Low-loss high-current rectifier with online identification of parallel branches and control method

Abstract

The application discloses a low-loss large-current rectifier with parallel branch online identification and a control method. The rectifier comprises at least two parallel controllable rectification branches, a branch current detection unit, a branch voltage detection unit, a temperature detection unit, an independent driving unit and a controller. The controller obtains the current and the conduction voltage drop of the same branch at two different current working points in the same rectification conduction period, determines the dynamic conduction parameter according to the voltage drop change and the current change between the two working points, and determines the connection dynamic resistance according to the device dynamic conduction parameter and the total dynamic conduction parameter of the branch. The controller determines the target current of each branch, adjusts the conduction window, and executes the pre-warning, derating, cutting-off or shutdown protection on the abnormal branch. The application does not need to connect a large-power current-sharing resistor in series, can reduce the current-sharing loss and improve the operation reliability.

Description

Technical Field

[0001] This invention relates to the field of power electronic conversion and power semiconductor technology, specifically to a low-loss high-current rectifier and control method for online identification of parallel branches. Background Technology

[0002] In high-current DC power supplies, industrial electrolysis power supplies, electroplating power supplies, charging power supplies, energy storage devices, welding power supplies, and communication power supplies, rectifier circuits typically need to carry large output currents. Due to limitations in the rated current, conduction losses, and heat dissipation capacity of individual power semiconductor devices, multiple power semiconductor devices or rectifier branches connected in parallel can be set in the rectifier bridge arm, with the rectified current being shared by multiple parallel branches.

[0003] However, even if multiple parallel power semiconductor devices use the same model, their actual conduction parameters may still differ. Furthermore, the length of the connecting conductors, the condition of the terminals, the tightness of the fastening, the contact interface condition, and the heat dissipation conditions of each branch may also vary. These differences can lead to uneven current distribution among the parallel branches, with branches with relatively lower on-resistance potentially carrying a larger current, resulting in higher temperature rise.

[0004] As operating time increases, the contact resistance of the connection interface may change due to thermal cycling, oxidation, loosening, or localized damage. Relying solely on device factory parameter selection or symmetrical conductor design is insufficient to compensate for the continuous parameter drift that occurs during the operation of high-current rectifiers.

[0005] Existing current sharing methods can set current sharing resistors in each parallel branch to limit the difference in branch currents using additional resistance. However, current sharing resistors generate additional conduction losses and heat, which is not conducive to improving rectification efficiency. Current sharing can also be achieved by detecting the current in each branch and adjusting the drive state of the power devices. However, adjusting only based on the magnitude of the branch current makes it difficult to distinguish whether the current bias is caused by changes in the performance of the power devices themselves, increased connection resistance, increased temperature, or abnormal drive.

[0006] The on-state voltage drop of power semiconductor devices typically includes both a fixed voltage drop component related to the device structure and a dynamic voltage drop component that varies with the current. If the on-state voltage drop is measured only at a single operating point, the measurement result is easily affected by the fixed voltage drop, sampling bias, temperature changes, and current magnitude, making it difficult to accurately reflect the actual conduction state of the branch.

[0007] Therefore, there is a need for a rectifier and control method that can identify the dynamic conduction state of parallel branches online during the operation of a high-current rectifier without relying on additional high-power current-sharing resistors, and correct the current distribution of each branch based on the identification results. Summary of the Invention

[0008] To address the aforementioned technical problems, this invention provides a low-loss, high-current rectifier and control method for online identification of parallel branches, thereby solving the problem of bias current generated in multiple parallel rectifier branches in the prior art due to discrete power device parameters, changes in connection resistance, and temperature differences.

[0009] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows: This invention discloses a low-loss, high-current rectifier with online identification of parallel branches, comprising an AC input unit, a rectifier circuit, a DC output unit, and a controller.

[0010] The rectifier circuit includes at least one rectifier bridge arm, and at least two controllable rectifier branches are arranged in parallel within the rectifier bridge arm. Each controllable rectifier branch includes a controllable power semiconductor device, a branch current detection unit, a branch voltage detection unit, and an independent drive unit.

[0011] The controllable power semiconductor device can be a MOSFET, a SiC MOSFET, an IGBT, or a composite rectifier unit consisting of a rectifier diode and a controllable bypass device. Preferably, the controllable power semiconductor device is a MOSFET or a SiC MOSFET to form a synchronous rectifier branch whose conduction start and end times can be independently controlled.

[0012] The branch current detection unit is used to detect the instantaneous branch current or the average branch current over one rectification conduction cycle of the corresponding controllable rectifier branch. The branch current detection unit can employ a shunt resistor, a Hall current sensor, a magnetoresistive current sensor, or a current detection structure based on the voltage drop of the connecting conductor.

[0013] The branch voltage detection unit is used to detect the on-state voltage drop of the corresponding controllable rectifier branch. The branch voltage detection unit preferably employs a Kelvin detection structure to separate the voltage sampling path from the main current path.

[0014] The independent drive unit is connected to the control terminal of the corresponding controllable power semiconductor device and is used to independently control the conduction start time, conduction end time and conduction duration of the corresponding controllable power semiconductor device according to the drive command output by the controller.

[0015] The controller is connected to each branch current detection unit, branch voltage detection unit, and independent drive unit. Within the same rectification conduction cycle, the controller acquires the branch current and branch conduction voltage drop of the same controllable rectifier branch at the first and second operating points.

[0016] The first operating point and the second operating point correspond to different branch currents. The time interval between the two operating points is less than the preset identification time, so that the device temperature, the temperature of the connecting conductor, and the state of the connection interface between the two operating points remain basically continuous.

