GB2551596A

GB2551596A - Method and apparatus for fault prediction of sub-module in flexible direct current transmission converter valve

Info

Publication number: GB2551596A
Application number: GB1620739.1A
Authority: GB
Inventors: Ji Panpan; Hu Liang; Zu Lifeng; Feng Min; Yang Qingbo; Zhao Yangyang; Ma Junjie; Li Kun
Original assignee: State Grid Corp of China SGCC; Xuji Group Co Ltd; XJ Electric Co Ltd
Current assignee: State Grid Corp of China SGCC; Xuji Group Co Ltd; XJ Electric Co Ltd
Priority date: 2016-06-13
Filing date: 2016-12-06
Publication date: 2017-12-27
Anticipated expiration: 2036-12-06
Also published as: GB2551596B; CN106026157B; GB201620739D0; CN106026157A

Abstract

Method and apparatus for fault prediction of a sub-module in a flexible direct current transmission converter valve by: determining if sub-module is in a charging or discharging state according to topology of sub-module and direction of collected bridge arm current, in connection with control instruction for the sub-module, the control instruction including switching in, switching out and blocking; determining a voltage rising or reducing rate according to the magnitude of current and capacitance of the sub-module or according to loss and capacitance of the sub-module; predicting a voltage after a pre-set time; comparing the actual voltage measured after the pre-set time and the predicted voltage to determine if a fault occurs. The voltage reducing rate may be determined according to the magnitude of the current and the capacitance of the sub-module if the sub-module is in a fast-discharging state, or according to the loss and capacitance if the sub-module is in a slow-discharging state.

Description

METHOD AND APPARATUS FOR FAULT PREDICTION OF SUB-MODULE IN FLEXIBLE DIRECT CURRENT TRANSMISSION

CONVERTER VALVE

TECHNICAL FIELD

[0001] The disclosure relates to the field of power electronic technology, and in particular to fault prediction of a sub-module in a flexible direct current transmission converter valve.

BACKGROUND

[0002] Flexible direct current transmission is a new generation of high voltage direct current transmission technology employing a voltage source converter. Due to use of fully controlled Insulated Gate Bipolar Translators (IGBTs) as power devices, the capability of turning off current becomes available, the problem that self-commutation cannot be implemented in the conventional thyristor-based direct current transmission is addressed and an additional independent decoupling control function for reactive power and active power is enabled. Therefore, the flexible direct current transmission is especially applicable to application scenarios such as networking of wind-driven generators, solar generators and other new energy-driven generators, asynchronous network interconnection, power supply for passive or isolated load, expansion or reconstruction of urban power distribution networks and the like.

[0003] In recent years, the flexible direct current transmission following a technical route of a Modular Multilevel Converter (MMC) has come into engineering applications in China. A flexible direct current transmission MMC, as a core master device of a flexible direct current transmission converter station, includes multiple sub-modules connected in series, to perform energy conversion between an alternating current system and a direct current system. Fig. lisa schematic diagram illustrating the topology of the main circuit of a three-phase MMC. As shown in Fig. 1, the topology includes six bridge arms, each of which is made up of n Sub-Modules (SM) and a converter reactor connected in series. Under the current technological conditions, the voltage each single sub-module can bear is no more than 2000Y. Generally, there are hundreds of serially connected sub-modules in each bridge arm of a flexible direct current transmission converter valve in engineering. Moreover, the number of the serially connected sub-modules in each bridge arm grows as the level of direct voltage rises.

[0004] In consideration of the huge number of the sub-modules, safe and reliable running of these sub-modules becomes extremely important. Currently, monitoring on running of sub-modules in a flexible direct current transmission converter valve only includes the monitoring on the capacitances of the sub-modules and a post-fault detection and protection for IGBTs. There is no online analysis means for sub-modules in service, and thus it is impossible to implement fault prediction of the sub-modules.

SUMMARY

[0005] The disclosure is intended to provide a method for fault prediction of a sub-module in a converter valve to address the problem existing in the conventional art that fault detection is delayed. The disclosure also provides an apparatus for fault prediction of a sub-module in a converter valve based on the flow of the fault prediction method.

