JPH0531380B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a failure diagnosis device for a thyristor valve for an AC/DC converter, and particularly to a failure diagnosis device that can reliably and automatically detect a failure of a thyristor in a high voltage thyristor valve. Regarding.

FIG. 1 shows an example of the configuration of an AC/DC converter. In the figure, 1 is a conversion transformer, a, b, and c are AC terminals of the transformer, and 2 is an AC/DC exchanger, which is composed of thyristor valves 21 to 26. 3 is a DC reactor, and 4, 4' are DC busbars, which are connected to the AC/DC converter of the other terminal via a DC power transmission line when the AC/DC converter is used for DC power transmission. The thyristor valve used here is usually constructed by connecting a number of thyristors in series, as shown in FIG. Since a very large number of thyristors are used, there is a probability that some of the thyristors will fail and become short-circuited, reducing the withstand pressure of the thyristor valve as a whole. However, in general, if the number of thyristors has some degree of redundancy and the number of thyristor failures is within an allowable range, operation can be continued with the failed thyristor left alone. However, in order to perform such operation, it is necessary to constantly monitor the number of failed thyristors and to confirm that the number of failed thyristors is within an allowable value. Therefore, a thyristor valve failure diagnosis device becomes indispensable for operating an AC/DC converter.

In FIG. 2, A and K are the anode and cathode terminals of the thyristor valve, respectively. _AL1 ～
_ALM is an anode reactor, S ₁ to S _N are thyristors, C ₁ to C _N and R ₁ to _RN are resistors for voltage division, and capacitors VD ₁ to _VDM are voltage dividers. In this example 8
One module is made up of two thyristors, two anode reactors, a voltage divider, etc., and the modules are connected in series to form a thyristor valve. VD ₁ to _VDM are voltage dividers provided for each module, for example, when a valve damping circuit is divided and provided for each module, or when a resistor for DC voltage division is provided for each module. The neon lamp NL ₁ , the light guide LG ₁ , and the detection device FD ₁ are conventional examples of the thyristor failure detection device.

Under normal conditions, thyristors S ₁ to S ₄ and S ₅ to S ₈ share equal voltage, and voltage dividers VD ₁ and VD ₂ also share equal voltage, so no voltage is applied to neon lamp NL ₁ . . However, for example, thyristor S ₁
When the lamp malfunctions and becomes short-circuited, the voltage balance is disrupted and the neon lamp NL ₁ emits light. The light is sent to the detection device FD ₁ at ground potential via the light guide LG ₁ , and the detection device FD ₁ detects the light to know if the thyristor is out of order.

However, if we consider a case where, for example, thyristors S ₁ and S ₇ have failed, even though the failure has occurred, the voltages of thyristors S ₁ - S ₄ and S ₅ - _{S 8} are balanced and the neon No voltage is applied to lamp NL ₁ , so it is not possible to detect a fault.
Furthermore, since the intensity of light from a neon lamp is not necessarily proportional to the applied voltage, it is difficult to clearly distinguish between a case where one thyristor has failed and a case where two thyristors have failed. Furthermore, with the increase in capacity and pressure resistance of AC/DC converters, the number of thyristor valves used increases, and it has traditionally been the responsibility of staff to detect failures and constantly monitor the number of failures. , is very difficult.

The object of the present invention is to avoid the above-mentioned drawbacks,
To provide a failure diagnosis device for a thyristor valve that can reliably know the failure location and number of failures in a thyristor.

A feature of the present invention is that a voltage detection circuit including light emitting elements connected in antiparallel is provided between the anode and cathode (AK) of the thyristor in order to detect the forward voltage and reverse voltage across the thyristor. If these voltages continuously exist during the non-conducting period, the thyristor is determined to be healthy, otherwise it is determined to be faulty, and information is assigned to each thyristor as to whether the forward voltage and reverse voltage are present (healthy) or absent (faulty). It is stored in the address of the designated storage area or in the bits of the address so that the location of the failure and the number of failures can be determined.

