CN111812435A - Method for diagnosing fault of static frequency converter based on BP neural network - Google Patents

Method for diagnosing fault of static frequency converter based on BP neural network Download PDF

Info

Publication number
CN111812435A
CN111812435A CN202010595473.6A CN202010595473A CN111812435A CN 111812435 A CN111812435 A CN 111812435A CN 202010595473 A CN202010595473 A CN 202010595473A CN 111812435 A CN111812435 A CN 111812435A
Authority
CN
China
Prior art keywords
failure
full
bridge
thyristor
radical
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202010595473.6A
Other languages
Chinese (zh)
Inventor
陈月营
张法
石祥建
王志
晏飞
肖海波
刘腾
张博
刘彦勇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huilong Branch Of State Grid Xin Yuan Co ltd
NR Engineering Co Ltd
Original Assignee
Huilong Branch Of State Grid Xin Yuan Co ltd
NR Engineering Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Huilong Branch Of State Grid Xin Yuan Co ltd, NR Engineering Co Ltd filed Critical Huilong Branch Of State Grid Xin Yuan Co ltd
Priority to CN202010595473.6A priority Critical patent/CN111812435A/en
Publication of CN111812435A publication Critical patent/CN111812435A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R19/00—Arrangements for measuring currents or voltages or for indicating presence or sign thereof
    • G01R19/04—Measuring peak values or amplitude or envelope of ac or of pulses
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/26—Testing of individual semiconductor devices
    • G01R31/2607—Circuits therefor
    • G01R31/263—Circuits therefor for testing thyristors
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/26—Testing of individual semiconductor devices
    • G01R31/27—Testing of devices without physical removal from the circuit of which they form part, e.g. compensating for effects surrounding elements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Power Engineering (AREA)
  • Rectifiers (AREA)

Abstract

The invention discloses a judgment method for diagnosing fault types of a static frequency converter based on a BP neural network technology. The invention classifies the failure types of the thyristors in detail. Based on the general structure of the static frequency converter, the invention provides the diagnosis steps when the static frequency converter has faults, establishes the BP neural network, diagnoses the thyristors in the bridge and the bridge of the static frequency converter, and finds out the failed thyristor. The invention provides a method for conveniently diagnosing the thyristor fault of a static frequency converter, which can effectively save manpower and material resources and maintain the safety and stability of a unit taking the static frequency converter as a core.