[0017] The controller determines the dynamic conduction parameters of the corresponding controllable rectifier branch based on the changes in branch on-state voltage drop and branch current between two operating points. Let the branch current and branch on-state voltage drop of the i-th controllable rectifier branch at the first operating point be I... i1 and V i1 The branch current and branch on-state voltage drop at the second operating point are respectively I i2 and V i2 Then the dynamic conduction parameter R of this branch i R can be determined according to the following formula: i = (V i2 -V i1 ) / (I i2 -I i1 ).

[0018] Among them, I should satisfy i2 with I i1 The currents are not equal, and the absolute value of the current change between the two is not less than the preset identification threshold. When the current change is less than the preset identification threshold, the controller discards the current sampling result and reselects an operating point or actively establishes a new operating point.

[0019] The controller determines the corresponding target current based on the dynamic conduction parameters of each controllable rectifier branch, and adjusts the conduction window of the corresponding controllable power semiconductor device based on the deviation between the actual branch current and the target current.

[0020] The conduction window refers to the time interval during which a controllable power semiconductor device is in a controlled conduction state within the rectification conduction cycle that allows the corresponding rectifier bridge arm to conduct. The conduction window is determined by at least two of the following parameters: conduction start time, conduction end time, and conduction duration.

[0021] When the actual current of a controllable rectifier branch is higher than the target current, the controller delays the start time of conduction of that branch, advances its end time of conduction, or shortens its conduction duration.

[0022] When the actual current of a controllable rectifier branch is lower than the target current and the branch is in a normal state, the controller advances the start time of the branch's conduction, delays the end time of its conduction, or extends the duration of its conduction.

[0023] Preferably, the controller performs complementary adjustments to multiple parallel controllable rectifier branches. When the conduction window of one branch shortens, the conduction window of at least one other branch that is in normal condition is correspondingly extended, so that the total output current of all parallel branches remains within a preset fluctuation range.

[0024] (III) Double-layer voltage detection scheme In a preferred embodiment, the branch voltage detection unit includes a first voltage detection component and a second voltage detection component.

[0025] The two sampling terminals of the first voltage detection component are respectively connected near the two main electrodes of the controllable power semiconductor device, and are used to detect the on-state voltage drop across the controllable power semiconductor device. The sampling wires of the first voltage detection component are arranged separately from the main current conductor to reduce the influence of the input connection conductor, output connection conductor, and connection interface voltage drop on the device voltage drop detection results.

[0026] The two sampling terminals of the second voltage detection component are connected to the branch input and branch output terminals of the corresponding controllable rectifier branch, respectively, to detect the total on-state voltage drop of the branch. The total on-state voltage drop of the branch includes the device on-state voltage drop of the controllable power semiconductor device, the input connection conductor voltage drop, the output connection conductor voltage drop, and the connection interface voltage drop.

[0027] The controller determines the dynamic on-state parameter R of the device based on the on-state voltage drop detected by the first voltage detection component at the first and second operating points. i_device :R i_device = (V i_device2 -V i_device1 ) / (I i2 -I i1 ).

[0028] The controller determines the total dynamic conduction parameter R of the branch based on the total on-state voltage drop detected by the second voltage detection component at the first and second operating points. i_total :R i_total = (V i_total2 -V i_total1 ) / (I i2 -I i1 ).

[0029] The controller determines the connection dynamic resistance R based on the difference between the total dynamic conduction parameters of the branch and the dynamic conduction parameters of the devices. i_connection :R i_connection =R i_total -R i_device .

[0030] The connection dynamic resistance is used to characterize the overall conduction state of the input connection conductor, output connection conductor, terminal block and connection interface.

[0031] In actual calculations, the controller can perform averaging, weighted averaging, median filtering, or low-pass filtering on multiple identification results to reduce measurement noise. When the calculated connection dynamic resistance is less than zero due to detection error, the result can be deemed invalid, or the connection dynamic resistance can be limited to not less than zero.

[0032] (iv) Temperature detection and normalization scheme Each controllable rectifier branch may also be equipped with a temperature detection unit. The temperature detection unit may be located on the package surface, heat sink, input connection conductor, output connection conductor, or near the connection terminal of the controllable power semiconductor device.

[0033] The controller stores the compensation relationship between dynamic conduction parameters and temperature, and converts the actual dynamic conduction parameters into normalized dynamic conduction parameters at the reference temperature based on the branch temperature detected by the temperature detection unit.

[0034] For the connecting conductor and the connection interface, the controller performs normalization compensation based on the temperature characteristics of the connection area. Let the actual temperature of the connection area be T. i The reference temperature is T. ref If the equivalent temperature coefficient of the connection region is α, then the normalized connection dynamic resistance can be expressed as: R i_connection_ref =R i_connection / [1+α(T i -T ref )].

[0035] The equivalent temperature coefficient α is obtained through prototype calibration; when the dynamic resistance is mainly generated by copper connecting conductors, the temperature coefficient of the corresponding conductor material can be used; when the contact interface resistance accounts for a large proportion or the temperature characteristics are nonlinear, piecewise functions, fitting curves, or table lookup methods can be used for temperature compensation. i Preferably, it refers to the temperature of the connection area detected near the connection terminal, connection conductor, or contact interface.

[0036] For controllable power semiconductor devices, temperature compensation can be performed using preset compensation coefficients, piecewise functions, or lookup tables, depending on the device type.

[0037] When the high-current rectifier is put into operation for the first time, restarted after maintenance, or in a preset low-load calibration state, the controller records the initial dynamic conduction parameters of the devices and the initial dynamic resistance of the connection of each branch as the reference parameters of the corresponding branch.

[0038] During subsequent operation, the controller will compare the real-time normalized parameters with the baseline parameters of the corresponding branch to identify long-term changes in the branch parameters, rather than being limited to comparing the absolute parameters between different branches.

[0039] (v) Methods for establishing the first and second working points The first and second operating points can be formed by natural current changes or active perturbations.