[0006] In view of above, the solutions of the disclosure include: [0007] a method for fault prediction of a sub-module in a flexible direct current transmission converter valve, which includes the following steps: [0008] step 1): determining whether the sub-module is in a charging state or a discharging state according to topology of the sub-module and a direction of a collected bridge arm current, in connection with a control instruction for the sub-module, wherein the control instruction includes: 'switching-in', 'switching-out' and 'blocking'; [0009] step 2): determining a voltage rising rate according to a magnitude of the current and capacitance of the sub-module if the sub-module is in a charging state, or determining a voltage reducing rate according to the magnitude of current and the capacitance of the sub-module or according to loss and the capacitance of the sub-module if the sub-module is in a discharging state; calculating, according to the voltage rising rate or the voltage reducing rate, a predictive voltage for an instant after a preset time; and recording the predictive voltage; and [0010] step 3): comparing an actual voltage at the instant after the preset time with the recorded predictive voltage; and predicting that the sub-module is running normally if the difference between the actual voltage and the predictive voltage is within a preset range, or predicting that a fault occurs in the sub-module if the difference between the actual voltage and the predictive voltage is out of the preset range.

[0011] Further, in step 1), the sub-module may be a half-bridge sub-module; and if the direction of current is 'positive' and the control instruction is 'blocking' or 'switching-in', the sub-module is in the charging state; if the control instruction is 'switching-out', regardless of whether the direction of current is 'positive' or 'negative', the sub-module is in a slow-discharging state; if the control instruction is 'blocking' and the direction of current is 'negative', the sub-module is in the slow-discharging state; or if the control instruction is 'switching-in' and the direction of the current is 'negative', the sub-module is in a fast-discharging state.

[0012] Further, in step 2), the voltage reducing rate may be determined according to the magnitude of the current and the capacitance of the sub-module if the sub-module is in the fast-discharging state; or the voltage reducing rate is determined according to the loss and the capacitance of the sub-module if the sub-module is in the slow-discharging state.

[0013] An apparatus for fault prediction of a sub-module in a flexible direct current transmission converter valve is also provided, which includes a first unit, a second unit and a third unit.

[0014] The first unit is configured to determine whether the sub-module is in a charging state or a discharging state according to topology of the sub-module and a direction of a collected bridge arm current, in connection with a control instruction for the sub-module, wherein the control instruction includes: 'switching-in', 'switching-out' and 'blocking'.

[0015] The second unit is configured to determine a predictive voltage, i.e., determine a voltage rising rate according to a magnitude of the current and capacitance of the sub-module if the sub-module is in a charging state, or determine a voltage reducing rate according to the magnitude of current and the capacitance of the sub-module or according to loss and the capacitance of the sub-module if the sub-module is in a discharging state; calculate, according to the voltage rising rate or the voltage reducing rate, the predictive voltage for an instant after a preset time; and record the predictive voltage.

[0016] The third unit is configured to an actual voltage at the instant after the preset time with the recorded predictive voltage; and predict that the sub-module is running normally if the difference between the actual voltage and the predictive voltage is within a preset range, or predict that a fault occurs in the sub-module if the difference between the actual voltage and the predictive voltage is out of the preset range.

Further, in the first unit, the sub-module may be a half-bridge sub-module; and if the direction of current is 'positive' and the control instruction is 'blocking' or 'switching-in', the sub-module is in the charging state; if the control instruction is 'switching-out', regardless of whether the direction of current is 'positive' or 'negative', the sub-module is in a slow-discharging state; if the control instruction is 'blocking' and the direction of current is 'negative', the sub-module is in the slow-discharging state; or if the control instruction is 'switching-in' and the direction of the current is 'negative', the sub-module is in a fast-discharging state.

[0017] Further, in the second unit, the voltage reducing rate may be determined according to the magnitude of the current and the capacitance of the sub-module if the sub-module is in the fast-discharging state; or the voltage reducing rate is determined according to the loss and the capacitance of the sub-module if the sub-module is in the slow-discharging state.

[0018] The method for fault prediction of a sub-module disclosed herein determines whether or not the running of a sub-module is abnormal by comparing an actual voltage with an expected voltage based on real-time monitoring on the capacitive voltage of the sub-module in combination with a voltage change trend expected by a control instruction ('switching-in', 'switching-out' or 'blocking').

[0019] For example, if the voltage of a sub-module is reduced while a control system commands the sub-module to be 'charged', or if the voltage rising rate of the sub-module is not in a normal range, then it can be predicted that a fault occurs in the sub-module. Likewise, if the voltage of a sub-module rises while a control system commands the sub-module to be 'discharged', or if the voltage reducing rate of the sub-module is out of the normal range, then it can be predicted that an under-voltage fault occurs in the sub-module.

[0020] Compared with the conventional art, the method and the apparatus disclosed herein are capable of predicting abnormal running of a sub-module, so as to take protective measures in advance, preventing the fault of the sub-module from causing more problems, reducing loss and improving reliability of operation of a converter valve and are therefore highly valuable in engineering applications. Except for the capacitive voltage of a sub-module which is an electrical parameter that must be collected during the running of the sub-module, the method disclosed herein requires no additional electrical parameter sampling and communication channel, and thus, fault prediction of the sub-module is merely based on a software algorithm. Therefore, the fault prediction method is easy to realize and is applicable to a sub-module controller or valve controller.