Description

Method for diagnosing fault of static frequency converter based on BP neural network
Technical Field
The invention belongs to the field of power electronic device fault diagnosis, and provides a method for diagnosing a failed thyristor by utilizing a BP neural network technology according to output voltages of a fully-controlled rectifier bridge and a fully-controlled inverter bridge when the thyristor of a static frequency converter fails.
Background
With the increasing capacity of the power grid, the pumped storage power station is applied more and more, and the pumped storage power station is used for peak load regulation and valley filling and accident standby. The pumped storage power station is the same as a common water turbine generator set in a power generation state, when the pumped storage power station is in a pumped state, the generator set works in a motor state, the generator set is started quickly and stably, impact on a power grid is reduced, and a set of common static frequency conversion device is arranged in the power station, so that the optimal treatment method in the technical and economic aspects is achieved.
The schematic diagram of the static frequency converter is shown in fig. 1, the static frequency converter converts the power frequency alternating current into the variable frequency alternating current with continuously adjustable frequency by using the thyristor converter, outputs the variable frequency current to the stator winding of the synchronous motor to form a stator rotating magnetic field, applies the exciting current to the rotor to form a rotor magnetic field, and the rotating stator magnetic field and the rotor magnetic field interact to pull the rotor to rotate, so that the starting of the unit can be realized.
Whether the full-control rectifier bridge and the full-control inverter bridge in the static frequency converter can realize current rectification and inversion is a key factor of whether the static frequency converter can work normally. Damage or failure of any thyristor on the quiescent frequency converter can result in failure of the quiescent frequency converter. The thyristor belongs to a device which is easy to damage, so that whether the thyristor which can find out faults in time is the guarantee of the working efficiency of the static frequency converter.
Disclosure of Invention
The invention aims to provide a method for diagnosing faults of a static frequency converter based on a BP neural network, which is used for quickly detecting the fault state of the static frequency converter in real time, diagnosing a failed thyristor for replacement and repair, stabilizing the working efficiency of the static frequency converter and improving the safety guarantee.
In order to achieve the above object, the present invention provides a method for diagnosing a fault of a static frequency converter based on a BP neural network, wherein the static frequency converter comprises a fully controlled rectifier bridge, a smoothing reactor and a fully controlled inverter bridge, an ac side of the fully controlled rectifier bridge is used as a three-phase input end of the static frequency converter, one end of a dc side of the fully controlled rectifier bridge is connected with one end of a dc side of the fully controlled inverter bridge through the smoothing reactor, and the other end of the dc side of the fully controlled rectifier bridge is directly connected with the other end of the dc side of the fully controlled inverter bridge; the alternating current side of the full-control inverter bridge is connected with a motor; the method comprises the following steps:
step 1: taking all faults which are possible to be used as network training samples under four conditions that only one thyristor is damaged in the full-control rectifier bridge, the upper thyristor and the lower thyristor of the same bridge arm of the full-control rectifier bridge are damaged, the two thyristors in the same half bridge of the full-control rectifier bridge are damaged and the two crossed thyristors of different half bridges of the full-control rectifier bridge are damaged when the trigger angle of the full-control rectifier bridge is 0, and taking a direct-current component X of the potential difference between the cathode and the anode of the full-control rectifier bridge1Amplitude of fundamental wave X2Second harmonic amplitude X3Third harmonic amplitude X4As an input value, training a BP neural network of the fully-controlled rectifier bridge; obtaining a corresponding relation between the output value and the full-control rectifier bridge fault thyristor;
step 2: taking all faults which are possibly caused by four conditions that only one thyristor is damaged in the full-control inverter bridge, the upper thyristor and the lower thyristor of the same bridge arm of the full-control inverter bridge are damaged, the two thyristors in the same half bridge of the full-control inverter bridge are damaged and the two crossed thyristors in different half bridges of the full-control inverter bridge are damaged when the trigger angle of the full-control inverter bridge is 0 as a network training sample, and taking the input voltage amplitude U of the full-control inverter bridge1Phase A output voltage amplitude U2B phase output voltage amplitude U3C phase output voltage amplitude U4As an input value, training a full-control inverter bridge BP neural network; obtaining a corresponding relation between the output value and the full-control inverter bridge fault thyristor;
and step 3: detecting the output voltages of a fully controlled rectifier bridge and a fully controlled inverter bridge of the static frequency converter in real time, entering a step 4 when detecting that the output voltage of the fully controlled rectifier bridge is abnormal, and entering a step 5 when detecting that the output voltage of the fully controlled inverter bridge is abnormal;
and 4, step 4: extracting direct current component X of potential difference of cathode and anode of full-control rectifier bridge1Amplitude of fundamental wave X2Second harmonic amplitude X3Third harmonic amplitude X4Inputting the full-control rectifier bridge BP neural network to obtain corresponding output, and positioning the failed thyristor according to the corresponding relation between the output value and the failed thyristor of the full-control rectifier bridge;
and 5: calculating input voltage amplitude U of full-control inverter bridge1Phase A output voltage amplitude U2B phase output voltage amplitude U3C phase output voltage amplitude U4And inputting the full-control inverter bridge BP neural network to obtain corresponding output, and positioning the failed thyristor according to the corresponding relation between the output value and the failed thyristor of the full-control inverter bridge.
Furthermore, the fully-controlled rectifier bridge has six thyristors VT1、VT2、VT3、VT4、VT5、VT6Wherein the thyristor VT1、VT4Are positioned on the same bridge arm; the thyristor VT3、VT6Are positioned on the same bridge arm; the thyristor VT5、VT2Are positioned on the same bridge arm;
the step 1 specifically comprises:
the failure types of the fully-controlled rectifier bridge thyristor are as follows:
the fault type is one, only one thyristor in the fully-controlled rectifier bridge is damaged;
the second fault type is that the upper thyristor and the lower thyristor of the same bridge arm of the fully-controlled rectifier bridge are damaged;
the fault type is three, wherein two thyristors in the same half bridge of the full-control rectifier bridge are damaged;
the fault type is four, two crossed thyristors of different half-bridges of the full-control rectifier bridge are damaged;
the full-control rectifier bridge BP neural network is divided into 3 layers, an input layer, a hidden layer and an output layer;