[0040] In the natural current variation mode, the controller selects two sampling moments within the monotonic variation range of the branch current during the same rectification conduction cycle. The first sampling moment corresponds to the first operating point, and the second sampling moment corresponds to the second operating point.

[0041] The two sampling times should avoid the rapid switching transients when the controllable power semiconductor device just starts to conduct, the drive oscillation range, and the commutation transients when it is about to turn off, in order to reduce the impact of parasitic inductance voltage drop and switching noise on the identification results.

[0042] In the active perturbation mode, the controller applies short-time pulse width modulation or conduction window perturbation to the controllable rectifier branch to be identified within the same rectifier conduction cycle, so that the branch forms a first conduction sampling sub-interval and a second conduction sampling sub-interval with different current magnitudes, and selects the first operating point and the second operating point in the stable part of the two conduction sampling sub-intervals respectively.

[0043] Both the first and second operating points should be located within the range where the controllable power semiconductor device is in the on state. The device off-state range, switching transient, commutation transient, or drive oscillation range should not be used as the identification operating point.

[0044] The amplitude of the active disturbance is less than the normal protection adjustment amplitude and meets at least one of the following limitations: the change in branch current does not exceed the preset proportion of the rated branch current; the fluctuation of the total output current of the high current rectifier does not exceed the preset allowable value; the DC output voltage ripple does not exceed the preset allowable value; and the peak current of the controllable power semiconductor device does not exceed the safety threshold.

[0045] When an active disturbance is applied to the branch to be identified, the controller can perform a compensation adjustment in the opposite direction to at least one other normal controllable rectifier branch. For example, when the conduction duration of the branch to be identified is shortened by a preset time, the conduction duration of the other normal branch is correspondingly extended to maintain the overall output current basically stable.

[0046] (vi) Target Current Determination Scheme The controller determines the current distribution weight of each branch based on the dynamic conduction parameters, branch temperature, and operating status of each branch.

[0047] When all controllable rectifier branches are in normal condition and the normalized dynamic conduction parameters of each branch are within the allowable range, each branch can have the same or basically the same current distribution weight.

[0048] Let I be the total current that the rectifier bridge arm needs to handle. total There are n normally controllable rectifier branches, and the current distribution weight of the i-th branch is W. i Then the target current I of this branch i_target It can be represented as: I i_target =Itotal ×W i / ΣW i .

[0049] Among them, ΣW i This represents the sum of the current distribution weights of all controllable rectifier branches currently in operation.

[0050] When the dynamic conduction parameter of a branch increases or the branch temperature rises, the controller reduces the current distribution weight of that branch; when the branch parameter returns to the normal range, its current distribution weight can be gradually restored.

[0051] The current allocation weight can be determined based on at least one of the following parameters: the rate of change of the device's dynamic conduction parameters relative to a reference value; the rate of change of the connection's dynamic resistance relative to a reference value; the deviation of the branch temperature relative to a reference temperature or temperature threshold; the number of historical anomalies in the branch; the branch's allowable current carrying capacity; and the branch's current operating status.

[0052] With the above settings, the present invention does not require all branches to share the current equally under all conditions. Instead, it implements current sharing when each branch is in normal condition and implements controlled derating when a branch has a slight abnormality.

[0053] (vii) Conductive window calibration scheme The controller periodically calculates the average actual current I of each branch. i_actual And determine the current deviation e i :e i =I i_target -I i_actual .

[0054] The controller adjusts the conduction window of the corresponding branch based on the current deviation. The conduction window can be adjusted using proportional control, proportional-integral control, step-by-step control, or lookup table control.

[0055] When using proportional-integral control, the correction amount Δt for the conduction duration i It can be expressed as: Δt i =K p ×e i +K i ×Σe i Among them, K p K is the proportional adjustment coefficient. i Σe is the integral adjustment coefficient. i This represents the cumulative current deviation within a preset time range.

[0056] To avoid significant changes in the total output due to the simultaneous increase in conduction time of multiple branches, the controller can perform zero-sum processing on the conduction window correction of multiple branches, so that the sum of the conduction window correction of multiple branches is close to zero.

[0057] The controller also sets the following constraints on the conduction window: the conduction start time must not be earlier than the allowable conduction time of the corresponding rectifier bridge arm; the conduction end time must not be later than the allowable turn-off time of the corresponding rectifier bridge arm; a preset dead time is maintained between adjacent commutator bridge arms; the correction amount of each conduction window does not exceed the preset maximum correction amount; the peak current of the branch does not exceed the peak current threshold; when the branch is in an abnormal state, the control to increase the conduction window is not executed.

[0058] (viii) Anomaly Identification and Protection Scheme The controller determines the operating status of each branch based on the device's dynamic conduction parameters, connection dynamic resistance, branch temperature, and branch current.

[0059] When the dynamic conduction parameters of a device exceed the abnormal threshold of the device, while the connection dynamic resistance is within the preset normal connection range, the corresponding branch is identified as a device performance degradation branch.

[0060] When the connection dynamic resistance exceeds the connection abnormality threshold, but the device dynamic conduction parameters are within the preset normal range, the corresponding branch is identified as the connection abnormality branch.

[0061] When the temperature of a branch exceeds the temperature threshold, the corresponding branch is identified as an overheated branch.

[0062] When the actual current of a branch cannot follow the target current for a long period of time, while the dynamic conduction parameters of the device, the dynamic resistance of the connection, and the temperature of the branch are all within the normal range, the corresponding branch can be identified as a driving abnormal branch or a detection abnormal branch.

