GB2555808A

GB2555808A - Electrical fault detection

Info

Publication number: GB2555808A
Application number: GB1618973.0A
Authority: GB
Inventors: Jayampathi Gajanayake Chandana; Dharmasiri Gamini Jayasinghe Shantha; Viswanathan Vaiyapuri; Kumar Gupta Amit; Nadarajan Sivakumar
Original assignee: Rolls Royce PLC
Current assignee: Rolls Royce PLC
Priority date: 2016-11-10
Filing date: 2016-11-10
Publication date: 2018-05-16

Abstract

A fault detection method is disclosed for a multiphase electrical device, where the device may be an inverter, a converter, a motor drive or a rectifier. The method comprises: measuring current (I) of each of a first phase (a) and a second phase (b) of the electrical device; calculating the root mean square value (RMS) of each measured current (la, lb); determining a current residual R by dividing magnitude of the difference of the RMS currents (la, lb) by the sum of their squares; comparing the current residual to a first fault threshold 36; and declaring a first fault 30 if the current residual exceeds the first fault threshold. The fault may be a current sensor fault. The current residual may also be compared to second and third thresholds 38, 40, and further faults 32 and 34 may be determined if it exceeds these thresholds. These faults may be open phase and open switch faults respectively.

Description

(71) Applicant(s):

Rolls-Royce pic (Incorporated in the United Kingdom)

Buckingham Gate, LONDON, SW1E 6AT, United Kingdom (72) Inventor(s):

Chandana Jayampathi Gajanayake Shantha Dharmasiri Gamini Jayasinghe Vaiyapuri Viswanathan

Amit Kumar Gupta Sivakumar Nadarajan (56) Documents Cited:

US 20100066551 A1 US 20080297089 A1 2011 IEEE Ninth International Conference on Power Electronics and Drive Systems (PEDS 2011), December 2011, IEEE, pp 956-961, Gajanayake et al., Fault tolerant control method to improve the torque and speed response in PMSM drive with winding faults (58) Field of Search:

INT CL G01R, H02K

Other: EPODOC, WPI, TXTA, INSPEC, XPIEEE, XPI3E. XPIOP, XPESP (74) Agent and/or Address for Service:

Rolls-Royce pic

IP Department (SinA-48), PO Box 31, DERBY, Derbyshire, DE24 8BJ, United Kingdom (54) Title of the Invention: Electrical fault detection

Abstract Title: Electrical fault detection by comparing RMS phase current values (57) A fault detection method is disclosed for a multiphase electrical device, where the device may be an inverter, a converter, a motor drive or a rectifier. The method comprises: measuring current (I) of each of a first phase (a) and a second phase (b) of the electrical device; calculating the root mean square value (RMS) of each measured current (la, lb); determining a current residual R by dividing magnitude of the difference of the RMS currents (la, lb) by the sum of their squares; comparing the current residual to a first fault threshold 36; and declaring a first fault 30 if the current residual exceeds the first fault threshold. The fault may be a current sensor fault. The current residual may also be compared to second and third thresholds 38, 40, and further faults 32 and 34 may be determined if it exceeds these thresholds. These faults may be open phase and open switch faults respectively.

Figure 3

At least one drawing originally filed was informal and the print reproduced here is taken from a later filed formal copy.

1/4

Figure 1

Figure 2

Figure 3

2/4

Figure 4

Figure 5

3/4 ρ

ω ω

0ί

ω ω

_Q ω

ω _C

0_

100 150 200 250

300 350

400 450 500

Figure 6

Time (ms)

Figure 7

4/4

Figure 8

Figure 9

Electrical Fault Detection

The present disclosure concerns a method to detect a fault in a multiphase electrical device. It also concerns a method to classify the detected fault. The method finds particular utility for detecting and classifying current sensor faults, open phase faults and open switch faults in an inverter fed motor drive.

Electrical devices such as inverter fed motor drives are subject to various fault conditions. One common fault is a current sensor fault. Such faults may affect the control of the electrical device, or other devices connected thereto, since current is often used in closed loop control schemes. Consequently other faults may be falsely detected by the sensor fault.