the number of input nodes of the BP neural network of the fully-controlled rectifier bridge is 4, and the direct-current component X of the potential difference between the cathode and the anode of the fully-controlled rectifier bridge is taken1Amplitude of fundamental wave X2Second harmonic amplitude X3Third harmonic amplitude X4As an input value;
taking 15 nodes of the hidden layer of the BP neural network of the fully-controlled rectifier bridge;
the number of output nodes of the BP neural network of the fully-controlled rectifier bridge is 6, and the output nodes are Y respectively1、Y2、Y3、Y4、Y5、Y6(ii) a Each node output value is 0 or 1; y is1Y2Y3When 001, the fault type is one; y is1Y2Y3When 010 is the number, the fault type II is shown; y is1Y2Y3When the value is 011, the fault type is represented as three; y is1Y2Y3When the number is 100, the fault type is four;
when Y is1Y2Y3At 001, there are 6 fault conditions for fault type one: y is4Y5Y6When 001, it represents VT1A failure; y is4Y5Y6When 010 indicates VT2A failure; y is4Y5Y6At 011, it represents VT3A failure; y is4Y5Y6When it is 100, it represents VT4A failure; y is4Y5Y6When 101, it represents VT5A failure; y is4Y5Y6110, denotes VT6A failure;
when Y is1Y2Y3At 010, there are 3 fault conditions for the second fault type: y is4Y5Y6When 001, it represents VT1、VT4A failure; y is4Y5Y6When 010 indicates VT2、VT5A failure; y is4Y5Y6At 011, it represents VT3、VT6A failure;
when Y is1Y2Y3When the fault type is 011, there are 6 fault conditions in the third fault type: y is4Y5Y6When 001, it represents VT1、VT3A failure; y is4Y5Y6When 010 indicates VT2、VT4A failure; y is4Y5Y6At 011, it represents VT3、VT5A failure; y is4Y5Y6When it is 100, it represents VT4、VT6A failure; y is4Y5Y6When 101, it represents VT1、VT5A failure; y is4Y5Y6110, denotes VT2、VT6A failure;
when Y is1Y2Y3At 100, there are 6 fault cases for the four fault types: y is4Y5Y6When 001, it represents VT1、VT2A failure; y is4Y5Y6When 010 indicates VT2、VT3A failure; y is4Y5Y6At 011, it represents VT3、VT4A failure; y is4Y5Y6When it is 100, it represents VT4、VT5A failure; y is4Y5Y6When 101, it represents VT5、VT6A failure; y is4Y5Y6110, denotes VT1、VT6And (4) failure.
Further, the normalization processing principle of the output value of the BP neural network diagnosis method of the fully-controlled rectifier bridge is Y1、Y2、Y3When the output value of (b) is greater than or equal to 0.5, 1 is assumed, and when less than 0.5, 0 is assumed.
Furthermore, the full-control inverter bridge is provided with six thyristors VT7、VT8、VT9、VT10、VT11、VT12Wherein the thyristor VT7、VT10Are positioned on the same bridge arm; the thyristor VT9、VT12Are positioned on the same bridge arm; the thyristor VT11、VT8Are positioned on the same bridge arm; the step 2 specifically comprises:
the failure type of the fully-controlled inverter bridge thyristor is as follows:
the fault type is one, only one thyristor in the full-control inverter bridge is damaged;
the second fault type is that the upper thyristor and the lower thyristor of the same bridge arm of the full-control inverter bridge are damaged;
the fault type is three, wherein two thyristors in the same half bridge of the full-control inverter bridge are damaged;
and the fault type is four, and two crossed thyristors of different half-bridges of the full-control inverter bridge are damaged.
The full-control inverter bridge BP neural network is divided into 3 layers, an input layer, a hidden layer and an output layer;
the number of input nodes of the BP neural network of the fully-controlled inverter bridge is 4, and the input voltage amplitude U of the fully-controlled inverter bridge is taken1Phase A output voltage amplitude U2B phase output voltage amplitude U3C phase output voltage amplitude U4As an input value;
taking 15 nodes of a hidden layer of the full-control inverter bridge BP neural network;
the number of output nodes of the BP neural network of the full-control inverter bridge is 6, and the output nodes are N respectively1、N2、N3、N4、N5、N6(ii) a Each node output value is 0 or 1; n is a radical of1N2N3When 001, the fault type is one; n is a radical of1N2N3When 010 is the number, the fault type II is shown; n is a radical of1N2N3When the value is 011, the fault type is represented as three; n is a radical of1N2N3When the number is 100, the fault type is four;
when N is present1N2N3At 001, there are 6 fault conditions for fault type one: n is a radical of4N5N6When 001, it represents VT7A failure; n is a radical of4N5N6When 010 indicates VT8A failure; n is a radical of4N5N6At 011, it represents VT9A failure; n is a radical of4N5N6When it is 100, it represents VT10A failure; n is a radical of4N5N6When 101, it represents VT11A failure; n is a radical of4N5N6110, denotes VT12A failure;
when N is present1N2N3At 010, there are 3 fault conditions for the second fault type: n is a radical of4N5N6When 001, it represents VT7、VT10A failure; n is a radical of4N5N6When 010 indicates VT8、VT11A failure; n is a radical of4N5N6At 011, it represents VT9、VT12A failure;
when N is present1N2N3When the fault type is 011, there are 6 fault conditions in the third fault type: n is a radical of4N5N6When 001, it represents VT7、VT9A failure; n is a radical of4N5N6When 010 indicates VT8、VT10A failure; n is a radical of4N5N6At 011, it represents VT9、VT11A failure; n is a radical of4N5N6When it is 100, it represents VT10、VT12A failure; n is a radical of4N5N6When 101, it represents VT7、VT11A failure; n is a radical of4N5N6110, denotes VT8、VT12A failure;
when N is present1N2N3At 100, there are 6 fault cases for the four fault types. N is a radical of4N5N6When 001, it represents VT7、VT8A failure; n is a radical of4N5N6When 010 indicates VT8、VT9A failure; n is a radical of4N5N6At 011, it represents VT9、VT10A failure; n is a radical of4N5N6When it is 100, it represents VT10、VT11A failure; n is a radical of4N5N6When 101, it represents VT11、VT12A failure; n is a radical of4N5N6110, denotes VT7、VT12And (4) failure.
Further, the normalization processing principle of the output value of the BP neural network diagnosis method of the fully-controlled inverter bridge is N4、N5、N6When the output value of (b) is greater than or equal to 0.5, 1 is assumed, and when less than 0.5, 0 is assumed.
The invention has the beneficial effects that: the invention classifies the failure types of the thyristors in detail, gives the diagnosis steps when the static frequency converter fails based on the general structure of the static frequency converter, establishes a BP neural network, diagnoses the thyristors in a bridge and a bridge of the static frequency converter, and finds out the failed thyristors. The invention provides a method for conveniently diagnosing the thyristor fault of a static frequency converter, which can effectively save manpower and material resources and maintain the safety and stability of a unit taking the static frequency converter as a core.
Drawings
Fig. 1 is a schematic diagram of a classical structure of a static frequency converter in the prior art.
Fig. 2 is a flow chart for diagnosing the fault type of the static frequency converter based on the BP neural network and positioning the fault.
Fig. 3 is a corresponding diagram of the relationship obtained after network training is performed on all the faults of the fully-controlled rectifier bridge by using the BP neural network.
Fig. 4 is a relationship corresponding graph obtained after network training is performed on all faults of the fully-controlled inverter bridge by using the BP neural network.
Detailed Description
