A method for detecting and locating circuit breaker faults in a three-phase inverter circuit of a UPS in a substation

Abstract

The present invention relates to a method for detecting and locating a circuit breaker fault in a three-phase inverter circuit of a substation UPS, comprising steps S1, selecting power devices T1 and D1 as research objects in the three-phase inverter circuit, performing fault simulation in MABLAB or Simulink, and obtaining simulation waveforms when power devices T1 and D1 fail; S2, analyzing the fault waveform of power device T1; S3, analyzing the fault waveform of power device D1; and S4, obtaining a method for detecting and locating a circuit breaker fault in a three-phase inverter circuit based on the fault analysis of steps S2 and S3. The present invention is scientifically and rationally designed. By extracting characteristic quantities after a three-phase inverter circuit fails and taking into account system load fluctuations and the impact during startup, a normalized DC method is selected and combined with voltage characteristics to diagnose which specific tube is faulty. Simulation verification is then performed to form a logic judgment table, thereby accurately locating the faulty device.

Description

Technical Field

[0001] The present invention belongs to the technical field of fault detection of power system substations, and in particular relates to a method for detecting and locating a circuit breaker fault of a three-phase inverter circuit of a UPS in a substation. Background Art

[0002] When a substation UPS uses DC emergency power supply, a three-phase inverter circuit is required to convert DC into three-phase AC for normal operation of the load. The three-phase inverter circuit often uses high-power electronic circuits to achieve DC to AC inversion.

[0003] However, the current safety assurance technology for high-power electronic circuits is still in the primary stage of passive safety assurance technology. The maintenance and repair basis adopted is based on the aging law of components under normal input and output conditions and environmental conditions, and does not consider the failure conditions under abnormal factors. It has great safety hazards and may even cause irreparable losses to the system. Therefore, fault diagnosis of electronic circuits is an inevitable development trend to improve system reliability and ensure safe operation of the system.

[0004] Research shows that the vast majority of three-phase inverter circuit faults are caused by short-circuit or open-circuit faults in switching tubes or diodes. Short-circuit faults often cause large short-circuit currents, which in turn turn into open-circuit faults. Therefore, we have researched and designed a detection method that can detect open-circuit faults in switching devices and diodes. By analyzing the causes of the fault waveforms and using the normalized current method for fault location and detection, we can accurately determine which tube is at fault and then perform targeted replacement. Summary of the Invention

[0005] The purpose of the present invention is to overcome the deficiencies of the prior art and to provide a method for detecting and locating a circuit breaker fault in a three-phase inverter circuit of a UPS in a substation.

[0006] The present invention solves the technical problem by the following technical solutions:

[0007] A method for detecting and locating a circuit breaker fault in a three-phase inverter circuit of a UPS in a substation, characterized in that the method comprises the following steps:

[0008] S1. In the three-phase inverter circuit, select power devices T1 and D1 as research objects, perform fault simulation in MABLAB or Simulink, and obtain the simulation waveforms when power devices T1 and D1 fail.

[0009] S2. Analyze the fault waveform of power device T1;

[0010] S3. Analyze the fault waveform of the power device D1;

[0011] S4. Based on the fault analysis of steps S2 and S3, a method for detecting and locating a circuit breaker fault in a three-phase inverter circuit is obtained.

[0012] Furthermore, the step S2 is specifically as follows:

[0013] a) Analysis of the reason why the positive half-cycle of the grid-side phase a current is zero in the three-phase inverter circuit

[0014] After T1 fails, the on and off conditions of the switch tube and freewheeling diode of phase a of the inverter are determined by the grid-side current i a The difference is due to the different flow directions:

[0015] ①i a ≥0 hours

[0016] When the switch tube T1 is working normally, a The current i in the half cycle ≥0 a The pathway consists of the following two paths:

[0017] a. When T1 is on and T2 is off, i a Circulates through T1;

[0018] b. When T1 is turned off and T2 is turned on, i a Circulates through D2;

[0019] In case ①, the switch tube T1 participates in the current i a Therefore, when T1 fails, the current i a It cannot continue to flow through T1, so the positive half cycle of phase a current is zero;

[0020] ②i a <0 o'clock

[0021] When T1 tube is not faulty, a In the half cycle when the current i a The pathway consists of the following two paths:

[0022] a. When T1 is on and T2 is off, i a Circulates through D1;

[0023] b. When T1 is turned off and T2 is turned on, i a Circulates through T2;

