch magnetic circuit yoke 30 are made of laminated silicon steel sheets, with the lamination direction parallel to the main direction of the magnetic flux. An insulating layer is provided between the laminated silicon steel sheets to suppress eddy current losses in the iron core.

[0059] In some embodiments, the materials of the magnetic column 11, the first end yoke 21, the second end yoke 22, and the branch magnetic circuit yoke 30 are selected according to the ripple frequency of the application scenario. When applied to a power frequency rectification system, oriented silicon steel sheets are used to balance low loss and easy processing characteristics; when applied to a medium frequency rectification scenario, amorphous alloy strips are used to expand the saturation magnetic flux density; when applied to a high frequency rectification scenario, ferrite materials are used to reduce high-frequency eddy current losses. This material selection method ensures that the iron loss and saturation margin are matched under various rectification systems.

[0060] See attached document Figure 1 As shown, the magnetic circuit yoke 30 includes a longitudinal connecting portion 31 and a transverse branch portion 32. The longitudinal connecting portion 31 is disposed between two magnetic pillars 11, extends vertically, and connects the first end magnetic yoke 21 and the second end magnetic yoke 22. The transverse branch portion 32 extends horizontally from the longitudinal connecting portion 31 to each of the two magnetic pillars 11, and is located at the height between two adjacent coils on the same magnetic pillar 11. The longitudinal connecting portion 31 and the transverse branch portion 32 intersect at the central position, forming a cross-shaped magnetic circuit yoke 30.

[0061] The sub-magnetic circuit yoke 30 can be assembled from independent longitudinal connecting part 31 iron core blocks and independent transverse branch part 32 iron core blocks by bonding or snap-fitting. In other embodiments, the sub-magnetic circuit yoke 30 can also be obtained from a one-piece iron core by bending or stamping. In this embodiment, there is no assembly joint surface between the longitudinal connecting part 31 and the transverse branch part 32, and the magnetic resistance is more continuous. This one-piece molding method also achieves the purpose of constructing the sub-magnetic circuit yoke 30 within the magnetic circuit frame to separate the sub-magnetic circuits 50.

[0062] See attached document Figure 1 As shown, in this embodiment, the four coils are a first coil 41 wound on the upper end of the left magnetic post 11, a second coil 42 wound on the lower end of the left magnetic post 11, a third coil 43 wound on the upper end of the right magnetic post 11, and a fourth coil 44 wound on the lower end of the right magnetic post 11. Each coil has two independent terminals for connecting to external circuits. All coils are electrically independent of each other, and the circuit loops of each coil are not electrically connected to each other.

[0063] The coils can be implemented using a concentrated winding method, where each coil is continuously wound on a corresponding section of the magnetic post 11, forming a single-layer or multi-layer cylindrical winding structure. In other embodiments, the coils can also be implemented using a split winding method, where each coil is divided into multiple layers of vertically lapped sub-windings, with heat dissipation gaps between each sub-winding. Both winding methods can be supported by the magnetic circuit frame of this embodiment.

[0064] See attached document Figure 1 and attached Figure 3 As shown, the corresponding sections of the transverse branch 32, the longitudinal connecting portion 31, the corresponding sections of the adjacent first end yoke 21 or second end yoke 22, and the corresponding section of the magnetic column 11 enclose a sub-magnetic circuit 50. In this embodiment, the four coils respectively form four sub-magnetic circuits 50, and the section of the magnetic column 11 where each coil is located is located on the left or right side of a sub-magnetic circuit 50, while the transverse branch 32 is located on the upper or lower side of the sub-magnetic circuit 50.

[0065] See attached document Figure 3 The diagram illustrates the operation of a single sub-magnetic circuit 50. After the first coil 41 is energized and generates magnetic flux, the flux is transmitted upward along the magnetic column 11 to the first end yoke 21, then flows along the first end yoke 21 towards the longitudinal connecting portion 31, enters the upper half of the longitudinal connecting portion 31, flows downward along the upper half of the longitudinal connecting portion 31, and then flows back through the corresponding arm of the transverse branch portion 32 to the lower end of the section where the first coil 41 is located in the left magnetic column 11, finally returning to the section where the first coil 41 is located to form a closed loop. This loop coincides exactly with the sub-magnetic circuit 50 formed by the four components mentioned above. The magnetic flux generated by the first coil 41 is confined within the sub-magnetic circuit 50 corresponding to the first coil 41 and does not diffuse into adjacent sub-magnetic circuits 50.