Current sensor faults may be one of three types: a gain error, an offset error and an open sensor error. A gain error, being an increase or decrease in the gain, results in both increased phase current and increased torque oscillations, at twice the fundamental frequency. An offset error applies a DC offset to the current in the faulty phase or phases and causes torque oscillation at the fundamental frequency. An open current sensor error, loss of the measurement, results in current ripples and torque ripples. This can cause instability in a motor drive system.

It is therefore desirable to detect current sensor faults, and to distinguish them from other faults such as open phase or open switch faults, so that mitigation actions can be taken to minimise the effects of the fault.

According to a first aspect there is provided a fault detection method for a multiphase electrical device, the method comprising steps to:

a) measure current of each of a first phase and a second phase of the electrical device;

b) calculate root mean square value of each measured current;

c) determine a current residual by dividing magnitude of the difference of the RMS currents by the sum of their squares;

d) compare the current residual to a first fault threshold; and

e) declare a first fault if the current residual exceeds the first fault threshold.

Advantageously the method enables different faults to be detected and distinguished. Advantageously the method enables open sensor and open phase faults to be distinguished. Advantageously the method only requires measurement of currents on two phases of a multiphase electrical device, such measurements being commonly performed on such devices. Therefore no additional sensors are required.

The first fault may comprise a current sensor fault.

The method may comprise further steps to:

a) compare the current residual to a second fault threshold; and

b) declare a second fault if the current residual exceeds the second fault threshold.

Advantageously the second fault threshold is distinct from the first fault threshold. The second fault may comprise an open phase fault. The second fault threshold may be lower than the first fault threshold.

The method may comprise further steps to:

a) compare the current residual to a third fault threshold; and

b) declare a third fault if the current residual exceeds the third fault threshold.

Advantageously the third fault threshold is distinct from the first fault threshold and the second fault threshold. The third fault may comprise an open switch fault. The third fault threshold may be lower than the second fault threshold. The third fault threshold may be lower than the first fault threshold.

The electrical device may comprise an inverter. The electrical device may comprise a motor drive. The electrical device may comprise a rectifier. The electrical device may comprise a converter.

The steps of the method may be iterated. The first fault may be declared when the current residual exceeds the first fault threshold on consecutive iterations, for example two or three consecutive iterations. The second fault may be declared when the current residual exceeds the second fault threshold on consecutive iterations, for example two or three consecutive iterations. The third fault may be declared when the current residual exceeds the third fault threshold on consecutive iterations, for example two or three consecutive iterations. Advantageously this reduces the possibility of falsely detecting a fault.

The method may comprise further steps to:

a) calculate a phase residual by dividing the difference of the RMS currents by their sum;

b) compare the phase residual to a negative threshold and declare a fault on the first phase if the negative threshold exceeds the phase residual; and

c) compare the phase residual to a positive threshold and declare a fault on the second phase if the phase residual exceeds the positive threshold.

Advantageously the method distinguishes which phase is faulty.

The method may include further steps to accommodate the detected fault. For example the method may include steps to reconstruct the current of the faulty phase. Advantageously operation of the electrical device may be maintained using the reconstructed phase current.

The skilled person will appreciate that except where mutually exclusive, a feature described in relation to any one of the above aspects may be applied mutatis mutandis to any other aspect. Furthermore except where mutually exclusive any feature described herein may be applied to any aspect and/or combined with any other feature described herein.

Embodiments will now be described by way of example only, with reference to the Figures, in which:

Figure 1 is a sectional side view of a gas turbine engine;

Figure 2 is a logic diagram of steps to calculate a current residual;

Figure 3 is a graph of current residual for three fault types;

Figure 4 is a logic diagram of steps to delay declaration of a fault;

Figure 5 is a graph showing phase residual and phase currents for a fault on phase a;

Figure 6 is a graph showing phase residual and phase currents for a fault on phase b;

Figure 7 is a logic diagram of steps to determine which phase is faulty;

Figure 8 is a logic diagram of steps to reconstruct a phase current;

Figure 9 is a graph of a reconstructed phase current.