[0024] In both cases, the switch tube T1 does not participate in the formation of i a The flow path, so the circuit breaker fault of T1 tube is a When <0, it has no effect on the actual working state of the main circuit, and its working topology remains unchanged, so the negative half cycle of phase a current remains normal;

[0025] b) In the three-phase inverter circuit, the grid side V ao 、V bo 、V co Analysis of the causes of voltage distortion

[0026] The three-phase inverter circuit adopts the SVPWM control mode of voltage closed loop. When T1 has a short circuit fault, the area III and area IV are not affected because during this period the lower tube T6 is turned on and T1 is not turned on, so it does not affect its normal operation. However, when T1 is turned on in areas V, VI, I, and II, V ao 、V bo 、V co The voltage will be affected and distorted;

[0027] It can be seen from the above working mode that when T1 is disconnected, the three-phase inverter becomes a two-phase inverter in some areas.

[0028]

[0029] From (1), we can get:

[0030] u ao =u no (2)

[0031] However, since the three-phase voltage is no longer symmetrical, u ao The value of is not equal to zero. When phase a is not conducting, point a is equivalent to point N, so u ao Equal to u no , nor is it zero.

[0032] Furthermore, the step S3 is specifically as follows:

[0033] When D1 fails, it can be seen from the simulation waveform that the negative half cycle of the grid-side a phase current decreases, and the grid-side V ao 、V bo 、V co The voltage is distorted, and the diode plays the role of freewheeling in the three-phase inverter circuit. When the load is a resistive-inductive load, the direction of the current cannot change suddenly at the moment the switch tube is switched. This requires a diode connected in parallel to freewheel to prevent a large voltage shock.

[0034] When D1 fails and the switch tube switches from T2 to T1, i a <0, since the load is a resistive-inductive load, the current cannot continue to flow through D1, so a large induced electromotive force will be caused in the inductor, and the current in the negative half cycle of phase a will also decrease.

[0035] Furthermore, the step S4 is specifically as follows:

[0036] 1) Extraction of fault characteristics

[0037] By analyzing the waveforms after T1 and D1 are disconnected and the symmetry of the circuit, it is found that the disconnection of T1 and D2 will cause the current of phase a to decrease in the positive half cycle, but the degree of reduction is different. However, the waveform of the voltage during the fault period is very different. Therefore, the grid-side phase voltage can be selected as the fault characteristic to determine the T1 and D2 disconnection faults. When T1 and T2 fail, the voltage waveforms are not much different, but the current waveforms are very different. Therefore, the grid-side phase current can be selected as the fault characteristic to determine the T1 and T2 disconnection faults. A diagnostic method combining current and voltage is adopted, firstly using current for preliminary detection and judgment, and then using voltage for final positioning.

[0038] 2) Normalized DC method

[0039] Considering that load fluctuations and startup shocks may cause large fluctuations in output current, thus affecting the diagnosis results, the current is normalized, the average current is taken, and then normalized. The current is compared with the empirical threshold, and finally a logical judgment is made. The normalized DC value of the three phases is defined as:

[0040]

[0041] Among them: I a,b,c(av) Respectively represent the average values of the three-phase currents;

[0042] X represents the base value of three-phase current;

[0043] The base value selected when normalizing the three-phase current on the grid side must be able to reflect the actual load situation under the three-phase inverter working condition. As the base value, where i d and i q are the components of the grid-side current on the d-axis and q-axis of the two-phase rotating coordinate system,

[0044]

[0045] The normalized expression of the three-phase current is as follows:

[0046]

[0047] When a phase a arm of the DC-AC three-phase inverter experiences a short-circuit fault, the grid-side current of phase a is normalized to obtain a simulated waveform. The waveforms when T1 and D1 are short-circuited are similar. Considering the symmetry of the circuit, determining only the current characteristics is insufficient to accurately locate the fault; voltage characteristics are also required.

[0048] 3) Fault detection and location of switching tubes and diodes

[0049] Through a large number of simulations, it is known that selecting ±0.45 as the threshold can detect system circuit breaker faults and avoid misdiagnosis caused by fluctuations during startup. The judgment formula is:

[0050]

[0051] Get its diagnostic waveform, and then combine it with the phase voltage u when T1 and D1 fail ao 、u bo 、u co From the waveform diagram, we can see that when the diode fails to operate, the voltage of phase A has a great impact, so the fault can be accurately located and positioned according to the logic judgment table.