[0063] The controller performs graded protection based on the severity of the anomaly: when the parameter reaches the warning threshold but not the fault threshold, it outputs a warning signal and records the anomaly information; when a branch has a minor anomaly, it reduces the target current and current distribution weight of that branch; when the branch parameter exceeds the fault threshold, it shuts down the corresponding independent drive unit and disconnects the abnormal branch from the parallel rectifier branch; when the number of normal branches is insufficient or the total carrying capacity of the remaining branches is lower than the load requirement, it reduces the output power of the high-current rectifier; when multiple branches have serious anomalies at the same time, it stops the operation of the high-current rectifier.

[0064] (ix) Control methods S1. Branch Status Acquisition: Acquire the branch current, branch on-state voltage drop, and branch temperature of each controllable rectifier branch. When a double-layer voltage detection structure is set up, the device on-state voltage drop of the controllable power semiconductor device and the total on-state voltage drop of the corresponding controllable rectifier branch are also acquired.

[0065] S2. Establishment of Operating Points: Within the same rectification conduction cycle, establish a first operating point and a second operating point with different current magnitudes for the controllable rectifier branch to be identified, both of which are located within the stable sampling range where the controllable power semiconductor device is in the conduction state. The first and second operating points can be formed naturally by changes in the branch current, or they can be actively formed by the controller applying short-time pulse width modulation or conduction window perturbation to the branch to be identified.

[0066] S3. Obtaining Operating Point Parameters: Obtain the first branch current and first branch on-state voltage drop at the first operating point, and the second branch current and second branch on-state voltage drop at the second operating point. In the dual-layer voltage detection method, obtain the device on-state voltage drop and the total branch on-state voltage drop at both operating points. If the current change between the two operating points is less than the preset identification threshold, the controllable power semiconductor device corresponding to any operating point is not in the on state, or switching oscillation, overcurrent, or commutation abnormality occurs during sampling, discard the current sampling data and re-execute step S2.

[0067] S4. Dynamic Parameter Identification: Based on the voltage drop change between the first and second branch conduction voltage drops, and the current change between the first and second branch currents, the dynamic conduction parameters of the controllable rectifier branch to be identified are determined. In the dual-layer voltage detection method, the dynamic conduction parameters of the device and the total dynamic conduction parameters of the branch are determined separately, and the connection dynamic resistance is determined based on the difference between the two. When setting up the temperature detection unit, the dynamic conduction parameters of the device and the connection dynamic resistance are normalized for temperature.

[0068] S5. Target Current Determination: Based on the dynamic conduction parameters, temperature, and operating status of each controllable rectifier branch, determine the current allocation weight of each branch, and determine the target current of each branch based on the total current of the rectifier bridge arm and the current allocation weight of each branch. For branches in normal condition, maintain the normal current allocation weight; for branches with increased dynamic conduction parameters or increased temperature but not reaching fault conditions, reduce their current allocation weight.

[0069] S6. Conductive Window Calibration: Based on the deviation between the actual current and the corresponding target current of each controllable rectifier branch, adjust the conduction start time, conduction end time, or conduction duration of the corresponding branch. When the actual current of a branch is higher than the target current, shorten the conduction window of that branch; when the actual current of a normal branch is lower than the target current, extend the conduction window of that branch. The conduction windows of multiple branches are preferably adjusted complementaryly to keep the total output current of all branches within a preset fluctuation range.

[0070] S7. Current Sharing Status Verification: In subsequent rectification conduction cycles, the branch current of each branch is reacquired to determine whether the current deviation between each branch is within the preset current sharing range. If the current deviation is not within the preset current sharing range, steps S5 and S6 are executed again; if the current deviation is within the preset current sharing range, the current conduction window is maintained, and dynamic parameter identification is re-executed according to the preset cycle.

[0071] S8. Anomaly Identification and Protection: Identify degraded branches, abnormally connected branches, and overheated branches based on device dynamic conduction parameters, connection dynamic resistance, and branch temperature. Depending on the anomaly type and severity, execute at least one protective operation, including reducing the target current, outputting a warning signal, disconnecting the abnormal branch, or stopping the operation of the high-current rectifier.

[0072] Compared with the prior art, the present invention has at least the following beneficial effects: 1. Achieve parallel branch current calibration without connecting a high-power current-sharing resistor in series, reducing additional conduction losses; 2. By using differential sampling at two operating points within the same rectification and conduction cycle, the impact of fixed on-state voltage drop and sampling zero-point deviation on the identification results is reduced; 3. The short time interval between the two operating points can reduce the interference of slow temperature changes on the dynamic parameter identification results; 4. It can acquire the dynamic conduction parameters of each parallel branch online, and compensate for the parameter dispersion of the device and the parameter drift caused by long-term operation; 5. By using a dual-layer voltage detection structure, the state of the power device itself is separated from the state of the branch connection; 6. Able to identify increased dynamic resistance in connections caused by loose terminals, oxidation of contact interfaces, and abnormalities in the connecting conductors; 7. Determine the target current based on the health status of each branch to prevent abnormal branches from continuing to bear large currents; 8. By using complementary adjustment of the conduction window, the total output current fluctuation is reduced while correcting the branch current distribution; 9. When a single branch malfunctions, the branch can be derated or disconnected to improve the continuous operation capability of the high-current rectifier; 10. The main improvements are focused on the detection circuit and controller, with a simple mechanical structure that is easy to implement on the basis of existing high-current rectifier topologies.

[0073] Of course, products implementing this invention need not achieve all of the above-mentioned technical effects at the same time. Attached Figure Description

[0074] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the overall structure of the low-loss, high-current rectifier of the present invention.

[0075] Figure 2 This is a schematic diagram of a single controllable rectifier branch and a double-layer voltage detection structure of the present invention.

[0076] Figure 3 This is a schematic diagram of the sampling timing of the first and second operating points within the same rectification and conduction cycle of the present invention.

[0077] Figure 4 This is a schematic diagram illustrating the complementary adjustment of the conduction windows of multiple parallel controllable rectifier branches according to the present invention.