[0021] The command 'charge' or 'discharge' is given in consideration of the control instruction 'switching-in', 'blocking' or 'switching-out' as well as the topology of a sub-module and the direction of the bridge arm current.

BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Fig. 1 is a schematic diagram illustrating a modular multilevel converter for flexible direct current transmission; [0023] Fig. 2 is a diagram illustrating the topology of a sub-module; [0024] Fig. 3 illustrates operation modes of the sub-module; and [0025] Fig. 4 is a flowchart illustrating the implementation of a fault prediction algorithm.

DETAILED DESCRIPTION

[0026] The disclosure is described below in detail with reference to the accompanying drawings.

[0027] Method Embodiment [0028] Each bridge arm of a modular multilevel converter for flexible direct current transmission includes serially connected sub-modules. Generally, current would not be turned on or off in the whole bridge arm. Instead, energy interchange between an alternating current system and a direct current system is realized through 'switching-in' or 'switching-out' of each sub-module. Actually, the sub-modules in a bridge arm are not switched in or switched out simultaneously, and thus, a control system is required to control each sub-module separately.

[0029] Fig. 1 is a schematic diagram illustrating the topology of the main circuit of a three-phase MMC. As shown in Fig. 1, the topology includes six bridge arms, each of which is made up of n Sub-Modules (SM) and a converter reactor connected in series. Each sub-module consists mainly of an IGBT half-bridge serving as a switch element and a direct current energy-storage capacitor which is connected with the IGBT half-bridge in parallel.

[0030] In an embodiment, an electrical topology of the sub-module is shown in Fig. 2. In each sub-module, a half-H bridge is formed by two IGBTs, each with an anti-parallel diode, and a direct current capacitor is connected with the two IGBTs in parallel. The operation states of the sub-module can be switched by changing the on-off states of the two IGBTs (T1 and T2) in the sub-module. Additionally, a bypass switch S is arranged at the alternating current output terminal of each sub-module to bypass a faulted sub-module, thereby isolating a fault.

[0031] In practical application, the upper and lower IGBTs in the sub-module cannot be turned on simultaneously. Normally, the sub-module has three operation states, as shown in Fig. 3.

[0032] (1) blocking: the upper and the lower IGBTs (ΤΙ, T2) are both turned off. In this case, the state of the sub-module depends on the forward conduction of the anti-parallel diodes (Dl, D2). When current flows through the diode Dl, the capacitor C is connected with the diode Dl in series in the bridge arm and charged; and when current flows through the diode D2, the capacitor C is bypassed, as shown in Fig. 3(1) and Fig. 3(4).

[0033] (2) switching-in: the T1 is turned on while the T2 is turned off. In this case, the output voltage of the sub-module is the voltage of the capacitor regardless of the direction of the current, which determines whether the capacitor is being charged or discharged, as shown in Fig. 3(2) and Fig. 3(5).

[0034] (3) switching-out: the T1 is turned off while the T2 is turned on. In this case, current flows through the T2 or the D2, and as the capacitor of the sub-module is always bypassed, the output voltage of the sub-module is 0, as shown in Fig. 3(3) and Fig. 3(6).

[0035] According to the operation states of the sub-module shown in Fig. 3, the states shown in Fig. 3(1) and Fig. 3(2) are defined as a capacitor-being-charged state, the state shown in Fig. 3(5) is defined as a capacitor-being-quickly-discharged state (discharging current is large), the states shown in Fig. 3(3), Fig. 3(4) and Fig. 3 (6) are defined as a capacitor-being-slowly-discharged state in which the capacitor is discharged through loss of the sub-module controller per se.

[0036] The main concept of the disclosure lies in that, since the charging or the discharging of the sub-module is controlled under control instructions issued from the controller, the controller may predict voltage change trend of the sub-module based on the issued control instructions; i.e., if the predictive voltage of the sub-module is significantly different from the actual feedback voltage, it is predicted that a fault occurs in the sub-module. A voltage rising rate or voltage reducing rate of the sub-module depends on the magnitude of the current, the capacitance of the sub-module and the loss of the sub-module.