[0052] The advantages and beneficial effects of the present invention are:

[0053] The present invention provides a method for detecting and locating a circuit breaker fault in a three-phase inverter circuit of a substation UPS. By extracting characteristic quantities after a three-phase inverter circuit fault and taking into account system load fluctuations and startup impacts, the normalized DC method is selected and combined with voltage characteristics to diagnose which specific tube is faulty. Simulation verification is then performed to form a logic judgment table, thereby accurately locating the faulty device. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A three-phase inverter circuit diagram used in the present invention;

[0055] Figure 2 This is the phase current waveform diagram of T1 fault of the present invention;

[0056] Figure 3 is the phase voltage u when T1 fails in the present invention ao 、u bo 、u co Waveform graph;

[0057] Figure 4 This is the phase current waveform diagram of D1 fault of the present invention;

[0058] Figure 5 is the phase voltage u when D1 fails in the present invention ao 、u bo 、u co Waveform graph;

[0059] Figure 6 This is the equivalent topology diagram of T1 failure of the present invention;

[0060] Figure 7 This is a schematic diagram of the SVPWM control method of the present invention;

[0061] Figure 8 This is a diagram of the working state of T1 after a fault of the present invention;

[0062] Figure 9 This is a normalized waveform diagram of the a-phase current after a fault in the three-phase inverter circuit of the present invention;

[0063] Figure 10 Diagnostic waveform diagram of T1 circuit breaker fault in the present invention;

[0064] Figure 11 This is the diagnostic waveform diagram after the D1 circuit breaker fault of the present invention. DETAILED DESCRIPTION

[0065] The present invention will be further described in detail below through specific examples. The following examples are only illustrative and not restrictive, and the scope of protection of the present invention cannot be limited thereto.

[0066] like Figure 1 The figure shows a typical three-phase inverter circuit, which consists of six IGBT / Diodes and adopts voltage closed-loop SVPWM control.

[0067] The power devices T1 and D1 of the phase a bridge arm are selected as the research objects. Due to the symmetry of the circuit, the failure of other switching tubes or diodes can be analogous to the failure of phase a, which will not be described here.

[0068] Perform fault simulation in MABLAB / Simulink, input voltage V DC =230V, L=0.1mH, R=2.9Ω, simulation time is 0.4s, circuit breaker fault occurs between 0.1 and 0.2s, the simulation waveforms of the above power tube faults are as follows Figures 2 to 5 shown.

[0069] From the above simulation waveforms, it can be seen that when a short-circuit fault occurs in the power tube, the output waveform is significantly different from that under normal circumstances. If these faults are not handled in time, it will cause fatal damage to the stable operation of the system.

[0070] 1. Waveform analysis when T1 fails

[0071] 1. Analysis of the reason why the positive half cycle of the grid-side phase a current is zero

[0072] After T1 fails, the on and off conditions of the switch tube and freewheeling diode of phase a of the inverter are determined by the grid-side current i a There are great differences due to the different circulation directions, which are discussed separately below.

[0073] ①i a ≥0 hours

[0074] It is not difficult to find from the working process of the inverter that when the switch tube T1 is working normally, a The current i in the half cycle ≥0 aThe pathway consists of the following two paths:

[0075] a. When T1 is on and T2 is off, i a Circulates through T1;

[0076] b. When T1 is turned off and T2 is turned on, i a Circulates through D2;

[0077] In case ①, the switch tube T1 participates in the current i a Therefore, when T1 fails, the current i a It cannot continue to flow through T1, so the positive half cycle of phase a current is zero.

[0078] ②i a <0 o'clock

[0079] When T1 tube is not faulty, a In the half cycle when the current i a The pathway consists of the following two paths:

[0080] a. When T1 is on and T2 is off, i a Circulates through D1;

[0081] b. When T1 is turned off and T2 is turned on, i a Circulates through T2;

[0082] In both cases, the switch tube T1 does not participate in the formation of i a The flow path, so the circuit breaker fault of T1 tube is a When <0, it has no effect on the actual working state of the main circuit, and its working topology remains unchanged, so the negative half cycle of phase a current remains normal. Its equivalent topology diagram is as follows: Figure 6 shown.