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Fig. 2 shows an embodiment of a method for diagnosing a fault of a stationary frequency converter based on a BP neural network, in which the stationary frequency converter includes a fully controlled rectifier bridge, a smoothing reactor and a fully controlled inverter bridge, an ac side of the fully controlled rectifier bridge is used as a three-phase input end of the stationary frequency converter, one end of a dc side of the fully controlled rectifier bridge is connected to one end of a dc side of the fully controlled inverter bridge through the smoothing reactor, and the other end of the dc side of the fully controlled rectifier bridge is directly connected to the other end of the dc side of the fully controlled inverter bridge; the AC side of the full-control inverter bridge is connected with the motor. The method comprises the following steps:
step 1: taking all faults which are possible to be used as network training samples under four conditions that only one thyristor is damaged in the full-control rectifier bridge, the upper thyristor and the lower thyristor of the same bridge arm of the full-control rectifier bridge are damaged, the two thyristors in the same half bridge of the full-control rectifier bridge are damaged and the two crossed thyristors of different half bridges of the full-control rectifier bridge are damaged when the trigger angle of the full-control rectifier bridge is 0, and taking a direct-current component X of the potential difference between the cathode and the anode of the full-control rectifier bridge1Amplitude of fundamental wave X2Second harmonic amplitude X3Third harmonic amplitude X4As an input value, training a BP neural network of the fully-controlled rectifier bridge; and obtaining the corresponding relation between the output value and the full-control rectifier bridge fault thyristor.
Step 2: taking all faults which are possibly caused by four conditions that only one thyristor is damaged in the full-control inverter bridge, the upper thyristor and the lower thyristor of the same bridge arm of the full-control inverter bridge are damaged, the two thyristors in the same half bridge of the full-control inverter bridge are damaged and the two crossed thyristors in different half bridges of the full-control inverter bridge are damaged when the trigger angle of the full-control inverter bridge is 0 as a network training sample, and taking the input voltage amplitude U of the full-control inverter bridge1Phase A output voltage amplitude U2B phase output voltage amplitude U3C phase output voltage amplitude U4As an input value, training a full-control inverter bridge BP neural network; and obtaining the corresponding relation between the output value and the full-control inverter bridge fault thyristor.
And step 3: detecting the output voltages of a fully controlled rectifier bridge and a fully controlled inverter bridge of the static frequency converter in real time, entering a step 4 when detecting that the output voltage of the fully controlled rectifier bridge is abnormal, and entering a step 5 when detecting that the output voltage of the fully controlled inverter bridge is abnormal;
and 4, step 4: extracting direct current component X of potential difference of cathode and anode of full-control rectifier bridge1Amplitude of fundamental wave X2Second harmonic amplitude X3Third harmonic amplitude X4Inputting the full-controlled rectifier bridge BP neural network to obtain corresponding output, and outputting the output value and the full-controlled rectifier bridgeThe fault thyristor is positioned according to the corresponding relation of the barrier thyristor;
and 5: calculating input voltage amplitude U of full-control inverter bridge1Phase A output voltage amplitude U2B phase output voltage amplitude U3C phase output voltage amplitude U4And inputting the full-control inverter bridge BP neural network to obtain corresponding output, and positioning the failed thyristor according to the corresponding relation between the output value and the failed thyristor of the full-control inverter bridge.
Fig. 1 shows a schematic diagram of a classical structure of a static frequency converter in the prior art. Full-controlled rectifier bridge has six thyristors VT1、VT2、VT3、VT4、VT5、VT6Wherein the thyristor VT1、VT4Are positioned on the same bridge arm; the thyristor VT3、VT6Are positioned on the same bridge arm; the thyristor VT5、VT2Are positioned on the same bridge arm. Full-controlled inverter bridge has six thyristors VT7、VT8、VT9、VT10、VT11、VT12Wherein the thyristor VT7、VT10Are positioned on the same bridge arm; the thyristor VT9、VT12Are positioned on the same bridge arm; the thyristor VT11、VT8Are positioned on the same bridge arm. The structure of fig. 1 below illustrates the method embodiment 2 of the present invention. In this embodiment, the BP neural network is divided into three layers, an input layer, a hidden layer, and an output layer. The number of input nodes for diagnosing the fully-controlled rectifier bridge and the fully-controlled inversion of the static frequency converter is 4, the number of hidden layers is 15, and the number of output nodes is 6.
The principle of the full-control rectifier bridge fault shows that the thyristor works in six conditions:
-case one, in which the number of failed thyristors is zero and the stationary frequency converter is in a normal operating state, at which time no BP neural network is needed to participate in the diagnosis;
case two, in which case the three thyristors of the upper or lower half-bridge of the fully controlled rectifier bridge are all broken, i.e. VT1、VT3、VT5Or VT2、VT4、VT6At the same time, the utility model is damaged,the load output voltage waveform is a zero line. Under the condition, only two damage conditions exist, the upper thyristor and the lower thyristor of the same bridge arm can be respectively replaced to diagnose the fault reason, and a BP neural network is not needed to participate in diagnosis;
case three, in which case only one thyristor of the fully controlled rectifier bridge is damaged. In this case, there may be (1) VT with thyristor damage1Damage; (2) VT2Damage; (3) VT3Damage; (4) VT4Damage; (5) VT5Damage; (6) VT6And (4) damage.
Case four, in which case the upper and lower thyristors of the same leg of the fully controlled rectifier bridge are damaged. In this case, there may be (1) VT with thyristor damage1、VT4Damage; (2) VT2、VT5Damage; (3) VT3、VT6And (4) damage.
Case five, in which case two of the thyristors of the same half bridge of the fully controlled rectifier bridge are damaged. In this case, there may be (1) VT with thyristor damage1、VT3Damage; (2) VT2、VT4Damage; (3) VT3、VT5Damage; (4) VT4、VT6Damage; (5) VT1、VT5Damage; (6) VT2、VT6And (4) damage.
-case six, in which case the two crossed thyristors of different half-bridges of the fully controlled rectifier bridge are damaged. In this case, there may be (1) VT with thyristor damage1、VT2Damage; (2) VT2、VT3Damage; (3) VT3、VT4Damage; (4) VT4、VT5Damage; (5) VT5、VT6Damage; (6) VT6、VT1And (4) damage.
The three conditions to the six conditions are respectively a fault type I, a fault type II, a fault type III and a fault type IV of the fully-controlled rectifier bridge in the application.
Taking all faults possibly occurring when the trigger angle of the fully-controlled rectifier bridge is 0 as network training samples, and taking potential difference between the cathode and the anode of the fully-controlled rectifier bridgeDirect current component X1Amplitude of fundamental wave X2Second harmonic amplitude X3Third harmonic amplitude X4And as an input value, training the BP neural network of the fully-controlled rectifier bridge. The output value Y of the full-control rectifier bridge BP neural network diagnosis method1、Y2、Y3、Y4、Y5、Y6Is 0 or 1. The normalization processing principle of the output value of the BP neural network diagnosis method is Y1、Y2、Y3、Y4、Y5、Y6When the output value of (b) is greater than or equal to 0.5, 1 is assumed, and when less than 0.5, 0 is assumed. The data corresponding to the trained samples are shown in fig. 3:
Y1Y2Y3when 001, the fault type is one; y is1Y2Y3When 010 is the number, the fault type II is shown; y is1Y2Y3When the value is 011, the fault type is represented as three; y is1Y2Y3When the number is 100, the fault type is four;
when Y is1Y2Y3At 001, there are 6 fault conditions for fault type one: y is4Y5Y6When 001, it represents VT1A failure; y is4Y5Y6When 010 indicates VT2A failure; y is4Y5Y6At 011, it represents VT3A failure; y is4Y5Y6When it is 100, it represents VT4A failure; y is4Y5Y6When 101, it represents VT5A failure; y is4Y5Y6110, denotes VT6A failure;
when Y is1Y2Y3At 010, there are 3 fault conditions for the second fault type: y is4Y5Y6When 001, it represents VT1、VT4A failure; y is4Y5Y6When 010 indicates VT2、VT5A failure; y is4Y5Y6At 011, it represents VT3、VT6A failure;
when Y is1Y2Y3When the fault type is 011, there are 6 fault conditions in the third fault type: y is4Y5Y6When 001, it represents VT1、VT3A failure; y is4Y5Y6When 010 indicates VT2、VT4A failure; y is4Y5Y6At 011, it represents VT3、VT5A failure; y is4Y5Y6When it is 100, it represents VT4、VT6A failure; y is4Y5Y6When 101, it represents VT1、VT5A failure; y is4Y5Y6110, denotes VT2、VT6A failure;
when Y is1Y2Y3At 100, there are 6 fault cases for the four fault types: y is4Y5Y6When 001, it represents VT1、VT2A failure; y is4Y5Y6When 010 indicates VT2、VT3A failure; y is4Y5Y6At 011, it represents VT3、VT4A failure; y is4Y5Y6When it is 100, it represents VT4、VT5A failure; y is4Y5Y6When 101, it represents VT5、VT6A failure; y is4Y5Y6110, denotes VT1、VT6And (4) failure.
The principle of the full-control inverter bridge fault shows that the thyristor works under six conditions:
-case one, in which the number of failed thyristors is zero and the stationary frequency converter is in a normal operating state, at which time no BP neural network is needed to participate in the diagnosis;
case two, in which case the three thyristors of the upper or lower half-bridge of the fully-controlled inverter bridge are all broken, i.e. VT8、VT12、VT10Or VT11、VT9、VT7And meanwhile, the voltage waveforms of the phase A, the phase B and the phase C are a zero straight line when the voltage is damaged. Under the condition, only two damage conditions exist, and the upper thyristor and the lower thyristor of the same bridge arm are respectively replacedThe fault reason can be diagnosed, and the BP neural network is not needed to participate in diagnosis;
case three, in which case only one thyristor in the fully controlled inverter bridge is damaged. In this case, there may be (1) VT with thyristor damage7Damage; (2) VT8Damage; (3) VT9Damage; (4) VT10Damage; (5) VT11Damage; (6) VT12And (4) damage.
Case four, in which case the upper and lower thyristors of the same arm of the fully controlled inverter bridge are damaged. In this case, there may be (1) VT with thyristor damage7、VT10Damage; (2) VT8、VT11Damage; (3) VT9、VT12And (4) damage.
Case five, in which case two of the thyristors of the same half bridge of the fully controlled inverter bridge are damaged. In this case, there may be (1) VT with thyristor damage7、VT9Damage; (2) VT8、VT10Damage; (3) VT9、VT11Damage; (4) VT10、VT11Damage; (5) VT7、VT11Damage; (6) VT8、VT12And (4) damage.
-case six, in which case the two crossed thyristors of different half-bridges of the fully controlled inverter bridge are damaged. In this case, there may be (1) VT with thyristor damage7、VT8Damage; (2) VT8、VT9Damage; (3) VT9、VT10Damage; (4) VT10、VT11Damage; (5) VT11、VT12Damage; (6) VT12、VT7And (4) damage.
The three to six conditions respectively correspond to a fault type one, a fault type two, a fault type three and a fault type four of the fully-controlled inverter bridge in the application.
Taking all faults possibly occurring when the trigger angle of the fully-controlled inverter bridge is 0 as network training samples, and taking the input voltage amplitude U of the fully-controlled inverter bridge1Phase A output voltage amplitude U2B phase output voltage amplitude U3C phase output voltage amplitude U4As input values, the BP neural network is trained. The BP neural network diagnosis method output value N1、N2、N3、N4、N5、N6Is 0 or 1. The normalization processing principle of the output value of the BP neural network diagnosis method is N1、N2、N3、N4、N5、N6When the output value of (b) is greater than or equal to 0.5, 1 is assumed, and when less than 0.5, 0 is assumed. The data corresponding to the trained samples are shown in fig. 4:
N1N2N3when 001, the fault type is one; n is a radical of1N2N3When 010 is the number, the fault type II is shown; n is a radical of1N2N3When the value is 011, the fault type is represented as three; n is a radical of1N2N3When the number is 100, the fault type is four;
when N is present1N2N3At 001, there are 6 fault conditions for fault type one: n is a radical of4N5N6When 001, it represents VT7A failure; n is a radical of4N5N6When 010 indicates VT8A failure; n is a radical of4N5N6At 011, it represents VT9A failure; n is a radical of4N5N6When it is 100, it represents VT10A failure; n is a radical of4N5N6When 101, it represents VT11A failure; n is a radical of4N5N6110, denotes VT12A failure;
when N is present1N2N3At 010, there are 3 fault conditions for the second fault type: n is a radical of4N5N6When 001, it represents VT7、VT10A failure; n is a radical of4N5N6When 010 indicates VT8、VT11A failure; n is a radical of4N5N6At 011, it represents VT9、VT12A failure;
when N is present1N2N3When the fault type is 011, there are 6 fault conditions in the third fault type: n is a radical of4N5N6When 001, it represents VT7、VT9Fault of；N4N5N6When 010 indicates VT8、VT10A failure; n is a radical of4N5N6At 011, it represents VT9、VT11A failure; n is a radical of4N5N6When it is 100, it represents VT10、VT12A failure; n is a radical of4N5N6When 101, it represents VT7、VT11A failure; n is a radical of4N5N6110, denotes VT8、VT12A failure;
when N is present1N2N3At 100, there are 6 fault cases for the four fault types. N is a radical of4N5N6When 001, it represents VT7、VT8A failure; n is a radical of4N5N6When 010 indicates VT8、VT9A failure; n is a radical of4N5N6At 011, it represents VT9、VT10A failure; n is a radical of4N5N6When it is 100, it represents VT10、VT11A failure; n is a radical of4N5N6When 101, it represents VT11、VT12A failure; n is a radical of4N5N6110, denotes VT7、VT12And (4) failure.
When the static frequency converter is in a normal working state, and the direct current component X of the potential difference between the cathode and the anode of the full-control rectifier bridge1Amplitude of fundamental wave X2Second harmonic amplitude X3Third harmonic amplitude X4Input voltage amplitude U of full-controlled inverter bridge1Phase A output voltage amplitude U2B phase output voltage amplitude U3C phase output voltage amplitude U4In a state of being detected in real time. Detecting the output voltage waveform abnormality of the full-control rectifier bridge at a certain moment, diagnosing by using a BP neural network, and inputting X1、X2、X3、X4To obtain Y1Y2Y3Y4Y5Y6Is 010010, knowing VT2、VT5And (4) failure.
When the static frequency converter is in normal operationState and fully controlling the DC component X of the potential difference between the cathode and the anode of the rectifier bridge1Amplitude of fundamental wave X2Second harmonic amplitude X3Third harmonic amplitude X4Input voltage amplitude U of full-controlled inverter bridge1Phase A output voltage amplitude U2B phase output voltage amplitude U3C phase output voltage amplitude U4In a state of being detected in real time. Detecting the output voltage waveform abnormality of the full-control inverter bridge at a certain moment, diagnosing by using a BP neural network, and inputting a U1、U2、U3、U4To obtain N1N2N3N4N5N6Is 100001, knowing VT7、VT8And (4) failure.