[0037] As shown in Fig. 4, assuming that the voltage of a sub-module is V0 and the change rate (rising rate or reducing rate) of the voltage is Δ, then the predictive voltage Vt of the sub-module can be calculated according to the following formula: Vt = V0+A-t (charged) or K —V0-A-1 (discharged). The controller compares the actual voltage k'action that is obtained by collecting a capacitive voltage with the predictive voltage Vt and predicts that a fault occurs in the sub-module if \Vt ~Vaction\ is greater than a preset value, or predicts that the sub-module is running normally if otherwise.

[0038] Specifically, a fault prediction algorithm may be implemented as follows: [0039] (1) if the direction of current is 'positive' and the instruction issued from the controller is 'blocking or 'switching-in', as shown in Fig. 3(1) and Fig. 3(2), the sub-module enters a charging state, and it is predicted that the voltage of the sub-module would rise,

Vt -V0+A-t, where the voltage rising rate A is determined according to the magnitude of current and the capacitance of the sub-module.

[0040] (2) if the instruction issued from the controller is 'switching-out', as shown in

Fig. 3(3) and Fig. 3(6), the sub-module enters a bypass state, regardless of whether the direction of current is 'positive' or 'negative', the voltage of the capacitor drops slowly due to the loss of the sub-module per se, and the sub-module is in a slow-discharging state, Vt - V0 -A, t, where the voltage reducing rate Δ i is determined according to the loss and the capacitance of the sub-module.

[0041] (3) if the instruction issued from the controller is 'switching-in' and the direction of current is 'negative', as shown in Fig. 3 (5), the sub-module enters a fast-charging state, and it is predicted that the voltage of the sub-module would rise fast, K=V0-A2-t, where the voltage reducing rate A 2 is determined according to the magnitude of current and the capacitance of the sub-module.

[0042] (4) if the instruction issued from the controller is 'blocking' and the direction of current is 'negative', as shown in Fig. 3(4), the sub-module enters a bypass state, the voltage of the capacitor decreases slowly due to the loss of the sub-module per se, and the sub-module is in a slow-discharging state, Vt =V0 -A{-t, where the voltage reducing rate A i is determined according to the loss and the capacitance of the sub-module.

[0043] During the foregoing process, the voltage rising rate Δ is determined according to the magnitude of the current at present and the capacitance of the sub-module, and the voltage reducing rate Δ1 is determined according to the loss and the capacitance of the sub-module; alternatively, the voltage reducing rate A2 is calculated according to the magnitude of the current at present and the capacitance of the sub-module using a well-known method which is not described here.

[0044] The controller calculates a predictive voltage Vt for an instant after a time period t according to the foregoing formula, records the predictive voltage Vt, compares, after the time period t lapses, the recorded predictive voltage with a sampled actual value Vacti0„, considers the sub-module normal if the difference between the predictive voltage and the sampled actual value Vactjon is within a reasonable range or predicts occurrence of a fault in the sub-module if the difference between the predictive voltage and the sampled actual value Vaction is out of the reasonable range and then sends an alarm signal.

[0045] It is to be noted that the foregoing embodiment is described by taking a half-bridge sub-module as an example. The method disclosed herein is also applicable to a full-bridge sub-module which also has states of switched-out, switched-in and blocked and in which current flows in a positive direction or in a negative direction or even to the topology of another type of sub-module.

[0046] Apparatus Embodiment [0047] An apparatus for fault prediction of a sub-module in a flexible direct current transmission converter valve includes a first unit, a second unit and a third unit.

[0048] The first unit is configured to determine whether the sub-module is in a charging state or a discharging state according to topology of the sub-module and a direction of a collected bridge arm current, in connection with a control instruction for the sub-module, wherein the control instruction includes: 'switching-in', 'switching-out' and 'blocking'.

[0049] The second unit is configured to determine a predictive voltage, i.e., determine a voltage rising rate according to a magnitude of the current and capacitance of the sub-module if the sub-module is in a charging state, or determine a voltage reducing rate according to the magnitude of current and the capacitance of the sub-module or according to loss and the capacitance of the sub-module if the sub-module is in a discharging state; calculate, according to the voltage rising rate or the voltage reducing rate, the predictive voltage for an instant after a preset time; and record the predictive voltage.

[0050] The third unit is configured to an actual voltage at the instant after the preset time with the recorded predictive voltage; and predict that the sub-module is running normally if the difference between the actual voltage and the predictive voltage is within a preset range, or predict that a fault occurs in the sub-module if the difference between the actual voltage and the predictive voltage is out of the preset range.

[0051] The apparatus disclosed herein is actually a solution which is implemented on a computer based on the foregoing method, that is, a software component. The software can be running in a sub-module controller or valve controller. The apparatus disclosed herein, which is actually a software component, can be understood with reference to the foregoing clearly described method and is therefore not described here repeatedly.