[0083] 2. Network side V ao 、V bo 、V co Analysis of the causes of voltage distortion

[0084] The distortion of the grid-side voltage is closely related to the control mode of the system, so it is necessary to analyze it in combination with the control mode. In this embodiment, the three-phase inverter circuit adopts the voltage closed-loop SVPWM control mode, such as Figure 7 It is the basic voltage space vector diagram of its control method.

[0085] There are many ways to combine the basic voltage vectors. This embodiment adopts the most common seven-segment combination, as shown in Table 1. Table 1 SVPWM output voltage vector seven-segment combination

[0086] Sector SVPWM seven-segment combination Sector SVPWM seven-segment combination I <![CDATA[V0V1V2V7V2V1V0]]> IV <![CDATA[V0V5V4V7V4V5V0]]> II <![CDATA[V0V3V2V7V2V3V0]]> V <![CDATA[V0V5V6V7V6V5V0]]> III <![CDATA[V0V3V4V7V4V3V0]]> VI <![CDATA[V0V1V6V7V6V1V0]]>

[0087] When T1 is short-circuited, regions III and IV are not affected because during this period the lower tube T6 is conducting and T1 is not conducting, thus it does not affect its normal operation. However, when T1 is conducting in regions V, VI, I, and II, V ao 、V bo 、V co The voltage will be affected and distorted.

[0088] For example, in area I, T1 is disconnected, and its working mode is as follows Figure 8 As shown, the working states are 000, 100, 110, 111, 110, 100, and 000 respectively.

[0089] It can be seen from the above working mode that when T1 is disconnected, three-phase inversion is converted to two-phase inversion in a part of the area.

[0090]

[0091] From (1), we can get:

[0092] u uo =u no (2)

[0093] However, since the three-phase voltage is no longer symmetrical, u ao The value of is not equal to zero. When phase a is not conducting, point a is equivalent to point N, so u ao Equal to u no , nor is it zero.

[0094] 2. Waveform analysis when D1 fails

[0095] When D1 fails, it can be seen from the above simulation waveform that the negative half cycle of the grid-side a phase current decreases, and the grid-side V ao 、V bo 、V co The voltage is distorted because the diode plays the role of freewheeling in the three-phase inverter circuit. When the load is a resistive-inductive load, the direction of the current cannot change suddenly at the moment the switch tube is switched. This requires a diode connected in parallel to continue the current to prevent a large voltage shock.

[0096] When D1 fails and the switch tube switches from T2 to T1, i a <0, since the load is a resistive-inductive load, the current cannot continue to flow through D1, so a large induced electromotive force will be caused in the inductor, and the current in the negative half cycle of phase a will also decrease.

[0097] 3. Three-phase inverter fault diagnosis

[0098] Through the above analysis, we have a general understanding of the characteristics of the output waveform after the circuit breaker fault of each component in the three-phase inverter circuit and the causes of waveform distortion. The following introduces the detection and location of the fault.

[0099] 1. Extraction of fault characteristics

[0100] By analyzing the waveforms after T1 and D1 are disconnected and the circuit symmetry, we found that both T1 and D2 disconnections will cause a decrease in the positive half-cycle current of phase a, albeit to varying degrees. However, the voltage waveforms during the fault period are significantly different. Therefore, the grid-side phase voltage can be used as the fault characteristic to identify T1 and D2 disconnection faults. However, the voltage waveforms of T1 and T2 faults are similar, while the current waveforms are significantly different. Therefore, the grid-side phase current can be used as the fault characteristic to identify T1 and T2 disconnection faults. This analysis reveals that selecting only one of these quantities as the fault characteristic cannot accurately locate the fault. Therefore, this embodiment adopts a combined current and voltage diagnostic approach, first using current for preliminary detection and judgment, and then using voltage for final location.

[0101] 2. Normalized DC method

[0102] Considering that load fluctuations and startup shocks may cause large fluctuations in output current, thus affecting the diagnostic results, normalizing the current to per unit can avoid this problem.

[0103] The normalized DC method is to take the average value of the current, normalize it, compare it with the empirical threshold, and finally make a logical judgment. The normalized DC value of the three phases is defined as:

[0104]

[0105] Where, I a,b,c(av) Respectively represent the average value of the three-phase current, and X represents the base value of the three-phase current. The base value selected when standardizing the three-phase current on the grid side must be able to reflect the actual load situation under the three-phase inverter working condition. In this embodiment, the grid side current space vector amplitude is selected As the base value, where i d and i q are the components of the grid-side current on the d-axis and q-axis of the two-phase rotating coordinate system,

[0106]

[0107] The normalized expression of the three-phase current is as follows:

[0108]

[0109] When a phase a bridge arm of the DC-AC three-phase inverter fails, the normalized simulation waveform of the phase a grid-side current is as follows: Figure 9 As shown, they are normal, T1 circuit breaker fault, and D1 circuit breaker fault.