[0066] The magnetic flux generated by each coil can take one of two paths at the yoke 30: a lateral return path and a longitudinal crossing path. The lateral return path is the path through which the magnetic flux is conducted to the longitudinal connection 31 via the lateral branch 32 and returns to the coil via the sub-magnetic circuit 50; the longitudinal crossing path is the path through which the magnetic flux travels along the magnetic column 11 across the lateral branch 32 to reach the section where the adjacent coil is located. Physically, the lateral return path directly connects the magnetic column 11 and the longitudinal connection 31 via the lateral branch 32, resulting in lower magnetic resistance; the longitudinal crossing path requires passing through a narrow section within the magnetic column 11 squeezed by the lateral branch 32 or an additional length bypassing the lateral branch 32, resulting in higher magnetic resistance. The magnetic flux naturally returns along the lateral return path with lower magnetic resistance, thus the magnetic flux is independently closed within the sub-magnetic circuit 50.

[0067] In some embodiments, the connection position between the longitudinal connecting portion 31 and the first end yoke 21 and the second end yoke 22 results in extremely low magnetic flux density and forms a local low magnetic field zero-point region because the magnetic flux from the two coil pairs converges in opposite directions at this position. A magnetic flux monitoring device or a tap can be embedded at this position to expand the sensing and tapping functions of the core without affecting the independent closing characteristics of the main magnetic circuit.

[0068] See attached document Figure 4As shown, the first coil 41 and the third coil 43 form an upper coil pair, both of which are adjacent to the first end yoke 21; the second coil 42 and the fourth coil 44 form a lower coil pair, both of which are adjacent to the second end yoke 22. The two coils in the same coil pair are wound in opposite directions, so that the two coils generate magnetic flux in the same direction in the corresponding sections of their respective magnetic pillars 11.

[0069] See attached document Figure 4 As indicated by the arrows, the first coil 41 generates an upward magnetic flux on the upper section of the left magnetic post 11, and the third coil 43 generates an upward magnetic flux on the upper section of the right magnetic post 11. The two magnetic fluxes enter the left and right sections of the first end yoke 21, which is separated by the longitudinal connecting part 31. The magnetic flux of the first coil 41 flows along the left section of the first end yoke 21 towards the longitudinal connecting part 31, and the magnetic flux of the third coil 43 flows along the right section of the first end yoke 21 towards the longitudinal connecting part 31. The two magnetic fluxes pass through the left and right sections of the upper end of the longitudinal connecting part 31. The magnetic flux enters the upper half of the longitudinal connecting section 31 at the merging position and flows downwards in the same direction within the upper half of the longitudinal connecting section 31. After merging, the magnetic flux splits at the height of the transverse branch 32. It flows back through the left arm of the transverse branch 32 to the lower end of the section where the first coil 41 is located in the upper section of the left magnetic column 11, forming the return flow of the sub-magnetic circuit 50 where the first coil 41 is located. It also flows back through the right arm of the transverse branch 32 to the lower end of the section where the third coil 43 is located in the upper section of the right magnetic column 11, forming the return flow of the sub-magnetic circuit 50 where the third coil 43 is located. The magnetic flux generated by the second coil 42 and the fourth coil 44 in the lower coil pair are symmetrically and independently closed in the lower half. The combined magnetic flux of the second coil 42 and the fourth coil 44 converges in the same direction in the lower half of the longitudinal connecting section 31 and then flows back through the left and right arms of the transverse branch 32 to their respective magnetic column 11 sections. The direction of magnetic flux flow is mirror-symmetrical with respect to the transverse plane where the transverse branch 32 is located. In the entire magnetic circuit frame, each end yoke independently carries the single coil magnetic flux injected by a magnetic column 11, and the upper and lower sections of the longitudinal connecting part 31 respectively carry the combined magnetic flux of the two coils of the upper coil pair and the lower coil pair.