Claims (5)

1. A method for diagnosing faults of a static frequency converter based on a BP neural network comprises a full-control rectifier bridge, a smoothing reactor and a full-control inverter bridge, wherein the alternating current side of the full-control rectifier bridge is used as the three-phase input end of the static frequency converter, one end of the direct current side of the full-control rectifier bridge is connected with one end of the direct current side of the full-control inverter bridge through the smoothing reactor, and the other end of the direct current side of the full-control rectifier bridge is directly connected with the other end of the direct current side of the full-control inverter bridge; the alternating current side of the full-control inverter bridge is connected with a motor; characterized in that the method comprises the following steps:
step 1: taking all faults which are possible to be used as network training samples under four conditions that only one thyristor is damaged in the full-control rectifier bridge, the upper thyristor and the lower thyristor of the same bridge arm of the full-control rectifier bridge are damaged, the two thyristors in the same half bridge of the full-control rectifier bridge are damaged and the two crossed thyristors of different half bridges of the full-control rectifier bridge are damaged when the trigger angle of the full-control rectifier bridge is 0, and taking a direct-current component X of the potential difference between the cathode and the anode of the full-control rectifier bridge1Amplitude of fundamental wave X2Second harmonic amplitude X3Third harmonic amplitude X4As an input value, training a BP neural network of the fully-controlled rectifier bridge; obtaining a corresponding relation between the output value and the full-control rectifier bridge fault thyristor;
step 2: taking the trigger angle of the full-control inverter bridge as 0In the case of all faults under four conditions that only one thyristor in the full-control inverter bridge is damaged, the upper thyristor and the lower thyristor of the same bridge arm of the full-control inverter bridge are damaged, the two thyristors in the same half bridge of the full-control inverter bridge are damaged, and the two crossed thyristors in different half bridges of the full-control inverter bridge are damaged, all the faults can be used as network training samples, and the input voltage amplitude U of the full-control inverter bridge is taken1Phase A output voltage amplitude U2B phase output voltage amplitude U3C phase output voltage amplitude U4As an input value, training a full-control inverter bridge BP neural network; obtaining a corresponding relation between the output value and the full-control inverter bridge fault thyristor;
and step 3: detecting the output voltages of a fully controlled rectifier bridge and a fully controlled inverter bridge of the static frequency converter in real time, entering a step 4 when detecting that the output voltage of the fully controlled rectifier bridge is abnormal, and entering a step 5 when detecting that the output voltage of the fully controlled inverter bridge is abnormal;
and 4, step 4: extracting direct current component X of potential difference of cathode and anode of full-control rectifier bridge1Amplitude of fundamental wave X2Second harmonic amplitude X3Third harmonic amplitude X4Inputting the full-control rectifier bridge BP neural network to obtain corresponding output, and positioning the failed thyristor according to the corresponding relation between the output value and the failed thyristor of the full-control rectifier bridge;
and 5: calculating input voltage amplitude U of full-control inverter bridge1Phase A output voltage amplitude U2B phase output voltage amplitude U3C phase output voltage amplitude U4And inputting the full-control inverter bridge BP neural network to obtain corresponding output, and positioning the failed thyristor according to the corresponding relation between the output value and the failed thyristor of the full-control inverter bridge.
2. The method for diagnosing the fault of the stationary frequency converter based on the BP neural network as claimed in claim 1, wherein: the full-control rectifier bridge is provided with six thyristors VT1、VT2、VT3、VT4、VT5、VT6Wherein the thyristor VT1、VT4Are positioned on the same bridge arm; the thyristor VT3、VT6Are positioned on the same bridge arm; the thyristorVT5、VT2Are positioned on the same bridge arm;
the step 1 specifically comprises:
the failure types of the fully-controlled rectifier bridge thyristor are as follows:
the fault type is one, only one thyristor in the fully-controlled rectifier bridge is damaged;
the second fault type is that the upper thyristor and the lower thyristor of the same bridge arm of the fully-controlled rectifier bridge are damaged;
the fault type is three, wherein two thyristors in the same half bridge of the full-control rectifier bridge are damaged;
the fault type is four, two crossed thyristors of different half-bridges of the full-control rectifier bridge are damaged;
the full-control rectifier bridge BP neural network is divided into 3 layers, an input layer, a hidden layer and an output layer;
the number of input nodes of the BP neural network of the fully-controlled rectifier bridge is 4, and the direct-current component X of the potential difference between the cathode and the anode of the fully-controlled rectifier bridge is taken1Amplitude of fundamental wave X2Second harmonic amplitude X3Third harmonic amplitude X4As an input value;
taking 15 nodes of the hidden layer of the BP neural network of the fully-controlled rectifier bridge;
the number of output nodes of the BP neural network of the fully-controlled rectifier bridge is 6, and the output nodes are Y respectively1、Y2、Y3、Y4、Y5、Y6(ii) a Each node output value is 0 or 1; y is1Y2Y3When 001, the fault type is one; y is1Y2Y3When 010 is the number, the fault type II is shown; y is1Y2Y3When the value is 011, the fault type is represented as three; y is1Y2Y3When the number is 100, the fault type is four;