With reference to Figure 1, a gas turbine engine is generally indicated at 10, having a principal and rotational axis 11. The engine 10 comprises, in axial flow series, an air intake 12, a propulsive fan 13, an intermediate pressure compressor 14, a high-pressure compressor 15, combustion equipment 16, a high-pressure turbine 17, an intermediate pressure turbine 18, a low-pressure turbine 19 and an exhaust nozzle 20. A nacelle 21 generally surrounds the engine 10 and defines both the intake 12 and the exhaust nozzle 20.

The gas turbine engine 10 works in the conventional manner so that air entering the intake 12 is accelerated by the fan 13 to produce two air flows: a first air flow into the intermediate pressure compressor 14 and a second air flow which passes through a bypass duct 22 to provide propulsive thrust. The intermediate pressure compressor 14 compresses the air flow directed into it before delivering that air to the high pressure compressor 15 where further compression takes place.

The compressed air exhausted from the high-pressure compressor 15 is directed into the combustion equipment 16 where it is mixed with fuel and the mixture combusted. The resultant hot combustion products then expand through, and thereby drive the high, intermediate and low-pressure turbines 17, 18, 19 before being exhausted through the nozzle 20 to provide additional propulsive thrust. The high 17, intermediate 18 and low 19 pressure turbines drive respectively the high pressure compressor 15, intermediate pressure compressor 14 and fan 13, each by suitable interconnecting shaft.

Other gas turbine engines to which the present disclosure may be applied may have alternative configurations. By way of example such engines may have an alternative number of interconnecting shafts (e.g. two) and/or an alternative number of compressors and/or turbines. Further the engine may comprise a gearbox provided in the drive train from a turbine to a compressor and/or fan.

The method will be described with respect to an inverter fed motor drive. It is also applicable to an inverter; an electronic converter; a rectifier, particularly an active rectifier; and other electrical devices.

An inverter fed motor drive is multi-phased, for example three-phased. A typical inverter fed motor drive or electronic converter has a first phase, a second phase and a third phase respectively labelled a, b, c. The current I of each phase is measured by conventional sensors to give instantaneous measured current values l_ai, Ibj and l_cj. Instantaneous current measurements laj, Ibj, Icj are stored in a buffer and used to calculate the root mean square RMS current values l_a, lb, lc- As is conventional the size of the buffer affects the smoothing of the RMS values and the computational power required to calculate the RMS values.

When the three phases of the inverter fed motor drive are balanced the RMS values of each phase current should be approximately the same; thus I_a~ I_h ~ I_c. By subtracting the RMS current of one phase, lb, from the RMS current of another phase, l_a, the difference in magnitude can be determined.

This difference is negligible where the phases are balanced but non-zero when the phases are unbalanced. The magnitude of the difference is dependent on the current drawn by an inverter fed motor. Thus a threshold current is inadequate to determine a fault condition since it would either be triggered by non-fault conditions, if set low, or would not be triggered by all fault conditions, if set too high.

The method normalises the RMS currents l_a, lb, lc using the sum of the squares of the RMS currents and calculates a current residual R. The current residual R is calculated from two of the RMS currents, for example l_a and lb. Calculation of the current residual R is shown in Figure 2 which implements the following equation in logic: R = . The current residual R is therefore calculated by first subtracting the RMS current of the second phase lb from the RMS current of the first phase l_a, and then by taking the magnitude or amplitude of the result. The result is then normalised by dividing by the sum of the square of the first RMS current and the square of the second RMS current.

As will be apparent the RMS currents used may be reversed or the third phase current l_c may be used with either the first phase current l_a or the second phase current lb without substantively changing the method.

Figure 3 is a graph showing the current residual R plotted against time for three types of fault which occur at around 0.2 seconds from the start of recording. A first fault 30 displays a relatively high current residual R, around 0.025 following initial perturbations. The first fault 30 is a current sensor fault. As described above, a current sensor fault may be a gain error, an offset error or an open sensor error. In this example the current sensor fault is an open sensor fault.

