JP5287036B2 - AC ΔI type fault detection method and fault detection device - Google Patents

Description

The present invention relates to an AC ΔI type fault detection method and fault detection device for detecting faults in AC power supply circuits that supply AC power from substations to electric cars via electric railroad tracks.

As a method for detecting a fault in this type of AC feeding circuit, for example, in an AC electric railway that supplies AC power from a substation to an electric car through a contact line, when the load current vector of the electric car is [I _L ] and the fault current vector when a fault occurs in the contact line is [I _S ],
[ΔI] = [ _IS ] - [ _IL ]
There has been proposed an AC ΔI type fault selection method in which a fault is selected when the magnitude of the difference vector [ΔI] calculated by the following formula exceeds a predetermined value (see, for example, Japanese Patent Application Laid-Open No. 2003-233363).
Japanese Patent Application Publication No. 9-93791

However, in the conventional example described in Patent Document 1, fault detection can be easily performed even in the case of PWM-controlled vehicles, but the current vector diagram when an electric vehicle load current flows in the overhead contact line is as shown in Figure 4(a). As shown in Figure 4(a), when a fault current _IS is simultaneously superimposed when an electric vehicle load current _IL flows, the current difference vector [ΔI ₁₁ ] that arises in the event of a fault, calculated by the following formula, becomes a small value.
[ΔI ₁₁ ] = [ _IS ] - [ _IL ]

On the other hand, when a normal electric car load current _IL flows between times t0 and t1 in Fig. 5(a), the feeding current shown by characteristic curve L1 as shown in Fig. 5(b) contains third harmonic components of about 20% or less as shown by characteristic curve L2, and during a fault from time t1 in Fig. 5(a) onwards, the feeding voltage drops, the electric car load current _IS containing harmonic components increases, and the harmonic components (third harmonic, etc.) of the feeding current decrease as shown by characteristic curve L2 in Fig. 5(b). By utilizing these factors and multiplying the current vector [ _IL ] by the suppression coefficient _KfL that suppresses the pre-fault current and further by multiplying the current vector [ _IS ] by the suppression coefficient _KfS that suppresses the post-fault current, _ΔI12 that rises during a fault appears to be large as shown in Fig. 4(b), making it easier to detect the fault. The suppression coefficient _KfS is approximately ₁ because the harmonic components (third harmonic, etc.) of the feeding current are approximately zero.
[ΔI ₁₂ ] = [ _IS ] x Kf _S - [ _IL ] x Kf _L

However, due to the current change during a fault, transient harmonic components (such as the third harmonic) may occur, and the harmonic content may exceed 100%. This influence makes it difficult to accurately detect AC ΔI type faults by suppressing harmonics, which is an unsolved issue.
Therefore, the present invention has been made with attention to the unresolved problems of the above-mentioned conventional examples, and has an object to provide an AC ΔI type fault detection method and fault detection device that can detect faults occurring in electric railway lines with high sensitivity by accurately determining the harmonic content before and after a fault without being affected by transient harmonic components generated by current changes during a fault.

In order to achieve the above object, the AC ΔI type fault detection method according to claim 1 is an AC ΔI type fault detection method for detecting faults in an AC power supply circuit that supplies AC power from a substation to an electric car through a train line by calculating a current difference vector ΔI, and is characterized in that it detects the current supplied to the train line, calculates a harmonic content based on the fundamental wave component and harmonic components of the detected current, detects a minimum harmonic content that is the minimum value from the high frequency content within a period going back a predetermined time from the present when a transient harmonic component occurs at the time of a fault, calculates a suppression coefficient based on the minimum harmonic content, multiplies the suppression coefficient by the fundamental wave component of the detected current to calculate a harmonic suppression current, calculates a current difference vector ΔI based on the current harmonic suppression current and the harmonic suppression current at a time going back beyond the present and beyond the predetermined time when the transient harmonic component generated at the time of a fault attenuates, and compares the current difference vector ΔI with a set value to detect the fault.