when Y is1Y2Y3At 001, there are 6 fault conditions for fault type one: y is4Y5Y6When 001, it represents VT1A failure; y is4Y5Y6When 010 indicates VT2A failure; y is4Y5Y6At 011, it represents VT3A failure; y is4Y5Y6When it is 100, it represents VT4A failure; y is4Y5Y6When 101, it represents VT5A failure; y is4Y5Y6110, denotes VT6A failure;
when Y is1Y2Y3At 010, there are 3 fault conditions for the second fault type: y is4Y5Y6When 001, it represents VT1、VT4A failure; y is4Y5Y6When 010 indicates VT2、VT5A failure; y is4Y5Y6At 011, it represents VT3、VT6A failure;
when Y is1Y2Y3When the fault type is 011, there are 6 fault conditions in the third fault type: y is4Y5Y6When 001, it represents VT1、VT3A failure; y is4Y5Y6When 010 indicates VT2、VT4A failure; y is4Y5Y6At 011, it represents VT3、VT5A failure; y is4Y5Y6When it is 100, it represents VT4、VT6A failure; y is4Y5Y6When 101, it represents VT1、VT5A failure; y is4Y5Y6110, denotes VT2、VT6A failure;
when Y is1Y2Y3At 100, there are 6 fault cases for the four fault types: y is4Y5Y6When 001, it represents VT1、VT2A failure; y is4Y5Y6When 010 indicates VT2、VT3A failure; y is4Y5Y6At 011, it represents VT3、VT4A failure; y is4Y5Y6When it is 100, it represents VT4、VT5A failure; y is4Y5Y6When 101, it represents VT5、VT6A failure; y is4Y5Y6110, denotes VT1、VT6And (4) failure.
3. The method for diagnosing the fault of the stationary frequency converter based on the BP neural network as claimed in claim 2, wherein: the normalization processing principle of the output value of the BP neural network diagnosis method of the fully-controlled rectifier bridge is Y1、Y2、Y3When the output value of (b) is greater than or equal to 0.5, 1 is assumed, and when less than 0.5, 0 is assumed.
4. The method for diagnosing the fault of the stationary frequency converter based on the BP neural network as claimed in claim 1, wherein: the full-control inverter bridge is provided with six thyristors VT7、VT8、VT9、VT10、VT11、VT12Wherein the thyristor VT7、VT10Are positioned on the same bridge arm; the thyristor VT9、VT12Are positioned on the same bridge arm; the thyristor VT11、VT8Are positioned on the same bridge arm; the step 2 specifically comprises:
the failure type of the fully-controlled inverter bridge thyristor is as follows:
the fault type is one, only one thyristor in the full-control inverter bridge is damaged;
the second fault type is that the upper thyristor and the lower thyristor of the same bridge arm of the full-control inverter bridge are damaged;
the fault type is three, wherein two thyristors in the same half bridge of the full-control inverter bridge are damaged;
and the fault type is four, and two crossed thyristors of different half-bridges of the full-control inverter bridge are damaged.
The full-control inverter bridge BP neural network is divided into 3 layers, an input layer, a hidden layer and an output layer;
the number of input nodes of the BP neural network of the fully-controlled inverter bridge is 4, and the input voltage amplitude U of the fully-controlled inverter bridge is taken1Phase A output voltage amplitude U2B phase output voltage amplitude U3C phase output voltage amplitude U4As an input value;
taking 15 nodes of a hidden layer of the full-control inverter bridge BP neural network;
the number of output nodes of the BP neural network of the full-control inverter bridge is 6, and the output nodes are N respectively1、N2、N3、N4、N5、N6(ii) a Each node output value is 0 or 1; n is a radical of1N2N3When 001, the fault type is one; n is a radical of1N2N3When 010 is the number, the fault type II is shown; n is a radical of1N2N3When the value is 011, the fault type is represented as three; n is a radical of1N2N3When the number is 100, the fault type is four;
when N is present1N2N3At 001, there are 6 fault conditions for fault type one: n is a radical of4N5N6When 001, it represents VT7A failure; n is a radical of4N5N6When 010 indicates VT8A failure; n is a radical of4N5N6At 011, it represents VT9A failure; n is a radical of4N5N6When it is 100, it represents VT10A failure; n is a radical of4N5N6When 101, it represents VT11A failure; n is a radical of4N5N6110, denotes VT12A failure;
when N is present1N2N3At 010, there are 3 fault conditions for the second fault type: n is a radical of4N5N6When 001, it represents VT7、VT10A failure; n is a radical of4N5N6When 010 indicates VT8、VT11A failure; n is a radical of4N5N6At 011, it represents VT9、VT12A failure;
when N is present1N2N3When the fault type is 011, there are 6 fault conditions in the third fault type: n is a radical of4N5N6When 001, it represents VT7、VT9A failure; n is a radical of4N5N6When 010 indicates VT8、VT10A failure; n is a radical of4N5N6At 011, it represents VT9、VT11A failure; n is a radical of4N5N6When it is 100, it represents VT10、VT12A failure; n is a radical of4N5N6When 101, it represents VT7、VT11A failure; n is a radical of4N5N6110, denotes VT8、VT12A failure;
when N is present1N2N3At 100, there are 6 fault cases for the four fault types. N is a radical of4N5N6When 001, it represents VT7、VT8A failure; n is a radical of4N5N6When 010 indicates VT8、VT9A failure; n is a radical of4N5N6At 011, it represents VT9、VT10A failure; n is a radical of4N5N6When it is 100, it represents VT10、VT11A failure; n is a radical of4N5N6When 101, it represents VT11、VT12A failure; n is a radical of4N5N6110, denotes VT7、VT12And (4) failure.
5. The method for diagnosing the fault of the stationary frequency converter based on the BP neural network as claimed in claim 4, wherein: the normalization processing principle of the output value of the full-control inverter bridge BP neural network diagnosis method is N4、N5、N6When the output value of (b) is greater than or equal to 0.5, 1 is assumed, and when less than 0.5, 0 is assumed.
CN202010595473.6A 2020-06-28 2020-06-28 Method for diagnosing fault of static frequency converter based on BP neural network Pending CN111812435A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010595473.6A CN111812435A (en) 2020-06-28 2020-06-28 Method for diagnosing fault of static frequency converter based on BP neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010595473.6A CN111812435A (en) 2020-06-28 2020-06-28 Method for diagnosing fault of static frequency converter based on BP neural network