[0110] As can be seen from the above waveform diagram, the waveforms when T1 and D1 are open are similar. Considering the symmetry of the circuit, it is not enough to determine the current characteristics alone to accurately locate the fault. The voltage characteristics must also be determined.

[0111] 3. Fault detection and location of switch tubes and diodes

[0112] Through a large number of simulations, it is known that selecting ±0.45 as the threshold can detect system circuit breaker faults and avoid misdiagnosis caused by fluctuations during startup. The judgment formula is:

[0113]

[0114] Its diagnostic waveform is as follows Figure 10 、 11 As shown, a) is to judge d a Is it greater than 0.45 waveform, b) is to judge d a Is the waveform graph less than -0.45?

[0115] Recombination Figure 3 and Figure 5 It can be seen that when the diode is open circuit fault, the phase A voltage has a large impact, so the fault can be accurately located. The logic judgment table is shown in Table 2.

[0116] Table 2 Logical judgment table for phase a fault

[0117] T1 failure T2 failure D1 failure D2 failure <![CDATA[γ a >0.45]]> 1 0 0 1 <![CDATA[γ a <-0.45]]> 0 1 1 0 <![CDATA[u ao Changes]]> Distortion Distortion There is a big positive impact There is a big negative impact

[0118] Although the embodiments and drawings of the present invention are disclosed for illustrative purposes, those skilled in the art will understand that various substitutions, changes and modifications are possible without departing from the spirit and scope of the present invention and the appended claims. Therefore, the scope of the present invention is not limited to the contents disclosed in the embodiments and drawings.

Claims

1. A method for detecting and locating a circuit breaker fault in a three-phase inverter circuit of a UPS in a substation, characterized by: The steps of the method are: 1) Extraction of fault characteristics In a three-phase inverter circuit, power devices T1 and D1 are selected as research objects. Fault simulation is performed in MABLAB or Simulink to obtain simulation waveforms of power devices T1 and D1 when they fail. By analyzing the waveforms and circuits after T1 and D1 are disconnected, the grid-side phase voltage is selected as the fault characteristic to determine the T1 and D2 disconnection faults, and the grid-side phase current is selected as the fault characteristic to determine the T1 and T2 disconnection faults. A diagnostic method combining current and voltage is adopted, with preliminary detection and judgment first performed by current, and then final positioning performed by voltage. 2) Normalized DC method The current is normalized, the average current is taken and then normalized, and then compared with the empirical threshold. Finally, a logical judgment is made and the normalized DC value of the three phases is defined as: Among them: I a,b,c(av) Respectively represent the average values of the three-phase currents; X represents the base value of three-phase current; The base value selected when normalizing the three-phase current on the grid side must be able to reflect the actual load situation under the three-phase inverter working condition. As the base value, where i d and i q are the components of the grid-side current on the d-axis and q-axis of the two-phase rotating coordinate system, The normalized expression of the three-phase current is as follows: When a phase a arm of the DC-AC three-phase inverter experiences a short-circuit fault, the grid-side current of phase a is normalized to obtain a simulated waveform. The waveforms when T1 and D1 are short-circuited are similar. Considering the symmetry of the circuit, determining only the current characteristics is insufficient to accurately locate the fault; voltage characteristics are also required. 3) Fault detection and location of switching tubes and diodes Through a large number of simulations, ±0.45 is selected as the threshold value to detect system circuit breaker faults and avoid misdiagnosis caused by fluctuations during startup. The judgment formula is: Get its diagnostic waveform, and then combine it with the phase voltage u when T1 and D1 fail ao 、u bo 、u co The waveform diagram shows that when the diode fails, the phase a voltage has a great impact. a is equal to 1, and u ao When there is distortion, it is determined to be a T fault, and γ a When T1 is greater than 0.45, γ a When it is less than -0.45, T2 fails; when d a is equal to 1, and u ao When there is a large impact, it is judged as a D fault; u ao is a positive impact, γ a When it is less than -0.45, D1 fails, u ao is a negative shock, γ a When it is greater than 0.45, D2 is faulty.

Images