A second fault 32 displays a medium current residual R, around 0.015 following initial perturbations. The second fault 32 is an open phase fault. An open phase fault is disconnection of the phase from the source. A third fault 34 displays a low current residual R, around 0.002 with minimal initial perturbation. The third fault 34 is an open switch fault. An open switch fault is the loss of switching capability of one or more of the switching devices of a power electronics converter. Advantageously the calculation of the current residual R according to the present method enables the different types of fault to be distinguished as shown in Figure 3.

A first fault threshold 36, represented by a horizontal line on the graph in Figure 3, is set at a value of current residual R. Similarly a second fault threshold 38 and a third fault threshold 40 may also be set. The first fault threshold 36 is preferably at the highest value of current residual R and the third fault threshold 40 at the lowest value of current residual R. The second fault threshold 38 may therefore be set at an intermediate value between the first and third fault thresholds 36, 40.

The first fault threshold 36 may be set at current residual R equal to 0.02, for example. This is a level at which any current residual R values calculated which are larger than, exceed, the threshold 36 are representative of a current sensor fault. Similarly the optional second fault threshold 38 may be set at current residual R equal to 0.01, for example, which is lower than the first fault threshold 36. This is a level at which any current residual R values calculated which are larger than, exceed, the threshold 38 are representative of an open phase fault, providing they are lower than the first fault threshold 36. Similarly the optional third fault threshold 40 may be set at current residual R equal to 0.001, for example, which is lower than the first fault threshold 36 and the second fault threshold 38. This is a level at which any current residual R values calculated which are larger than, exceed, the threshold 40 are representative of an open switch fault, providing they are lower than the second fault threshold 38.

If the calculated current residual R exceeds the first fault threshold 36 the method declares a current sensor fault. Mitigation actions may then be instigated to protect the inverter fed motor drive. If the calculated current residual R exceeds the second fault threshold 38 but does not exceed the first fault threshold 36 the method declares an open phase fault. If the calculated current residual R exceeds the third fault threshold 40 but does not exceed either the first or second fault thresholds 36, 38 the method declares an open switch fault.

Alternatively the fault thresholds 36, 38, 40 may be set at levels where only the average or maximum current residuals R and not all the calculated values of current residual R exceed the threshold.

The steps of the method may be iterated over time. Thus on each iteration the instantaneous phase currents l_ai, lbj> l_cj are measured and the RMS currents l_a, Ib, lc calculated. Then the current residual R can be calculated and compared to at least the first fault threshold 36. If the iterated current residual R exceeds the first fault threshold 36 a current sensor fault 30 can be declared.

The method may optionally require the current residual R calculated in two or more consecutive iterations to exceed the first fault threshold 36 before the current sensor fault 30 is declared. Alternatively the method may optionally require that the current residual R calculated in any two or more of a predefined number of consecutive iterations, for example five iterations, exceeds the first fault threshold 36 before the current sensor fault 30 is declared. Advantageously single spurious current measurements do not trigger fault declaration.

Similarly the current residual R can be calculated and compared to at least the second fault threshold 38. If the iterated current residual R exceeds the second fault threshold 38 an open phase fault 32 can be declared.

The method may optionally require the current residual R calculated in two or more consecutive iterations to exceed the second fault threshold 38 before the open phase fault 32 is declared. Alternatively the method may optionally require that the current residual R calculated in any two or more of a predefined number of consecutive iterations, for example five iterations, exceeds the second fault threshold 38 before the open phase fault 32 is declared. Advantageously single spurious current measurements do not trigger fault declaration.

Similarly the current residual R can be calculated and compared to at least the third fault threshold 40. If the iterated current residual R exceeds the third fault threshold 40 an open switch fault 34 can be declared.

The method may optionally require the current residual R calculated in two or more consecutive iterations to exceed the third fault threshold 40 before the open switch fault 34 is declared. Alternatively the method may optionally require that the current residual R calculated in any two or more of a predefined number of consecutive iterations, for example five iterations, exceeds the third fault threshold 40 before the open switch fault 34 is declared. Advantageously single spurious current measurements do not trigger fault declaration.