[0078] Figure 5 This is a flowchart of the online identification and current sharing calibration control method for parallel rectifier branches according to the present invention.

[0079] In the diagram: 1. Low-loss high-current rectifier; 10. AC input unit; 20. Rectifier circuit; 21. Controllable rectifier branch; 211. Controllable power semiconductor device; 212. Branch input terminal; 213. Branch output terminal; 214. Input connection conductor; 215. Output connection conductor; 22. Rectifier bridge arm; 30. DC output unit; 40. Controller; 41. Data acquisition module; 42. Dynamic parameter identification module; 43. Target current determination module; 44. On / off state. Window calibration module; 45. Current sharing status verification module; 46. Anomaly identification and protection module; 50. Branch current detection unit; 60. Branch voltage detection unit; 61. First voltage detection component; 611. First device voltage sampling terminal; 612. Second device voltage sampling terminal; 62. Second voltage detection component; 621. First branch voltage sampling terminal; 622. Second branch voltage sampling terminal; 70. Temperature detection unit; 80. Independent drive unit; 90. Protection unit. Detailed Implementation

[0080] The following will describe in detail the implementation of the present invention with reference to the accompanying drawings and embodiments, so that the process of how the present invention uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0081] Example 1: Overall Structure of a High-Current Rectifier like Figure 1 As shown, a low-loss high-current rectifier 1 with parallel branch online identification includes an AC input unit 10, a rectifier circuit 20, a DC output unit 30, a controller 40, and a protection unit 90.

[0082] The AC input unit 10 is used to connect the AC power to be rectified; it can be single-phase or three-phase AC power. The rectifier circuit 20 is located between the AC input unit 10 and the DC output unit 30, and is used to convert the AC power to DC power. The rectifier circuit 20 can be a single-phase full-bridge rectifier circuit, a three-phase full-bridge rectifier circuit, or a multi-pulse rectifier circuit.

[0083] This embodiment takes a three-phase full-bridge synchronous rectifier circuit as an example. The three-phase full-bridge synchronous rectifier circuit includes multiple rectifier bridge arms 22, and each rectifier bridge arm 22 includes three controllable rectifier branches 21 connected in parallel.

[0084] Each controllable rectifier branch 21 includes a controllable power semiconductor device 211. In this embodiment, the controllable power semiconductor device 211 is a SiC MOSFET module. The SiC MOSFET module has a first main electrode, a second main electrode, and a control electrode. The first main electrode is connected to the branch input terminal 212, the second main electrode is connected to the branch output terminal 213, and the control electrode is connected to the independent drive unit 80.

[0085] In other embodiments, the controllable power semiconductor device 211 may also be a silicon-based MOSFET, IGBT, or a composite power unit consisting of a rectifier diode and a controllable bypass device.

[0086] Each controllable rectifier branch 21 is equipped with an independent branch current detection unit 50, a branch voltage detection unit 60, a temperature detection unit 70, and an independent drive unit 80.

[0087] The controller 40 includes a data acquisition module 41, a dynamic parameter identification module 42, a target current determination module 43, a conduction window calibration module 44, a current sharing status verification module 45, and an anomaly identification and protection module 46. The data acquisition module 41 is used to synchronously acquire the branch current, device on-state voltage drop, total on-state voltage drop, and branch temperature of each controllable rectifier branch 21; the dynamic parameter identification module 42 is used to calculate the device dynamic conduction parameters, total branch dynamic conduction parameters, and connection dynamic resistance based on two different current operating points within the same rectification conduction cycle; the target current determination module 43 is used to determine the current allocation weight and target current based on the dynamic parameters, branch temperature, and operating status of each branch; the conduction window calibration module 44 is used to adjust the conduction window based on the deviation between the actual current and the target current of each branch; the current sharing status verification module 45 is used to determine whether the current deviation of each branch enters the preset current sharing range; the anomaly identification and protection module 46 is used to identify device performance degradation, connection anomalies, overheating, drive anomalies, and detection anomalies, and to execute warnings, derating, branch disconnection, or system shutdown.

[0088] Example 2: Branch Detection Structure like Figure 2 As shown, a controllable rectifier branch 21 includes a branch input terminal 212, a branch output terminal 213, a controllable power semiconductor device 211, an input connection conductor 214, and an output connection conductor 215.

[0089] Branch input terminal 212 is connected to the first main electrode of controllable power semiconductor device 211 via input connection conductor 214, and the second main electrode of controllable power semiconductor device 211 is connected to branch output terminal 213 via output connection conductor 215. Independent drive unit 80 is connected to the control electrode of controllable power semiconductor device 211.

[0090] The branch current detection unit 50 is connected in series or coupled in the main current path of the branch. This embodiment uses a Hall current sensor to reduce the additional conduction loss generated by the current sampling element.

[0091] The branch voltage detection unit 60 includes a first voltage detection component 61 and a second voltage detection component 62. The first voltage detection component 61 includes a first device voltage sampling terminal 611 and a second device voltage sampling terminal 612, which are respectively located close to the two main electrodes of the controllable power semiconductor device 211 and connected to the controller 40 through independent low-current sampling wires, so that the device conduction voltage drop detection path does not pass through the input connection conductor 214 and the output connection conductor 215 that carry the main current.

[0092] The second voltage detection component 62 includes a first branch voltage sampling terminal 621 and a second branch voltage sampling terminal 622. The first branch voltage sampling terminal 621 is located at the branch input terminal 212, and the second branch voltage sampling terminal 622 is located at the branch output terminal 213. The total forward voltage drop of the branch detected by the second voltage detection component 62 includes the voltage drop across the input connection conductor 214, the controllable power semiconductor device 211, the output connection conductor 215, and the corresponding connection interfaces.