[0070] The first end yoke 21 and the second end yoke 22 are each divided into two segments by the longitudinal connecting part 31. Each segment of the first end yoke 21 and each segment of the second end yoke 22 carries only the magnetic flux of one coil. The magnetic flux of the two coils in the same coil converges in the same direction in the corresponding half segment of the longitudinal connecting part 31. The steady-state magnetic flux carried by any half segment of the longitudinal connecting part 31 is twice the steady-state magnetic flux carried by a single magnetic column 11 segment.

[0071] In some embodiments, the number of coils wound on the two magnetic pillars 11 is equal, and the two coils located at the same position on the two magnetic pillars 11 are symmetrically arranged about the middle plane between the two magnetic pillars 11; the upper and lower sections of the longitudinal connecting portion 31 and the upper and lower coils are mirror-symmetrical about the transverse plane where the transverse branch portion 32 is located. The magnetic circuit frame achieves mirror symmetry in both the left-right and up-down directions, making the geometric characteristics of the four sub-magnetic circuits 50 completely identical.

[0072] See attached document Figure 1 As shown, air gaps are provided at at least two different locations among the magnetic column 11, the first end yoke 21, the second end yoke 22, and the longitudinal connecting portion 31. The gap length and number of air gaps at each location are configured differently based on the steady-state magnetic flux density carried at that location. In this embodiment, the steady-state magnetic flux carried by any half of the longitudinal connecting portion 31 is twice the steady-state magnetic flux carried by a single segment of the magnetic column 11, and the total gap length of the air gaps provided on the longitudinal connecting portion 31 is greater than the total gap length of the air gaps provided on a single magnetic column 11.

[0073] The total air gap length of each sub-magnetic circuit 50 is equal among all sub-magnetic circuits 50, so that each coil exhibits the same inductance value under the same current.

[0074] An air gap is provided on the transverse branch 32, and the air gap on the transverse branch 32 includes a shim 60 with adjustable thickness. The gap length of the air gap in the transverse branch 32 is smaller than the gap length of the air gap provided on the magnetic column 11 in the same sub-magnetic circuit 50, and is mainly used to compensate for the magnetic reluctance deviation between the sub-magnetic circuits 50 caused by the core machining tolerance and assembly gap during the assembly stage.

[0075] The air gap can be achieved by inserting a single section of non-magnetic material shim 60 into the mating surface of the magnetic circuit frame, and the thickness of the shim 60 is the length of the air gap. In other embodiments, the air gap can also be achieved by stacking multiple thin shims 60, and the total thickness of the stacked shims 60 is the length of the air gap. The stacking method allows for fine adjustment of the total length of the air gap by increasing or decreasing the number of thin shims 60.

[0076] To further accommodate the air gap differentiation, this embodiment differentiates the configuration of different positions in the magnetic circuit frame based on both material and core cross-sectional area. The longitudinal connecting portion 31 is made of a magnetic material with a saturation magnetic flux density greater than that of the magnetic column 11. The cross-sectional area of ​​the longitudinal connecting portion 31 is greater than that of the magnetic column 11. The cross-sectional areas of the first end yoke 21 and the second end yoke 22 of each segment are smaller than the cross-sectional area of ​​the magnetic column 11.

[0077] Since the steady-state magnetic flux carried by any half of the longitudinal connecting portion 31 is twice that of the single magnetic column 11, if the cross-sectional area remains the same, the steady-state magnetic flux density of the longitudinal connecting portion 31 will also be twice. By increasing the cross-sectional area of ​​the longitudinal connecting portion 31, the actual steady-state magnetic flux density of the longitudinal connecting portion 31 is reduced to less than twice, thus approaching the saturation critical point together with the magnetic column 11 and the end yoke half. The synergistic optimization of the three dimensions of material, cross-section, and air gap further reduces the overall core weight and manufacturing cost.

[0078] In other embodiments, a permanent magnet is embedded in the air gap provided on the longitudinal connecting portion 31. The magnetization direction of the permanent magnet is opposite to the steady-state direction of the magnetic flux on the longitudinal connecting portion 31, causing the operating point of the longitudinal connecting portion 31 to deviate from the saturation region. This advanced embodiment uses a passive magnetization source to supplement the DC bias effect, further expanding the anti-saturation margin.