The AC ΔI type fault detection device according to claim 2 is an AC ΔI type fault detection device that detects a fault in an AC feeding circuit that supplies AC power from a substation to an electric car through an overhead contact line by calculating a current difference vector ΔI,
a current detection unit that detects a current supplied to the electric rail; a filter unit that extracts a fundamental wave component and a harmonic component of the detected current; a harmonic content calculation unit that calculates a harmonic content based on the extracted fundamental wave component and harmonic component; a high frequency content storage unit that sequentially stores the calculated high frequency content; a minimum harmonic content detection unit that detects a minimum harmonic content that is the minimum value among a current high frequency content calculated by the harmonic content calculation unit and a high frequency content within a period going back a predetermined time from the present when a transient harmonic component occurs at the time of a fault that is stored in the high frequency content storage unit; a harmonic suppression current calculation unit that calculates a harmonic suppression current by multiplying a suppression coefficient calculated by the harmonic suppression current calculation unit by the fundamental wave component of the current extracted by the filter unit; a harmonic suppression current memory unit that sequentially stores the harmonic suppression current calculated by the harmonic suppression current calculation unit; a current difference vector calculation unit that calculates a current difference vector ΔI based on the present harmonic suppression current calculated by the harmonic suppression current calculation unit and the harmonic suppression current stored in the harmonic suppression current memory unit at a point going back from the present to a predetermined time period during which transient harmonic components generated in the event of a fault decay; and a fault detection unit that compares the calculated current difference vector ΔI with a set value to detect a fault.

According to the present invention, when calculating the suppression coefficient for calculating the harmonic suppression current before and after the fault, the suppression coefficient Kf is calculated based on the minimum value of the harmonic content rate for a certain period of time that exceeds the period during which the transient harmonic components generated by the current change at the time of the fault exist. Therefore, when a fault occurs in the electric rail of the AC power feeder circuit, the fault can be detected reliably and quickly, and the AC power feeder circuit breaker can be opened when this fault is detected. This has the effect of preventing damage to the fault point and various equipment and minimizing the expansion of the fault.

In addition, harmonic currents (such as third harmonics) contained in electric vehicle load currents are common when electric vehicles are diode rectifier vehicles or thyristor phase control vehicles, and the fault detection method and device of the present invention are effective for this purpose. Although the harmonic currents contained in the electric vehicle load current of PWM controlled vehicles are decreasing, there is still the need to calculate the current difference vector ΔI so as not to be affected by the transient harmonic components generated by current changes during a fault, and the present invention makes it possible to perform accurate AC ΔI type fault detection regardless of the train control method.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a basic configuration diagram showing an AC power feeding system to which an AC ΔI type fault detection device according to the present invention is applied.
In FIG. 1 , reference numeral 1 denotes an AC power feeding system. In this AC power feeding system 1, AC power output from a substation 2 is supplied to an electric rail line 3, and the AC power supplied to the electric rail line 3 is supplied via a pantograph 6 to an electric car 5 running on rails 4.

The AC voltage applied to the contact line 3 is stepped down by an instrument transformer 7 and input as an input voltage to an AC ΔI type fault detection device 8. An input current Ii detected by an instrument current transformer 9 arranged on the contact line 3 is input to this AC ΔI type fault detection device 8.
The AC ΔI type fault detection device 8 is constituted by, for example, a microcomputer, and its functional block diagram is as shown in FIG.

That is, first, the input current Ii detected by the instrument current transformer 9 is supplied to a filter unit 11, which extracts the fundamental wave component If1 and third harmonic component If3 of the electric vehicle load current. The extracted fundamental wave component If1 and third harmonic component If3 are then supplied to a harmonic content calculation unit 12. When the fundamental wave component If1 is approximately zero, this harmonic content calculation unit 12 sets the harmonic content nf(t) to the maximum value of "1," and when the fundamental wave component If1 is not approximately zero, it calculates the harmonic content nf(t) according to the following formula (1).
nf(t)=|If3|÷|If1|...(1)

The harmonic content rates nf(t) calculated by the harmonic content rate calculation unit 12 are then stored in a harmonic content rate storage unit 13 that is composed of a shift register set to a predetermined number of stages.
Then, of the harmonic content nf(t) calculated by the harmonic content calculation unit 12, the harmonic content nf(t) stored in the harmonic content storage unit 13 within a predetermined time exceeding the period during which transient harmonic components generated by current changes during a fault exist is supplied to the minimum harmonic content calculation unit 14. For example, if the predetermined time exceeding the period during which transient harmonic components generated by current changes during a fault exist is two cycles, the harmonic content nf(t) at the latest time t, the harmonic content nf(t-1) at time t-1 one cycle ago, and the harmonic content nf(t-2) at time t-2 two cycles ago are supplied to the minimum harmonic content calculation unit 14.