Figure 4 is a logic diagram of an optional feature. In the logic diagram the first, second and third fault thresholds 36, 38, 40 are shown as logic blocks to which the current residual R is compared. A delay 50 is applied to each current residual R value which exceeds one or more of the thresholds 36, 38, 40. The delay 50 is introduced in order to allow the current residual R to settle to a constant level, although as can be seen in Figure 3 the steady state current residual R oscillates. Without the delay 50 the current residual R would first cross the lowest fault threshold, the third fault threshold 40, and so an open switch fault would be declared. After a further time the current residual R would cross the next threshold, the second fault threshold 38, and so an open phase fault would be declared. Finally the highest threshold, the first fault threshold 36, would be crossed by the current residual R and so a current sensor fault would be declared. Advantageously by introducing the delay 50 only one fault type is declared, corresponding to the value of the steady state current residual R, so there is no ambiguity as to the type of fault which has occurred.

The delay 50 may be relatively short, for example much less than 0.1s. In the example shown in Figure 3 the current residual R crosses the first fault threshold 36 within 0.02s of the fault occurring at 0.2s. The delay 50 may be longer, provided that this is not detrimental to the possible mitigation actions which can be applied to address the fault. For example, the delay 50 may be 0.1s. A delay 50 of this length would advantageously permit the current residual R to settle to a steady state oscillation for an open phase fault, line 32 in Figure 3. Advantageously the second fault threshold 38 could be closer to the expected value of current residual R in this case.

The method may include further steps once a current sensor fault 30 is declared. The further steps determine on which phase a, b of the inverter fed motor drive the fault has arisen. The further steps include calculating a phase residual P. The phase residual P is calculated from the RMS current values l_a, lb. Specifically the difference of the RMS currents is divided by the sum of the

RMS currents: P = ——— . Where there is no fault the phase residual P is

I_a+I_h approximately zero because I_a «I_h. However, where there is a fault on one of the phases a, b the phase residual P is significantly non-zero. In particular, where no instantaneous current l_aj, lbj is measured the phase residual P will be ±1.

The phase residual P calculated for a fault on the first phase a is plotted in Figure 5 against time, line 42. At approximately 200s the phase residual P drops from zero to minus one.

The phase residual P can be compared to a negative threshold 44 as shown on the graph in Figure 5. The negative threshold 44 can be set at any value which is less than zero but may advantageously be set closer to minus one, for example at -0.5 or below. Thus any phase residuals P which are close to zero but not exactly equal to zero will not cross the negative threshold 44. The method may therefore declare a fault on the first phase a if the negative threshold 44 exceeds the phase residual P, line 42. This is shown by the step change in the phase a graph in Figure 5 whereas there is no change in the phase b graph.

The phase residual P can also be compared to a positive threshold 46. The phase residual P calculated for a fault on the second phase b is plotted in Figure 6 against time, line 48. At approximately 200s the phase residual P rises from zero to one.

The phase residual P can be compared to the positive threshold 46 as shown on the graph in Figure 6. The positive threshold 46 can be set at any value which is greater than zero but may advantageously be set closer to one, for example at 0.5 or above. Thus any phase residuals P which are close to zero but not exactly equal to zero will not cross the positive threshold 46. The method may therefore declare a fault on the second phase b if the phase residual P, line 48, exceeds the positive threshold 46. This is shown by the step change in the phase b graph in Figure 6 whereas there is no change in the phase a graph.

As will be readily understood, the graphs in Figure 5 and Figure 6 are mirror inversions. A logic diagram of the optional further steps to determine which phase is faulty is shown in Figure 7.

Figure 8 is a logic diagram of further optional steps to reconstruct the phase current l_a where a current sensor fault 30 has been declared. The steps are appropriate where there is no fault on the phase a but no phase current l_a can be measured because the current sensor is faulty. Therefore it is preferable to perform the optional steps described with respect to Figure 5, Figure 6 and

Figure 7 to test the phase residual P against the positive and negative thresholds 46, 44 to ensure that there is no fault on the phase.