[0079] See attached document Figure 2 As shown, the DC rectification system of this embodiment includes a rectifier transformer 10, four three-phase rectifier bridges, and the DC smoothing reactor described in the previous embodiment. The primary side of the rectifier transformer 10 is connected to the high-voltage AC bus via a circuit breaker; the secondary winding of the rectifier transformer 10 is divided into four groups, with the phases of each group of secondary windings staggered to form a multi-pulse rectification scheme; each group of secondary windings is connected to the AC side of a three-phase rectifier bridge; the DC output circuits of the four three-phase rectifier bridges are respectively connected in series with a coil of the DC smoothing reactor, and the four DC circuits converge into a common DC positive bus and DC negative bus at the coil output terminals.

[0080] See attached document Figure 2 As shown, the first coil 41, second coil 42, third coil 43, and fourth coil 44 of the DC smoothing reactor respectively handle ripple suppression for the outputs of the four three-phase rectifier bridges. Since the four coils achieve independent magnetic flux closure within the same magnetic circuit frame through the yoke 30, and each coil is electrically independent, changes in the output current of one three-phase rectifier bridge will not affect the inductance value of the circuit containing the other three-phase rectifier bridges through the reactor. The DC rectifier system of this embodiment is suitable for applications requiring multiple independent DC smoothing, such as 24-pulse rectifier systems.

[0081] In some embodiments, the two coils in the same coil pair are respectively connected to the DC output circuit of two three-phase rectifier bridges with complementary phases. The common-mode noise component between the two rectifier bridges is naturally suppressed in the end yokes due to reverse cancellation, thus achieving differential suppression capability without the need for additional filtering circuits.

[0082] This embodiment also provides an inductance calibration method for the above-mentioned DC smoothing reactor, including the following steps S1 to S4.

[0083] S1. Apply AC test signals of the same frequency and amplitude to the two coils in each coil pair, and measure the inductance of the two coils respectively.

[0084] Specifically, the two terminals of the AC test signal are connected to the two terminals of the coil under test respectively; during the test, the terminals of the other coil in the same coil pair, except for the one being measured, are kept open to avoid mutual inductance interference to the inductance value under test caused by magnetic flux coupling between adjacent coils.

[0085] The AC test signal can be a small-signal sinusoidal voltage or current signal, with the current amplitude generated in the coil being less than the coil's rated DC current. Small-signal AC signals prevent the iron core from entering the saturation region, thus avoiding a large deviation between the measured inductance value and the steady-state inductance value. In other embodiments, the AC test signal can also be a rectangular pulse signal, with the coil inductance value deduced by measuring the time constant of the pulse current response. Both test signal implementation methods can obtain the inductance value of each coil in the linear operating region.

[0086] S2. Calculate the difference in inductance between the two coils in the same coil pair, and use it as the inductance difference of the coil pair.

[0087] Specifically, the inductance measurements of the two coils in the same coil pair are denoted as L1 and L2. In some embodiments, the inductance difference ΔL is given in absolute difference form: ΔL = |L1−L2|. In other embodiments, the inductance difference is given in relative deviation ratio form: δL = 2|L1−L2| / (L1+L2). Both expressions can be used as the basis for comparison with a preset threshold in the subsequent step S3.

[0088] S3. In response to the inductance difference exceeding the preset threshold, the thickness of the pads 60 on the transverse branches 32 in the two sub-magnetic circuits 50 where the coil pair is located is adjusted respectively, so that the inductance difference is reduced to below the preset threshold.

[0089] A deviation of the inductance difference from zero indicates that the total magnetic reluctance of the two sub-magnetic circuits 50 containing the same coil is inconsistent; the coil with the smaller magnetic reluctance has a larger inductance, and the coil with the larger magnetic reluctance has a smaller inductance. In response, the thickness of the shim 60 in the air gap of the transverse branch 32 of the sub-magnetic circuit 50 with the smaller inductance is reduced, or the thickness of the shim 60 in the air gap of the transverse branch 32 of the sub-magnetic circuit 50 with the larger inductance is increased, so that the total magnetic reluctance of the two sub-magnetic circuits 50 tends to be consistent.