The minimum harmonic content calculation unit 14 calculates the minimum harmonic content nfmin by performing the calculation of the following equation (2).
nfmin=min{nf(t), nf(t-1), ..., nf(t-P)} ...(2)
Here, min{ } is a function for finding the minimum value in { }, and P is the number of cycles of a predetermined time exceeding the period during which transient harmonic components generated by current changes during a fault exist.

The harmonic content nf(t) when transient harmonics occur when a fault occurs is greater than the harmonic content nf(t) before the fault occurs and when the transient harmonics are attenuated during the fault, so by calculating the minimum harmonic content nfmin as described above, the harmonic content nf(t) when transient harmonics occur when a fault occurs is excluded.

The minimum harmonic content rate nfmin calculated by the minimum harmonic content rate calculation unit 14 is supplied to the suppression coefficient calculation unit 15. The suppression coefficient calculation unit 15 first calculates the suppression coefficient Kf by performing the calculation of the following equation (3).
Kf=1-nfmin×K...(3)
Here, K is a correction coefficient, and since the third harmonic contained in the feeding current is about 20% or less when the electric vehicle load current is flowing, K is set to, for example, 3.33 to ensure the maximum current suppression effect and to prevent the suppression coefficient Kf from becoming a negative value. However, since Kf may become a negative value, in such a case, Kf is reset to 0. For example, when the main wave component If1 is approximately zero, nfmin=1, which corresponds to such a case. Then, the suppression coefficient Kf calculated by the suppression coefficient calculation unit 15 and the fundamental wave component If1 extracted by the filter unit 11 described above are supplied to the harmonic suppression current calculation unit 16. In this harmonic suppression current calculation unit 16, the harmonic suppression current Ifs(t) is calculated by the following formula (4).
Ifs(t)=If1×Kf...(4)

The harmonic suppression current Ifs(t) calculated by the harmonic suppression current calculation unit 16 is stored in sequence in a harmonic suppression current storage unit 17 that is constituted by, for example, a shift register.
If the cycle in which the period in which the transient harmonic components exist at the time of fault occurrence is exceeded and the transient harmonics attenuate is denoted as Q, then the harmonic suppression current Ifs(t) calculated by the harmonic suppression current calculation unit 16 and the harmonic suppression current If(t-Q) Q cycles earlier stored in the harmonic suppression current storage unit 17 are supplied to the current difference vector calculation unit 18. For example, when the period in which the transient harmonic components exist at the time of fault occurrence is exceeded in 3 to 5 cycles and the transient harmonics attenuate, one of the harmonic suppression currents If(t-3), If(t-4) and If(t-5), for example If(t-4), is supplied to the current difference vector calculation unit 18.

The current difference vector calculation unit 18 calculates the current difference vector ΔIf by performing the calculation of the following equation (5).
ΔIf=|Ifs(t)-Ifs(t-Q)|...(5)
By performing the calculation of this equation (5), Ifs(t) represents the harmonic suppression current after the fault and Ifs(t-Q) represents the harmonic suppression current before the fault, and therefore, the following equations (6) and (7) can be used.
[I _S ]×Kf _S =Ifs(t)...(6)
[I _L ]×Kf _L =Ifs(t-Q)...(7)

Therefore, the above equation (5) calculates a current difference vector [ΔI] that is not affected by transient harmonic components that occur due to current changes during a fault.
[ΔI] = [ _IS ] x Kf _S - [ _IL ] x Kf _L ... (8)
The current difference vector ΔI calculated by this current difference vector calculation unit 18 is supplied to a fault detection unit 19. This fault detection unit 19 judges whether the input current difference vector ΔI is equal to or larger than a predetermined value ΔIs, and determines that no fault has occurred when ΔI<ΔIs, and determines that a fault has occurred when ΔI≧ΔIs, and outputs a fault detection signal BS that opens the AC feeder circuit breaker 20.