Publications (1)

Publication Number Publication Date
CN111812435A true CN111812435A (en) 2020-10-23

Family

ID=72855507

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010595473.6A Pending CN111812435A (en) 2020-06-28 2020-06-28 Method for diagnosing fault of static frequency converter based on BP neural network

Country Status (1)

Country Link
CN (1) CN111812435A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112485629A (en) * 2020-11-26 2021-03-12 电子科技大学 IGBT converter health assessment method based on harmonic analysis

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106054053A (en) * 2016-06-15 2016-10-26 西安科技大学 Method for fault diagnosis of three-phase full-bridge rectifier circuit thyristor
CN107153735A (en) * 2017-04-28 2017-09-12 无锡开放大学 Motor driven systems PWM inverter method for diagnosing faults
CN108872751A (en) * 2018-07-05 2018-11-23 西南交通大学 A kind of method for diagnosing faults of three level Cascade H-Bridge Inverter neural network based
CN109085436A (en) * 2018-08-17 2018-12-25 南京航空航天大学 Non-intrusion type three-phase full-bridge inverter combined failure diagnostic method
CN110084106A (en) * 2019-03-19 2019-08-02 中国地质大学（武汉） Microgrid inverter method for diagnosing faults based on wavelet transformation and probabilistic neural network
CN110208721A (en) * 2019-07-09 2019-09-06 西南交通大学 A kind of method for diagnosing faults and device cascading three-level inverter
CN111178430A (en) * 2019-12-28 2020-05-19 上海电机学院 Particle swarm optimization-based rectifier circuit fault diagnosis method for neural network

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106054053A (en) * 2016-06-15 2016-10-26 西安科技大学 Method for fault diagnosis of three-phase full-bridge rectifier circuit thyristor
CN107153735A (en) * 2017-04-28 2017-09-12 无锡开放大学 Motor driven systems PWM inverter method for diagnosing faults
CN108872751A (en) * 2018-07-05 2018-11-23 西南交通大学 A kind of method for diagnosing faults of three level Cascade H-Bridge Inverter neural network based
CN109085436A (en) * 2018-08-17 2018-12-25 南京航空航天大学 Non-intrusion type three-phase full-bridge inverter combined failure diagnostic method
CN110084106A (en) * 2019-03-19 2019-08-02 中国地质大学（武汉） Microgrid inverter method for diagnosing faults based on wavelet transformation and probabilistic neural network
CN110208721A (en) * 2019-07-09 2019-09-06 西南交通大学 A kind of method for diagnosing faults and device cascading three-level inverter
CN111178430A (en) * 2019-12-28 2020-05-19 上海电机学院 Particle swarm optimization-based rectifier circuit fault diagnosis method for neural network

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112485629A (en) * 2020-11-26 2021-03-12 电子科技大学 IGBT converter health assessment method based on harmonic analysis

Similar Documents

Publication Publication Date Title
Li et al. A fast diagnosis method for both IGBT faults and current sensor faults in grid-tied three-phase inverters with two current sensors
El Khil et al. Diagnosis of open-switch and current sensor faults in PMSM drives through stator current analysis
Liu et al. Fault localization strategy for modular multilevel converters under submodule lower switch open-circuit fault
Jlassi et al. A single method for multiple IGBT, current, and speed sensor faults diagnosis in regenerative PMSM drives
Selvaraj et al. Fault tolerant operation of parallel-connected 3L-neutral-point clamped back-to-back converters serving to large hydro-generating units
de Freitas et al. Rectifier topologies for permanent magnet synchronous generator on wind energy conversion systems: A review
Jlassi et al. Multiple open-circuit faults diagnosis in back-to-back converters of PMSG drives for wind turbine systems
Ko et al. Fault diagnosis of three-parallel voltage-source converter for a high-power wind turbine
Pires et al. A multilevel fault-tolerant power converter for a switched reluctance machine drive
Hu et al. Fault diagnosis and tolerance with low torque ripple for open-switch fault of IM drives
CN111123162A (en) Three-phase transformer short-circuit parameter online monitoring method based on fundamental wave positive sequence component
Peng et al. Open-circuit fault diagnosis and fault tolerance for shunt active power filter
Ronanki et al. Quasi-online low-frequency impedance monitoring scheme for submodule capacitors in modular multilevel converters
CN111812435A (en) Method for diagnosing fault of static frequency converter based on BP neural network
Zhang et al. A simple open-circuit fault diagnosis method for grid-tied T-type three-level inverters with various power factors based on instantaneous current distortion
Jlassi et al. IGBTs and current sensors fault diagnosis in direct-drive PSMG wind turbine systems using adaptive thresholds
Ni et al. Overview on fault-tolerant four-switch three-phase voltage source converters
CN115579931A (en) Harmonic voltage suppression method, system, device and medium for hybrid Direct Current (DC) transmission system
Zdiri et al. SSTPI-IM reconfiguration and diagnostic under OCF appearance used in PV system
Guo et al. A double-voltage vector based model predictive control method for three phase four-switch fault-tolerant converter
Gjerde et al. Fault tolerance of a 10 MW, 100 kV transformerless offshore wind turbine concept with a modular converter system
Zhang et al. Leg-shared fault-tolerant predictive control of four-quadrant motor drive systems
Garces et al. Power collection array for improved wave farm output based on reduced matrix converters
Feng et al. A novel fault diagnosis method for power converter of doubly salient electro-magnetic motor
Yu et al. A sensor fault diagnosis method for rectifier used in traction systems

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20201023

RJ01 Rejection of invention patent application after publication