[0090] The thickness of the shim 60 can be adjusted using a continuously adjustable stepping screw device. Each rotation of the screw corresponds to a small step change in the thickness of the shim 60, facilitating precise approximation of the target thickness. In other embodiments, discrete prefabricated thin shims 60 can also be used for adjustment. Each thin shim 60 corresponds to a fixed thickness level, and the target thickness is achieved by replacing shims 60 at different levels or stacking multiple thin shims 60. Both adjustment methods can be completed during the assembly stage.

[0091] S4. Perform S1 to S3 sequentially on each coil pair until all coil pairs reach a state where the inductance difference is less than the preset threshold.

[0092] The preset threshold can be set with reference to the relative deviation of the rated inductance value, so that the deviation between the actual inductance value of each coil and the target value is controlled within the allowable range of engineering. In some embodiments, a preset upper limit for the number of iterations is set. If the adjustment of the same coil pair exceeds the upper limit and still fails to converge, an alarm for a defective iron core is triggered.

[0093] After calibration using the inductance calibration method, the total air gap length of each sub-magnetic circuit 50 is equal among all sub-magnetic circuits 50.

[0094] This embodiment also provides a protection method for the DC smoothing reactor during operation. After the DC smoothing reactor is calibrated using the aforementioned inductance calibration method and put into operation, the inductance ratio ρ = L1 / L2 of the two coils in each coil pair is continuously monitored.

[0095] Under normal operating conditions, the inductance values ​​of both coils in the same coil pair are close to the target value, and the inductance ratio ρ is close to 1.0. When there is a short circuit between coil turns, the inductance of the coil drops sharply, and the inductance ratio deviates from 1.0. When the coil is open-circuited, the coil has no inductance, and the inductance ratio tends to infinity or zero. In response to the inductance ratio ρ deviating from the preset normal operating range, an alarm signal indicating an abnormal state in the coil pair is triggered. The alarm signal can be used to disconnect the abnormal circuit or prompt maintenance personnel to perform maintenance.

[0096] In some embodiments, the voltage and current values ​​of the ripple component superimposed on the DC component in the main circuit are extracted, and the inductance value of each coil is inferred from the equivalent impedance at the ripple frequency, thus achieving non-invasive continuous measurement of the inductance during operation. In other embodiments, a small-signal AC test component with an amplitude much smaller than the rated DC current is injected into the coil, and the coil inductance value is inferred from the voltage response of the injected small signal. Neither of these lower-level implementations requires disconnecting the main circuit.

[0097] The operational protection method uses only the working winding itself as a sensor, without adding any auxiliary windings or dedicated monitoring devices. The operational protection method is an extension of the inductance calibration method used in the assembly stage, and both share the same differential inductance measurement framework. The above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A DC smoothing reactor, characterized in that, include: Two spaced magnetic pillars and a first end yoke and a second end yoke respectively connected to the two ends of the two magnetic pillars, the two magnetic pillars, the first end yoke and the second end yoke enclose to form a magnetic circuit frame; At least four coils are divided into two groups and wound on two magnetic pillars respectively, and at least two coils are arranged on each magnetic pillar along the extension direction of the magnetic pillar; At least one sub-magnetic circuit yoke, the sub-magnetic circuit yoke including a longitudinal connecting portion and a transverse branch portion; the longitudinal connecting portion is disposed between two magnetic pillars and connects the first end magnetic yoke and the second end magnetic yoke; the transverse branch portion extends from the longitudinal connecting portion to the magnetic pillar and is located between two adjacent coils on the same magnetic pillar; the transverse branch portion, the corresponding section of the longitudinal connecting portion, the corresponding section of the adjacent first end magnetic yoke or second end magnetic yoke, and the corresponding section of the magnetic pillar enclose a sub-magnetic circuit, the sub-magnetic circuit yoke confining the magnetic flux generated by each coil within the corresponding sub-magnetic circuit; The coil is wound only on the magnetic post, and the coil is not wound on the longitudinal connecting part, the transverse branch part, the first end yoke and the second end yoke; the number of sub-magnetic circuits is equal to the number of coils, and only one coil is wound in each sub-magnetic circuit, and each coil is wound only on the corresponding section of the magnetic post in one sub-magnetic circuit; Each coil is electrically independent of the others, and each coil has two independent terminals.