FIG. 3 shows an embodiment of the present invention. The same symbols as in the previous figure indicate the same thing, so I will only explain the different ones. VL ₁ to VL _N are voltage detection circuits between AK that detect the voltage across AK of each thyristor, and LG ₁ to LG. _N is a light guide that sends the voltage signal detected by the A-to-AK voltage detection circuit to the fault diagnosis device FM at ground potential by means of an optical signal. This fault diagnosis device FM is comprised of a computer, such as a microcomputer, in order to automatically monitor and record the presence or absence of forward voltage and/or reverse voltage in each thyristor. The detailed operation of this device will be described later, but first,
FIG. 4 shows a specific example of the A-K voltage detection circuit. FIG. 4 shows an example of the i-th thyristor in FIG. 3 (i=1 to N). Here, PD _i is a photodiode connected in antiparallel and emits an optical signal when a current flows through the diode, that is, when a voltage is applied between AK of thyristor S _i . Here, the photodiodes were connected in antiparallel in order to be able to detect both the forward and reverse voltages between the thyristors AK, but it is also possible to connect only one of them. This optical signal is transmitted to the aforementioned fault diagnosis device via the light guide LG _i .
Guided by FM. R _li is a limiting resistor to limit the current flowing to PD _i , and the voltage detection circuit between AK and VL _i is
In this case, it is composed of a series-connected circuit of R _li and PD _i , but any method can be used as long as it can detect the presence or absence of voltage between AK, and the detection method is not limited to this method.

FIG. 5 shows a circuit diagram of a circuit provided inside the failure diagnosis device for converting optical signals into electrical signals.
V _c is a DC power supply, and PT _i is a phototransistor, which becomes conductive when an optical signal is sent from the light guide LG _i . R _ci is a limiting resistor that limits the current flowing through the phototransistor PT _i . Therefore, when the thyristor _{S i shown} in FIG. 4 is healthy and a forward or reverse voltage is applied during its non-conducting period, the output O i becomes zero because an optical signal is sent from PD _i to make PT _i conductive. On the other hand, if S _i is faulty and no forward or reverse voltage is applied during its non-conducting period, no optical signal is output from PD _i , so PT _i remains non-conducting and the output O _i is “1”. (potential of V _c ).

The fault diagnosis device FM is a device that automatically checks whether the state of each thyristor during the non-conducting period is “1” or “0”. The operating principle of this device will be explained using FIG.

In FIG. 6, 1 is a thyristor valve 21~
Voltage waveform between AK of one phase among 26, 2 is on-pulse, 3 is on-pulse of other phase, 4 and 5 are each forward, reverse voltage detection signal, 6 is output O _i signal, 7 is microcontroller interrupt signal , 8 indicates the operating status of the microcomputer, S indicates processing (fault diagnosis), T
is a diagram showing a waiting period (normal processing). The thyristor becomes conductive due to the on-pulse,
The voltage across AK becomes zero. Normally, the next phase of the on-pulse is applied after an electrical angle of 120 degrees after the on-pulse, so the thyristor that was conducting becomes non-conducting after a certain overlap period. A reverse voltage appears between A and K of the thyristor, which causes the thyristor to become non-conductive, then a forward voltage is applied again, and an on-pulse is applied again 360 degrees in electrical angle from the on-pulse, changing the state from non-conductive to conductive. If the thyristor is healthy, this operation is repeated. When a thyristor fails, it is normally in a short-circuit state, so neither forward voltage nor reverse voltage appears across the AK of this thyristor. The non-conducting period of the thyristor valve can be known from the on-pulses of the phases that lead the thyristor valve by 60°, 120°, and 180° in electrical angle. In Figure 6, an on-pulse P _t (self-phase off-pulse) of the other phase that is 120 degrees ahead of the on-pulse P _p of this thyristor valve interrupts the microcomputer to diagnose the failure of each thyristor that makes up this thyristor valve. It shows the case.

FIG. 7 shows a block diagram of the schematic configuration of the failure diagnosis device FM. IRQ is an interrupt signal, MPU is a microprocessor that forms the core of a microcontroller, and ROM
is a memory circuit that only performs reading, and is a fault diagnostic device.
The contents of commands for processing performed by FM are stored here. RAM is a readable/writable memory circuit that temporarily stores the state of each thyristor shown in Figure 5 above.
It also temporarily stores the calculation results. DI is an input interface circuit that takes in the state I _o of each thyristor (signal based on O _i in Figure 5).
The address signal outputted from the aforementioned microprocessor MPU (transferred via the address bus AB) is decoded by the decoder DEC, and the input gate of the input interface circuit DI is opened at an appropriate time. The state (data) of each thyristor is transferred to the memory circuit RAM mentioned above through the data bus DB.
be taken in. DO is an output interface circuit. Similar to the operation in the input interface circuit DI, the output gate of the output interface circuit DO is opened by another output signal of the decoder DEC mentioned above, and the content of the output signal is sent to the display circuit DIS. indicate.