Moreover, the AC ΔI type fault detection device 8 executes a fault detection process shown in FIG.
This fault detection process is executed as a timer interrupt process at predetermined time intervals, as shown in FIG. 3. First, in step S1, the contact line current Ii detected by the instrument current transformer 9 is read in. Next, the process proceeds to step S2, where filtering is performed to extract the fundamental wave component If1 and the third harmonic component If3 of the contact line current Ii.

Then, proceed to step S3 to determine whether the absolute value of the extracted fundamental wave component If1 is equal to or less than a predetermined value δI and is approximately "0" (|If1|<δI). If the fundamental wave component If1 is approximately "0", proceed to step S4 to set the harmonic content rate nf(t) to "1" and then proceed to step S6. If the fundamental wave component If1 is not approximately "0", proceed to step S5 to calculate the harmonic content rate nf(t) by performing the calculation of formula (1) above.

Then, proceed to step S6, where the calculated harmonic content nf(t) is stored in a harmonic content storage area configured as a shift register formed in a RAM or the like, and then proceed to step S7. Here, if the number of cycles of a predetermined time exceeding the time during which transient harmonic components generated by current changes during a fault exist is P, the harmonic content nf(t) calculated in step S4 or S5 and the harmonic content nf(t-1) from one cycle ago to the harmonic content nf(t-P) P cycles ago stored in the harmonic content storage area are read, and the minimum harmonic content nfmin is calculated by performing the calculation of formula (2) above.

Next, the process proceeds to step S8, where the suppression coefficient Kf is calculated by performing the calculation of equation (3) based on the minimum harmonic content rate nfmin calculated in step S7, and then the process proceeds to step S9, where it is determined whether the calculated suppression coefficient Kf is a negative value.If Kf<0, the process proceeds to step S10, where the calculated suppression coefficient Kf is changed to "0" and then the process proceeds to step S11.If Kf≧0, the process proceeds directly to step S11.
In step S11, the harmonic suppression current Ifs(t) is calculated by multiplying the suppression coefficient Kf calculated in step S8 or S10 by the fundamental wave component If1 extracted in step S2 using the above equation (4), and then the process proceeds to step S12.

In step S12, the harmonic suppression current Ifs(t) calculated in step S11 is stored in a harmonic suppression current storage area configured as a shift register formed in a RAM or the like, and then the process proceeds to step S13, where the harmonic suppression current Ifs(t) calculated in step S12 and the harmonic suppression current Ifs(t-Q) stored in the harmonic suppression current storage area, which is beyond the period during which the transient harmonic component exists at the time of the fault occurrence and is the number of cycles Q before the transient harmonic decays, are read, and based on these, the calculation of equation (5) is performed to calculate the current difference vector ΔI, and then the process proceeds to step S14.

In this step S14, it is determined whether the current difference vector ΔI calculated in step S13 is equal to or greater than a preset value ΔIs. If ΔI < ΔIs, it is determined that no fault has occurred in the contact line 3, and the timer interrupt process is terminated and the process returns to a predetermined main program. If ΔI ≥ ΔIs, it is determined that a fault has occurred in the contact line 3, and the process proceeds to step S15, where a fault signal BS is output to the AC feeder circuit breaker 20, and the fault detection process is terminated.

In the processing of FIG. 3, the processing of step S2 corresponds to the filter unit 11, the processing of steps S3 to S5 corresponds to the harmonic content calculation unit 12, the processing of step S6 corresponds to the harmonic content storage unit 13, the processing of step S7 corresponds to the minimum harmonic content calculation unit 14, the processing of steps S8 to S9 corresponds to the suppression coefficient calculation unit 15, the processing of step S11 corresponds to the harmonic suppression current calculation unit 16, the processing of step S12 corresponds to the suppression current storage unit 17, the processing of step S13 corresponds to the current difference vector calculation unit 18, and the processing of steps S14 and S15 corresponds to the fault detection unit 19.