2. The DC smoothing reactor according to claim 1, characterized in that, The magnetic flux generated by each coil forms a lateral return path and a longitudinal crossing path at the yoke of the magnetic circuit; the lateral return path is the path through which the magnetic flux is conducted to the longitudinal connection via the lateral branch and returns to the coil via the sub-magnetic circuit; the longitudinal crossing path is the path through which the magnetic flux crosses the lateral branch along the magnetic column to reach the section where the adjacent coil is located; the magnetic resistance of the lateral return path is lower than the magnetic resistance of the longitudinal crossing path.

3. The DC smoothing reactor according to claim 1, characterized in that, Each pair of coils, each wound around one of the two magnetic pillars and adjacent to the same first end yoke or the same second end yoke, forms a coil pair. The two coils in each coil pair are wound in opposite directions, such that the two coils generate magnetic flux in the same direction within the corresponding section of their respective magnetic pillars. The first end yoke and the second end yoke are each divided into two segments by the longitudinal connecting portion, and each segment of the first end yoke and each segment of the second end yoke carries the magnetic flux of only one coil. The magnetic flux of the two coils in the same coil pair flows through their respective segments of the first end yoke or the second end yoke to the longitudinal connecting portion, and converges in the same direction within the longitudinal connecting portion.

4. The DC smoothing reactor according to claim 3, characterized in that, The steady-state magnetic flux carried by the longitudinal connecting portion is twice that carried by the corresponding section of a single magnetic column; air gaps are provided at at least two different positions of the magnetic column, the first end yoke, the second end yoke, and the longitudinal connecting portion, and the gap length and number of the air gaps at different positions are configured differently based on the steady-state magnetic flux density at each position; the total gap length of the air gaps provided on the longitudinal connecting portion is greater than the total gap length of the air gaps provided on a single magnetic column.

5. The DC smoothing reactor according to claim 3, characterized in that, The longitudinal connecting portion is made of a magnetic material with a saturation magnetic flux density greater than that of the magnetic column; the cross-sectional area of ​​the longitudinal connecting portion is greater than that of the magnetic column; the cross-sectional area of ​​each segment of the first end yoke and the cross-sectional area of ​​each segment of the second end yoke are both smaller than that of the magnetic column.

6. A DC rectifier system, characterized in that, include: Rectifier transformer; Multiple rectifier bridges are connected to the output side of the rectifier transformer; The DC smoothing reactor as described in any one of claims 1 to 5, wherein each of the coils of the DC smoothing reactor is connected in series in the DC output circuit of one of the rectifier bridges.

7. A method for calibrating the inductance of a DC smoothing reactor as described in any one of claims 3 to 5, characterized in that, An air gap is provided on the lateral branch of the DC smoothing reactor, and the air gap on the lateral branch includes a shim with adjustable thickness; the inductance calibration method includes the following steps: S1. Apply an AC test signal of the same frequency and amplitude to each of the two coils in each coil pair, and measure the inductance value of each of the two coils respectively; S2. Calculate the difference in inductance values ​​between the two coils in the same coil pair as the inductance difference value of the coil pair; S3. In response to the inductance difference exceeding a preset threshold, adjust the thickness of the pads on the transverse branches of the two sub-magnetic circuits where the coil pair is located, so that the inductance difference is reduced to below the preset threshold; S4. Perform S1 to S3 sequentially on each coil pair until all coil pairs reach a state where the inductance difference is less than the preset threshold. After calibration using the aforementioned inductance calibration method, the total air gap length of each of the sub-magnetic circuits is equal among the sub-magnetic circuits.

8. The inductance calibration method according to claim 7, characterized in that, The AC test signal in S1 is a small-signal sinusoidal voltage signal or current signal, and the current amplitude generated by the AC test signal in the coil is less than the rated DC current of the coil; when S1 is executed, the terminals of the other coil in the coil pair, except for the one being measured, remain open.

9. The inductance calibration method according to claim 7, characterized in that, During the operation of the DC smoothing reactor after it has been calibrated by the inductance calibration method, the inductance ratio of the two coils in each coil pair is continuously monitored. In response to the inductance ratio deviating from the preset normal operating range, an alarm signal indicating an abnormal state in the coil pair is triggered.

Images