[0093] The temperature detection unit 70 is disposed near the package surface or heat dissipation substrate of the controllable power semiconductor device 211. In other embodiments, temperature detection points may be added near the connection terminals of the input connection conductor 214 or the output connection conductor 215. To improve differential recognition accuracy, the branch current detection unit 50, the first voltage detection component 61, and the second voltage detection component 62 are synchronously sampled by the same sampling clock.

[0094] Example 3: Dual Operating Point Sampling like Figure 3As shown, a rectification conduction cycle includes a conduction start phase, a stable conduction phase, and a conduction end phase. The controller 40 selects a first sampling time t1 and a second sampling time t2 during the stable conduction phase within the same monotonic variation range. The first sampling time t1 corresponds to the first operating point P1, and the second sampling time t2 corresponds to the second operating point P2.

[0095] The branch current at the first operating point P1 is I. i1 The branch current at the second operating point P2 is I. i2 The two satisfy |I i2 -I i1 |≥ΔI min , where ΔI min The preset identification threshold is used. The first sampling time t1 and the second sampling time t2 avoid the power device's turn-on transients and turn-off transients, and the time interval between them is no greater than the preset identification time Δt. max .

[0096] Data acquisition module 41 acquires branch current I at two operating points respectively. i1 I i2 Device on-state voltage drop V i_device1 V i_device2 and the total conduction voltage drop V of the branch circuit i_total1 V i_total2 The dynamic parameter identification module 42 calculates R. i_device = (V i_device2 -V i_device1 ) / (I i2 -I i1 ), R i_total = (V i_total2 -V i_total1 ) / (I i2 -I i1 ) and R i_connection =R i_total -R i_device The controller 40 continuously performs identification for multiple rectifier conduction cycles and applies median filtering and moving average to the multiple consecutive valid identification results.

[0097] As a non-restrictive example of a parameter, let the branch current I of the i-th controllable rectifier branch at the first operating point be... i1 400A, device on-state voltage drop V i_device1 The total on-state voltage drop of the branch is 0.48V. i_total1 The branch current I at the second operating point is 0.68V. i2 600A, device on-state voltage drop V i_device2 The total on-state voltage drop of the branch is 0.60V. i_total2 It is 0.86V. At this time, R i_device=(0.60-0.48) / (600-400)=0.60mΩ, R i_total =(0.86-0.68) / (600-400)=0.90mΩ, R i_connection =0.90 - 0.60 = 0.30 mΩ.

[0098] If the temperature of the connection area is T i 75℃, reference temperature T ref Given a temperature of 25℃, the equivalent temperature coefficient α obtained through calibration is 0.0036 / ℃, then R i_connection_ref =0.30 / [1 + 0.0036 × (75 - 25)] ≈ 0.254 mΩ. If the initial dynamic resistance of this branch at the reference temperature is 0.250 mΩ, then the change in the normalized result relative to the initial value is approximately 1.6%, which indicates that the connection status is still within the normal range. In this example, the preset identification threshold ΔI min It can be set to 50A, with a preset recognition time Δt. max It can be set to 200μs.

[0099] Example 4: Establishment of Active Working Points When the load of the high-current rectifier is stable, and the natural variation of the branch current within a single rectifier conduction cycle cannot meet the identification requirements, the controller 40 actively establishes the first operating point and the second operating point through short-time pulse width modulation or conduction window perturbation.

[0100] The controller 40 first controls multiple parallel controllable rectifier branches 21 to conduct simultaneously according to the normal conduction window, and sets a first conduction sampling sub-interval and a second conduction sampling sub-interval within the allowable conduction interval of the branch to be identified. The controller 40 applies a small pulse width or conduction window adjustment to the branch to be identified between the two sampling sub-intervals, so that the branch current in the two sampling sub-intervals forms a distinguishable difference; both operating points are selected after the controllable power semiconductor device 211 is in the conduction state and the current fluctuation enters a stable range.

[0101] To compensate for short-term changes in the current of the branch to be identified, the controller 40 simultaneously adjusts the conduction window of at least one other normal branch in the opposite direction. The controller does not use the terminal voltage when the device is turned off as the conduction voltage drop, nor does it perform dynamic parameter calculations during turn-on, turn-off, or commutation transients.

[0102] During active disturbance, the anomaly identification and protection module 46 monitors the total output current, DC output voltage, and peak current of each branch in real time. When any parameter exceeds the allowable range for active identification, the active disturbance is immediately terminated, and the conduction window before the disturbance is restored.

[0103] Example 5: Target Current and Conduction Window Calibration like Figure 4 As shown, the three parallel controllable rectifier branches are the first branch, the second branch, and the third branch, and all three have the same nominal conduction window.

[0104] When the actual current of the first branch is detected to be higher than its target current, the actual current of the second branch is lower than its target current, and the dynamic conduction parameters and temperature of the second branch are within the normal range, the controller 40 shortens the actual conduction window of the first branch and correspondingly extends the actual conduction window of the second branch; the actual current of the third branch is within the allowable range, and its actual conduction window basically coincides with the nominal conduction window.

[0105] After the conduction window is adjusted, the controller 40 reacquires the average current of the three branches in subsequent rectification conduction cycles. When the current deviation of the three branches all enters the preset current sharing range, the adjusted conduction window is maintained. When the connection dynamic resistance of the first branch exceeds the warning threshold, the controller 40 reduces its current allocation weight; when the connection dynamic resistance continues to increase and exceeds the fault threshold, the independent drive unit 80 corresponding to the first branch is turned off, and the first branch is disconnected.