In order to reconstruct the phase current l_a it is first necessary to obtain the measured phase current l_bj of a healthy phase b and to calculate the RMS current lb. The RMS current lb is multiplied by the square root of 2 and then by a sinusoidal function, sin(0+ 2^/), where Θ is angle and 2^/ shifts the healthy phase current l_b by 120° because there are three equally spaced phases. Thus the reconstructed phase current I_a = -/2.1 _h sin(# + ²^). The reconstructed phase current l_a is shown in Figure 9.

The method is applicable in motor drives for marine vehicle applications including hybrid or fully electric propulsion systems are used. It is also applicable for hybrid trains and other rail vehicles; hybrid or fully electric motor vehicles; and industrial applications including electric motor drives. The method is also applicable for grid-connected inverters for connecting solar and/or wind energy to a national or local electric grid, for example using three phase inverters and active rectifiers.

It will be understood that the invention is not limited to the embodiments above-described and various modifications and improvements can be made without departing from the concepts described herein. Except where mutually exclusive, any of the features may be employed separately or in combination with any other features and the disclosure extends to and includes all combinations and sub-combinations of one or more features described herein.

Claims

1. A fault detection method for a multiphase electrical device, the method comprising steps to:

a. measure current (I) of each of a first phase (a) and a second phase (b) of the electrical device;

b. calculate root mean square value (RMS) of each measured current (la, lb);

c. determine a current residual (R) by dividing magnitude of the difference of the RMS currents (la, lb) by the sum of their squares;

d. compare the current residual (R) to a first fault threshold (36); and

e. declare a first fault (30) if the current residual (R) exceeds the first fault threshold (36).

2. A method as claimed in claim 1 wherein the first fault (30) comprises a current sensor fault.

3. A method as claimed in claim 1 or claim 2 comprising further steps to:

a. compare the current residual (R) to a second fault threshold (38); and

b. declare a second fault (32) if the current residual (R) exceeds the second fault threshold (38).

4. A method as claimed in claim 3 wherein the second fault (32) comprises an open phase fault.

5. A method as claimed in claim 3 or claim 4 wherein the second fault threshold (38) is lower than the first fault threshold (38).

6. A method as claimed in any preceding claim comprising further steps to:

a. compare the current residual (R) to a third fault threshold (40); and

b. declare a third fault (34) if the current residual (R) exceeds the third fault threshold (40).

7. A method as claimed in claim 6 wherein the third fault (34) comprises an open switch fault.

8. A method as claimed in claim 6 or claim 7 wherein the third fault threshold (40) is lower than the second fault threshold (38) and/or is lower than the first fault threshold (36).

9. A method as claimed in any preceding claim wherein the electrical device comprises any one of an inverter; a motor drive; a rectifier; a converter.

10. A method as claimed in any preceding claim wherein the steps of the method are iterated.

11. A method as claimed in claim 10 wherein the first fault (30) is declared when the current residual (R) exceeds the first fault threshold (36) on consecutive iterations.

12. A method as claimed in claim 10 when dependent on any of claims 3, 4 or 5 wherein the second fault (32) is declared when the current residual (R) exceeds the second fault threshold (38) on consecutive iterations.

13. A method as claimed in claim 10 when dependent on any of claims 6, 7 or 8 wherein the third fault (34) is declared when the current residual (R) exceeds the third fault threshold (40) on consecutive iterations.

14. A method as claimed in any preceding claim comprising further steps to:

a. calculate a phase residual (P) by dividing the difference of the RMS currents (la, lb) by their sum;

b. compare the phase residual (P) to a negative threshold and declare a fault on the first phase (a) if the negative threshold exceeds the phase residual (P); and

c. compare the phase residual (P) to a positive threshold and declare a fault on the second phase (b) if the phase residual (P) exceeds the positive threshold.

15. A method of the kind set forth substantially as described herein with reference to and as illustrated in the accompanying drawings.

Intellectual

Property

Office

Application No: GB1618973.0 Examiner: Mr Shane Keane