[0052] Although certain preferred embodiments of the disclosure have been illustrated herein, these embodiments are not to be construed as limiting the disclosure. It should be appreciated that any solution that can be easily devised by those of ordinary skill in the art by making a little modification on the foregoing embodiments within the concept of the disclosure to be used as a variation, substitute or modification of the solutions disclosed herein to substantially realize the same function or purpose with the disclosure shall fall within the protection scope of the disclosure.

Claims

WHAT IS CLAIMED IS:

1. A method for fault prediction of a sub-module in a flexible direct current transmission converter valve, comprising: step 1): determining whether the sub-module is in a charging state or a discharging state according to topology of the sub-module and a direction of a collected bridge arm current, in connection with a control instruction for the sub-module, wherein the control instruction includes: 'switching-in', 'switching-out' and 'blocking'; step 2): determining a voltage rising rate according to a magnitude of the current and capacitance of the sub-module if the sub-module is in a charging state, or determining a voltage reducing rate according to the magnitude of current and the capacitance of the sub-module or according to loss and the capacitance of the sub-module if the sub-module is in a discharging state; calculating, according to the voltage rising rate or the voltage reducing rate, a predictive voltage for an instant after a preset time; and recording the predictive voltage; and step 3): comparing an actual voltage at the instant after the preset time with the recorded predictive voltage; and predicting that the sub-module is running normally if the difference between the actual voltage and the predictive voltage is within a preset range, or predicting that a fault occurs in the sub-module if the difference between the actual voltage and the predictive voltage is out of the preset range.

2. The method for fault prediction of the sub-module in the flexible direct current transmission converter valve according to claim 1, wherein in step 1), the sub-module is a half-bridge sub-module; and if the direction of current is 'positive' and the control instruction is 'blocking' or 'switching-in', the sub-module is in the charging state; if the control instruction is 'switching-out', regardless of whether the direction of current is 'positive' or 'negative', the sub-module is in a slow-discharging state; if the control instruction is 'blocking' and the direction of current is 'negative', the sub-module is in the slow-discharging state; or if the control instruction is 'switching-in' and the direction of the current is 'negative', the sub-module is in a fast-discharging state.

3. The method for fault prediction of the sub-module in the flexible direct current transmission converter valve according to claim 2, wherein in step 2), the voltage reducing rate is determined according to the magnitude of the current and the capacitance of the sub-module if the sub-module is in the fast-discharging state; or the voltage reducing rate is determined according to the loss and the capacitance of the sub-module if the sub-module is in the slow-discharging state.

4. An apparatus for fault prediction of a sub-module in a flexible direct current transmission converter valve, comprising: a first unit, configured to determine whether the sub-module is in a charging state or a discharging state according to topology of the sub-module and a direction of a collected bridge arm current, in connection with a control instruction for the sub-module, wherein the control instruction includes: 'switching-in', 'switching-out' and 'blocking'; a second unit, configured to determine a predictive voltage, i.e., determine a voltage rising rate according to a magnitude of the current and capacitance of the sub-module if the sub-module is in a charging state, or determine a voltage reducing rate according to the magnitude of current and the capacitance of the sub-module or according to loss and the capacitance of the sub-module if the sub-module is in a discharging state; calculate, according to the voltage rising rate or the voltage reducing rate, the predictive voltage for an instant after a preset time; and record the predictive voltage; and a third unit, configured to compare an actual voltage at the instant after the preset time with the recorded predictive voltage; and predict that the sub-module is running normally if the difference between the actual voltage and the predictive voltage is within a preset range, or predict that a fault occurs in the sub-module if the difference between the actual voltage and the predictive voltage is out of the preset range.

5. The apparatus for fault prediction of the sub-module in the flexible direct current transmission converter valve according to claim 4, wherein in the first unit, the sub-module is a half-bridge sub-module; and if the direction of current is 'positive' and the control instruction is 'blocking' or 'switching-in', the sub-module is in the charging state; if the control instruction is 'switching-out', regardless of whether the direction of current is 'positive' or 'negative', the sub-module is in a slow-discharging state; if the control instruction is 'blocking' and the direction of current is 'negative', the sub-module is in the slow-discharging state; or if the control instruction is 'switching-in' and the direction of the current is 'negative', the sub-module is in a fast-discharging state.

6. The apparatus for fault prediction of the sub-module in the flexible direct current transmission converter valve according to claim 5, wherein in the second unit, the voltage reducing rate is determined according to the magnitude of the current and the capacitance of the sub-module if the sub-module is in the fast-discharging state; or the voltage reducing rate is determined according to the loss and the capacitance of the sub-module if the sub-module is in the slow-discharging state.