Next, the operation of the above embodiment will be described with a concrete example.
If we assume that the train load current flowing through the contact line 3 continues to be approximately zero, the input current Ii input to the AC ΔI type fault detection device 8 will also be approximately zero, and the fundamental wave component If1 extracted in step S2 will also be approximately zero, and will be a value below the set value δI. As a result, the harmonic content rate nf(t) calculated by moving from step S3 to step S4 will continue to be "1." The number of cycles P of the predetermined time exceeding the time during which transient harmonic components generated by current changes at the time of a fault exist is set to 2 cycles, and the number of cycles Q during which the period during which transient harmonic components exist at the time of a fault and the transient harmonics attenuate is set to 4 cycles.

As a result, the minimum harmonic content nfmin calculated by the minimum harmonic content calculation unit 14 becomes "1", but since the suppression coefficient Kf calculated by the above formula (3) becomes a negative value, the process moves from step S9 to step S10 and the suppression coefficient Kf is set to "0". As a result, the suppression current Ifs(t) calculated by the above formula (4) becomes Ifs(t) = If1 x Kf = If1 x 0 = 0, which is approximately the same as the input current Ii, and becomes zero. And since the suppression current Ifs(t-4) four cycles ago also becomes "0", the current difference vector ΔI calculated in step S14 also becomes "0", which is less than the predetermined value ΔIs, so it is determined that the power supply system is normal, the timer interrupt is terminated, and the process returns to the predetermined main program.

When the train load current starts to flow from a state where the train load current is approximately zero, the harmonic content rate nf(t) changes from nf(t) = 1 when the train load current is approximately zero to nf(t-1) when the train load current begins to flow. When the suppression coefficient Kf calculated by the above formula (3) becomes negative, the process moves to step S10 and the suppression coefficient Kf is changed to "0". Therefore, even if the high-frequency content rate nf(t) due to the transient harmonics increases, the suppression current Ifs(t) calculated by the above formula (4) becomes "0", and the current difference vector ΔI calculated by the above formula (5) remains "0".

Thereafter, when the train load current stabilizes and the transient harmonics subside, the third harmonic component If3 becomes 20% or less of the fundamental component If1, and the harmonic content nf(t) becomes 0.2 or less.
For this reason, nf(t)<nf(t-1), nf(t-2), the minimum harmonic content nfmin=nf(t), and the suppression coefficient Kf is calculated based on this minimum harmonic content nfmin using the above equation (3).

Therefore, the suppression coefficient Kf is calculated without being affected by transient harmonics, and the suppression coefficient Kf is multiplied by the fundamental wave component If1 to calculate a relatively small suppression current Ifs(t) corresponding to the harmonic content rate nf(t), and the suppression current Ifs(t-4) from four cycles earlier is subtracted from this suppression current Ifs(t) to calculate the current difference vector ΔI. Therefore, the calculated current difference vector ΔI does not exceed the set value ΔIs, and it is possible to reliably prevent a mistaken judgment that a fault has occurred due to transient harmonics.

When the AC power supply system fails while this train load current is flowing, as shown in FIG. 5(b), the power supply voltage drops, the electric car load current containing harmonic components increases, and after the transient harmonics subside, the third harmonic component If3 of the power supply current decreases to near zero. Therefore, the suppression coefficient Kfs for suppressing the post-fault current at this time is Kfs ≈ 1.0, which is larger than the suppression coefficient Kf = 0.33 when the harmonic component content rate nf(t-4) of the pre-fault current is set to the normal maximum value of 0.2, and the fundamental wave component If1 also increases, so that the current difference vector ΔI calculated by the above-mentioned formula (5) becomes a large value and exceeds the set value ΔIs, it is determined that a failure has occurred, a fault signal BS is supplied to the AC power supply line breaker 20, and the AC power supply line breaker 20 is controlled to open. As a result, damage accidents to the fault point and various equipment can be prevented in advance, and the expansion of the fault can be minimized.