[0106] As a non-restrictive example of current sharing calibration, suppose the total current to be handled by rectifier arm 22 is 1500A, and the initial current allocation weight of the three normal branches is 1, then the target current of each branch is 500A. If the average actual currents of the first, second, and third branches before calibration are 560A, 460A, and 480A respectively, the controller 40 can shorten the conduction window of the first branch by 5μs, extend the conduction window of the second branch by 5μs, and keep the conduction window of the third branch unchanged. After several consecutive control cycles, the average currents of the three branches are adjusted to, for example, 505A, 498A, and 497A, all within the preset current sharing range of ±2% of the target current.

[0107] In this example, increasing the normalized connection dynamic resistance by 20% relative to the initial reference value can be set as a warning condition, increasing it by 50% can be set as a branch disconnection condition, and 100°C can be set as the branch temperature protection threshold. The above values ​​are only used to illustrate the parameter calculation and control process of the controller. The actual values ​​can be calibrated experimentally based on the device's rated current, connection structure, heat dissipation conditions, and load characteristics.

[0108] Example 6: Control Method like Figure 5As shown, the online identification and current sharing calibration control method for parallel rectifier branches in this embodiment is executed sequentially from S1 to S8. S1 collects branch current, branch on-state voltage drop, and branch temperature; S2 establishes a first operating point and a second operating point, both within the stable conduction range of the devices; S3 acquires the operating point parameters and determines whether the sampled data is valid. If invalid, it returns to S2; if valid, it proceeds to S4 for dynamic parameter identification; S5 determines the target current; S6 calibrates the conduction window; S7 verifies the current sharing status. If the current deviation does not enter the preset current sharing range, it returns to S5; if it enters the preset current sharing range, it proceeds to S8; S8 performs anomaly identification and protection, and returns to S1 after the end of this control cycle.

[0109] Example 7: Simplified Detection Scheme In applications where the requirements for distinguishing abnormal locations are less stringent, only the second voltage detection component 62 can be set, without the first voltage detection component 61. The controller 40 calculates the total dynamic conduction parameters of the branch based on the dual-operating-point changes in the total on-state voltage drop and the branch current, and uses these parameters as the basis for target current allocation and conduction window calibration. This simplified scheme cannot directly separate the device's dynamic conduction parameters from the connection dynamic resistance, but it can still identify changes in the complete branch conduction state and achieve online current sharing calibration.

[0110] Example 8: Diode Composite Rectification Scheme In another embodiment, each controllable rectifier branch 21 includes a rectifier diode and a bypass MOSFET corresponding to the rectifier diode. When the rectifier diode meets the forward conduction condition, the controller 40 controls the bypass MOSFET to turn on, so that the main rectified current is transmitted through the bypass MOSFET to reduce the rectified voltage drop.

[0111] When the bypass MOSFET needs to shorten its conduction window, a portion of the rectified current is temporarily transferred through the rectifier diode. The controller 40 adjusts the average current of each branch within one rectification conduction cycle by regulating the conduction window of the bypass MOSFETs in different branches. In this embodiment, the rectifier diode can also maintain the basic rectification path when the controller starts up, the drive fails, or the bypass MOSFET is turned off, thereby improving the fault tolerance of the high-current rectifier.

[0112] The above description is merely a preferred embodiment of the present invention. Any equivalent substitutions or transformations made by those skilled in the art regarding the type of controllable power semiconductor device, rectification topology, sampling method, parameter compensation method, conduction window adjustment method, and abnormal threshold setting method without departing from the technical concept of the present invention shall fall within the protection scope of the present invention.

Claims

1. A low-loss, high-current rectifier with online identification of parallel branches, comprising a rectifier circuit and a controller, wherein the rectifier circuit comprises at least two controllable rectifier branches connected in parallel, characterized in that: Each of the controllable rectifier branches is equipped with a controllable power semiconductor device, a branch current detection unit, a branch voltage detection unit, and an independent drive unit; The controller is connected to the branch current detection unit, the branch voltage detection unit and the independent drive unit respectively, and is used to acquire the branch current and branch conduction voltage drop of the same controllable rectifier branch at the first operating point and the second operating point with different current magnitudes within the same rectification conduction cycle, and determine the dynamic conduction parameters of the corresponding controllable rectifier branch according to the ratio of the change in branch conduction voltage drop to the change in branch current between the two operating points. The controller is also used to determine the corresponding target current based on the dynamic conduction parameters of each controllable rectifier branch, and to adjust the conduction window of the corresponding controllable power semiconductor device based on the deviation between the actual current of each controllable rectifier branch and the corresponding target current.

2. The low-loss, high-current rectifier with online identification of parallel branches according to claim 1, characterized in that: The branch voltage detection unit includes a first voltage detection component and a second voltage detection component. The two sampling terminals of the first voltage detection component are respectively connected near the two main electrodes of the controllable power semiconductor device, and are used to detect the on-state voltage drop of the controllable power semiconductor device; The two sampling terminals of the second voltage detection component are respectively connected to the branch input terminal and the branch output terminal of the corresponding controllable rectifier branch, and are used to detect the total on-state voltage drop of the branch, including the controllable power semiconductor device, the branch connecting conductor and the connection interface.

3. The low-loss, high-current rectifier with online identification of parallel branches according to claim 2, characterized in that: The controller is used to determine the dynamic conduction parameters of the controllable power semiconductor device based on the device conduction voltage drop detected by the first voltage detection component at the first operating point and the second operating point. The controller is further configured to determine the total dynamic conduction parameter of the corresponding controllable rectifier branch based on the total conduction voltage drop of the branch detected by the second voltage detection component at the first operating point and the second operating point, and to determine the connection dynamic resistance of the corresponding controllable rectifier branch based on the difference between the total dynamic conduction parameter of the branch and the dynamic conduction parameter of the device.