FIG. 8 shows the software flow of calculation processing of this failure diagnosis device FM. Along with startup 80, the microcomputer performs normal processing 81, such as displaying the position and number of failed thyristors. On-pulse of other phase, that is, electrical angle than on-pulse of own phase.
When a pulse advanced by 60°, 120°, or 180° is output from a pulse generating circuit (not shown), this becomes an interrupt signal IRQ, and the microcomputer enters interrupt processing 82-86. In this case, normal processing is temporarily put on standby. In the interrupt processing, data reading 83 of the _state of each thyristor is carried out, checking and memorizing whether the state is "1" (failure) or "0" (healthy),
If there is a failure, the location and time are displayed 84. In this interrupt processing, it is determined 85 whether or not all the thyristors of the corresponding phase have been inspected, and when the inspection has been completed, the interrupt processing is terminated 86. Thereafter, the process returns to the original normal processing 81 and repeats the normal processing 81 until the next interrupt signal IRQ arrives. The states of the thyristors of the individual thyristor valves are stored in the temporary memory circuit RAM, and each state of the thyristor of _the module _M1 is stored in each bit as shown in FIG. It can be stored in a pair-to-one correspondence, and the number of thyristors in one phase whose status is “1” (failure) (one in the figure) and its position (S ₅ in the figure) can be easily determined. can be known. It is also easy to issue an alarm when the number of failures exceeds the redundancy level. These processes are performed in the normal process 81 as described above.

This fault diagnosis does not need to be performed at high speed, and in order to ensure a reliable fault diagnosis, it is necessary to perform the fault diagnosis several times and make sure that the thyristor status is "1" (failure) in a row.
It is best to judge this thyristor as malfunctioning only when this occurs. This operation can be easily performed using microcomputer software.

Also, in the data shown in Figure 9, thyristor 1
The state of each word is divided into 1 bit and stored, but if the storage capacity is not limited, it is also possible to associate one thyristor with 1 word, and if the storage capacity is limited, it is possible to It is also possible to use a data processing method in which only the position and time of the thyristor are stored, and since these can be easily performed by operating the software of a microcomputer, a detailed explanation will not be given here.

In addition, if a thyristor valve causes commutation failure, the state of each thyristor of the thyristor valve where commutation failure has occurred becomes "1", so in this case, the commutation failure is detected, for example, by a signal from a relay, etc. If this fault diagnosis device is locked, even more reliable operation can be expected.

Furthermore, in the above-mentioned embodiment, the failure diagnosis of each thyristor of the thyristor valve is performed by an interrupt command, and the case is entered into a standby state when the diagnosis is completed. However, in order to improve the reliability of the diagnosis,
It is good to perform repeated diagnosis, and this operation can be easily handled using microcontroller software.

In addition, in the above embodiment, failure diagnosis is performed using electrical angles.
Although the explanation has been given on the case where the rotation is performed once every 360° (one cycle), it is clear that the same effect can be obtained even if the rotation is performed once every several cycles.

As described above, according to the present invention, regardless of the number of thyristors used, it is possible to reliably know the location of failure and the number of failures of thyristors, and this can be done automatically. Furthermore, since both forward voltage and reverse voltage are detected, deterioration of the element can also be detected. Further, according to the present invention, by performing multiple diagnosis, erroneous diagnosis can be prevented and the reliability of the failure diagnosis device can be improved.

[Brief explanation of the drawing]

FIG. 1 is a configuration diagram of an AC/DC converter, FIG. 2 is a conventional example of a thyristor valve failure detection method, and FIG. 3 is a thyristor valve failure diagnosis device according to the present invention.
Figure 4 is an example of the A-to-AK voltage detection circuit shown in Figure 3, Figure 5 is an example of the photo-electrical conversion circuit, Figure 6 is a diagram of the operating principle of the failure diagnosis device, and Figure 7 is failure diagnosis. FIG. 8 is a schematic block diagram of the device, FIG. 8 is a software flowchart of the failure diagnosis device, and FIG. 9 is an example of data of the failure diagnosis device. 1...Conversion transformer, 2...AC/DC converter, 3...
...DC reactor, 4,4'...DC bus, 21
~26...Thyristor valve.

Claims (1)

[Scope of Claims] 1. A device for detecting failure of a thyristor of an AC/DC reversible converter having a thyristor valve in each phase formed by connecting a plurality of thyristor modules in series, each comprising a plurality of thyristors connected in series, Between the anode and cathode of each thyristor,
a voltage detection circuit that includes at least two unidirectional conduction type light emitting elements connected in antiparallel and detects the presence or absence of a forward voltage and a reverse voltage based on the presence or absence of an optical signal from the unidirectional conduction type light emitting elements; During one non-conducting period of the thyristor valve from the generation of the ON pulse of the thyristor valve, the voltage detection circuit repeatedly detects the presence or absence of the forward voltage and reverse voltage, and the state in which there is no forward voltage and reverse voltage continues. 1. A failure diagnosis device for a thyristor valve, comprising a failure diagnosis section that determines a thyristor detected as having failed.