Furthermore, harmonic currents (third harmonics, etc.) contained in electric vehicle load currents are often found when the electric vehicle is a diode rectifier vehicle or a thyristor phase controlled vehicle, for which the fault detection method and apparatus of the present invention are effective. Although the harmonic currents contained in the electric vehicle load current of a PWM controlled vehicle are in a decreasing trend, there is still the need to calculate the current difference vector ΔI so as not to be affected by transient harmonic components generated by current changes during a fault, and the present invention makes it possible to perform accurate AC ΔI type fault detection regardless of the electric train control method.
In the above embodiment, the minimum value for a predetermined time (e.g., three points between two cycles) exceeding the time when transient harmonic components exist is obtained, and the calculation point of the pre-fault current is set to a cycle between three cycles and five cycles before. However, this is not limited to this, and the value two cycles before may be used to match the above-mentioned predetermined time.

1 is a schematic configuration diagram showing an embodiment in which the present invention is applied to an AC power feeding system. FIG. 2 is a functional block diagram of an AC ΔI type fault detection device. 4 is a flowchart showing an example of a fault detection process performed by the AC ΔI type fault detection device. 1 is a vector diagram illustrating a fault detection principle when a load current and a fault current overlap in a conventional example. 4 is a characteristic diagram illustrating the relationship between the load current and the fault current, and transient harmonics when the current changes. FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1...AC feeding system 2...Substation 3...Electric railway line 4...Rail 5...Electric car 7...Instrument transformer 8...AC ΔI type fault detection device 9...Instrument current transformer 11...Filter section 12...Harmonic content calculation section 13...Harmonic content storage section 14...Minimum harmonic content calculation section 15...Suppression coefficient calculation section 16...Harmonic suppression current calculation section 17...Harmonic suppression current storage section 18...Current difference vector calculation section 19...Fault detection section 20...AC feeding line circuit breaker

Claims (2)

1. An AC ΔI type fault detection method for detecting a fault in an AC feeding circuit that supplies AC power from a substation to an electric car through an overhead contact line by calculating a current difference vector ΔI, comprising the steps of:
Detecting a current supplied to the electric rail;
Calculating a harmonic content based on the fundamental wave component and the harmonic component of the detected current;
A minimum harmonic content rate is detected, which is the minimum value among the high frequency content rates within a period going back from the present to a predetermined time period during which a transient harmonic component occurs at the time of the fault;
Calculating a suppression coefficient based on the minimum harmonic content rate;
multiplying the suppression coefficient by the fundamental wave component of the detected current to calculate a harmonic suppression current;
A current difference vector ΔI is calculated based on the current harmonic suppression current and the harmonic suppression current at a point going back beyond a predetermined time period during which the transient harmonic components generated during the fault attenuate;
The AC ΔI type fault detection method comprises comparing the current difference vector ΔI with a set value to detect a fault. 1. An AC ΔI type fault detection device that detects a fault in an AC feeding circuit that supplies AC power from a substation to an electric car through a contact line by calculating a current difference vector ΔI,
a current detection unit that detects a current supplied to the electric rail; and a filter unit that extracts a fundamental wave component and a harmonic component of the detected current;
a harmonic content calculation unit that calculates a harmonic content based on the extracted fundamental wave component and harmonic components;
a high frequency content storage unit that sequentially stores the calculated high frequency content;
a minimum harmonic content detection unit that detects a minimum harmonic content, which is the minimum value, from the current harmonic content calculated by the harmonic content calculation unit and the harmonic content within a period going back from the present to a predetermined time during which a transient harmonic component is generated at the time of a fault, which is stored in the harmonic content storage unit;
a harmonic suppression current calculation unit that calculates a suppression coefficient based on the detected minimum harmonic content rate, and multiplies the calculated suppression coefficient by the fundamental wave component of the current extracted by the filter unit to calculate a harmonic suppression current;
a harmonic suppression current storage unit that sequentially stores the harmonic suppression currents calculated by the harmonic suppression current calculation unit;
a current difference vector calculation unit that calculates a current difference vector ΔI based on the current harmonic suppression current calculated by the harmonic suppression current calculation unit and the harmonic suppression current stored in the harmonic suppression current storage unit at a time going back from the present time to a time exceeding a predetermined time during which a transient harmonic component generated during a fault attenuates;
a fault detection unit that compares the calculated current difference vector ΔI with a set value to detect a fault;
An AC ΔI type fault detection device comprising:

Images