4. The low-loss, high-current rectifier with online identification of parallel branches according to claim 1, characterized in that: Each of the controllable rectifier branches is also provided with a temperature detection unit, which is used to detect the temperature of the corresponding controllable power semiconductor device or the branch connection conductor. The controller stores the compensation relationship between dynamic conduction parameters and temperature, and is used to perform temperature normalization processing on the dynamic conduction parameters of the corresponding controllable rectifier branch based on the temperature detected by the temperature detection unit.

5. The low-loss, high-current rectifier with online identification of parallel branches according to claim 1, characterized in that: The first and second operating points are formed by the natural variation of the branch current within the same rectification conduction cycle. Alternatively, the controller may apply short-time pulse width modulation or conduction window perturbation to the controllable rectifier branch to be identified within the same rectifier conduction cycle, so that the controllable rectifier branch to be identified forms the first operating point and the second operating point in two conduction sampling sub-intervals with different current magnitudes. The first operating point and the second operating point are both located in the stable sampling range when the controllable power semiconductor device is in the on state, and the absolute value of the change in branch current between the two is not less than the preset identification threshold.

6. The low-loss, high-current rectifier with online identification of parallel branches according to claim 5, characterized in that: When the controller actively forms the first operating point and the second operating point through short-time pulse width modulation or conduction window perturbation, the controller simultaneously performs reverse compensation adjustment on the conduction window of at least one other normal controllable rectifier branch so that the total output current of each controllable rectifier branch is kept within a preset fluctuation range.

7. The low-loss, high-current rectifier with online identification of parallel branches according to any one of claims 1 to 6, characterized in that: Each of the controllable rectifier branches is also provided with a temperature detection unit, which is used to detect the branch temperature of the corresponding controllable power semiconductor device, branch connecting conductor, or connecting area. The controller is used to determine the current allocation weight of the corresponding controllable rectifier branch based on the dynamic conduction parameters, branch temperature and operating status of each controllable rectifier branch, and to determine the corresponding target current based on the current allocation weight. When the dynamic conduction parameter or branch temperature of a controllable rectifier branch exceeds the corresponding threshold, the controller reduces the current distribution weight of the controllable rectifier branch and increases the current distribution weight of at least one controllable rectifier branch that is in a normal state. When the actual current of a controllable rectifier branch is higher than the corresponding target current, the controller delays the start time of the conduction of the controllable rectifier branch, advances its end time of conduction, or shortens its conduction duration. When the actual current of a controllable rectifier branch is lower than the corresponding target current and is in a normal state, the controller advances the start time of conduction of the controllable rectifier branch, delays its end time of conduction, or extends its conduction duration.

8. A control method for a low-loss, high-current rectifier with online identification of parallel branches, applied to a rectifier having at least two parallel controllable rectifier branches, characterized in that, Includes the following steps: S1, Branch Status Acquisition: Acquire the branch current, branch on-state voltage drop, and branch temperature of each controllable rectifier branch; S2. Identification of operating point establishment: Within the same rectification conduction cycle, the controllable rectifier branch to be identified forms a first operating point and a second operating point with different current magnitudes, and both the first operating point and the second operating point are located in the stable sampling range where the controllable power semiconductor device is in the conduction state. S3. Obtaining operating point parameters: Obtain the first branch current and the first branch conduction voltage drop at the first operating point, and the second branch current and the second branch conduction voltage drop at the second operating point. S4. Dynamic parameter identification: Based on the voltage drop change between the first branch conduction voltage drop and the second branch conduction voltage drop, and the current change between the first branch current and the second branch current, determine the dynamic conduction parameters of the controllable rectifier branch to be identified. S5. Target current determination: Determine the target current of each controllable rectifier branch based on the dynamic conduction parameters of each controllable rectifier branch. S6. Conduction window calibration: Based on the deviation between the actual current of each controllable rectifier branch and the corresponding target current, adjust the conduction start time, conduction end time, or conduction duration of the corresponding controllable rectifier branch. S7. Current sharing status verification: In the subsequent rectification conduction cycle, the branch current of each controllable rectifier branch is reacquired. Based on the reacquired branch current, it is determined whether the current deviation between each controllable rectifier branch is within the preset current sharing range. If it is not within the preset current sharing range, steps S5 and S6 are executed again.

9. The control method for a low-loss, high-current rectifier with online identification of parallel branches according to claim 8, characterized in that, Step S3 further includes: The device on-state voltage drop of the controllable power semiconductor device in the controllable rectifier branch to be identified at the first operating point and the second operating point, and the total on-state voltage drop of the controllable rectifier branch to be identified at the first operating point and the second operating point are obtained respectively. Step S4 includes: The dynamic conduction parameters of the device are determined based on the ratio of the change in device on-state voltage drop to the change in branch current between two operating points. The total dynamic conduction parameters of the branch are determined based on the ratio of the change in total conduction voltage drop of the branch to the change in branch current between the two operating points. The connection dynamic resistance of the controllable rectifier branch to be identified is determined based on the difference between the total dynamic conduction parameter of the branch and the dynamic conduction parameter of the device.

10. The control method for a low-loss, high-current rectifier with online identification of parallel branches according to claim 9, characterized in that, It also includes step S8, anomaly identification and protection: Obtain the branch temperature of each controllable rectifier branch; When the dynamic conduction parameter of the device exceeds the abnormal threshold of the device, and the connection dynamic resistance is within the preset normal connection range, the corresponding controllable rectifier branch is determined as the device performance degradation branch. When the connection dynamic resistance exceeds the connection abnormality threshold and the device dynamic conduction parameter is within the preset normal range of the device, the corresponding controllable rectifier branch is determined as the connection abnormality branch. When the branch temperature of the corresponding controllable rectifier branch exceeds the temperature threshold, it is identified as an overheated branch. Depending on the type and severity of the anomaly, at least one protective operation is performed on the corresponding controllable rectifier branch, including reducing the target current, outputting a warning signal, disconnecting the branch, or stopping the rectifier operation.

Images