WO2005062456A1

WO2005062456A1 - Detecting a failure in a converter-load system

Info

Publication number: WO2005062456A1
Application number: PCT/EP2004/013287
Authority: WO
Inventors: Sven Gebhardt
Original assignee: Bombardier Transportation Gmbh
Priority date: 2003-12-20
Filing date: 2004-11-19
Publication date: 2005-07-07
Also published as: DE112004002512B4; GB2409355A; DE112004002512T5; GB0329543D0

Abstract

The invention that relates to a method of detecting a failure in a system which comprises a converter and a load, in particular a driving motor of a railroad traction vehicle, wherein the load is connected to the converter via a line, at least one electric current carried by the line is measured, the converter is controlled by a controlling device and the controlling device uses values of the measured current for operating the system. It is proposed to determine a local minimum value (MIN) and a local maximum value (MAX) of the measured current (I). The local minimum value (MIN) and the local maximum value (MAX) are used to decide whether there is a failure of the system.

Description

Detecting a Failure in a Converter-Load System

Description

The present invention relates to a method for detecting a failure in a system which comprises a converter and a load, in particular a driving motor of a railroad traction vehicle, wherein the load is connected to the converter via a. line, in particular an alternating current line having three phases. Furthermore, the present invention relates to a corresponding arrangement. In particular, the present invention relates to the field of electric high power applications, such as the delivery of electric energy to a driving motor of a railroad traction vehicle.

For many applications, a converter comprises a direct current side and an alternating current side with three alternating current phases. The direct current side is connected to a direct current intermediate circuit. The electric load is connected to the three alternating current phases on the alternating current side. A controlling device controls the operation of the converter, wherein the phase currents of the three alternating current phases are used as input variables of the controlling process. For example, it is possible to use a controlling device, which is adapted to generate pulse width modulation (PWM) signals. Switching signals, which correspond to the PWM signals, are transferred to the converter and effect the switching of electronic valves of the converter. However, other methods may be performed for controlling the converter. Alternatively, the converter may be a DC/DC converter, for example, or another type of converter.

There are several errors and malfunctions that might occur during the operation of the system: One or more of the three phases of the alternating current connection could be interrupted or the converter itself might malfunction. One or more of the three phases might be connected to ground. Furthermore, a sensor, used to detect a phase current of one of the phases might be defect and/or an energy supply for operating the sensor could fail. A signal connection from the detector to the controlling device could be interrupted or not realised (possibly due to a loose connector). In addition, or alternatively, any device (such as an amplifier), which is used to process the signal from the detector could be defect.

If measuring the phase currents malfunctions (including a malfunction of processing the measuring signal), the controlling device will use a wrong input value and, as a consequence, the controlling device will try to adapt the phase current. In other words: the controlling device will output control signals to the converter in order to compensate the wrong input value. If the measured phase current is permanently too small (e. g. zero due an interruption of the line or of the measurement processing equipment), the converter and the load can be damaged or even destroyed.

It is an object of the present invention to provide a method for detecting a failure in a system that comprises a converter and a load, so that the failure can reliably be detected and so that the detection is sufficiently fast to avoid damage of the system or of parts of the system. Furthermore, the detection should be low in effort and should not interfere with the normal operation of the system. A further object of the invention is to provide a corresponding arrangement.

Converters usually comprise a plurality of electronic valves which are switched on and off by the controlling device in order to operate the converter. The measured current, which is a function of time, therefore comprises a plurality of local minima and a plurality of local maxima due to the switching actions. For example, in the case of an inverter which is connected to a direct current intermediate circuit, the voltage of the intermediate circuit and any inductivity that is involved in the process effect an increase or a decrease of the line current, when electronic valves are switched on or off during normal operation of the converter. However, if there is an interruption, for example at the line, at the load or at the intermediate circuit, or if there is a failure of the current measurement equipment, an increase or decrease does not take place or is significantly smaller compared to its level during normal operation.

It is proposed to evaluate at least one local minimum value and at least one local maximum value and to detect a failure of the system using these local extreme values. In particular, the following is proposed: A method for detecting a failure in a system which comprises a converter and a load, in particular a driving motor of a railroad traction vehicle, wherein - the load is connected to the converter via a line, in particular an alternating current line having three phases, at least one electric current carried by the line is measured, the converter is controlled by a controlling device, the controlling device uses values of the measured current for operating the system, - a local minimum value and a local maximum value of the measured current are determined and the local minimum value and the local maximum value are used to decide whether there is a failure of the system.

For example, it is possible to measure the current carried by the line using a current sensor. This can be attached to the line, to the converter and/or to the load. In other words: it is not necessary to measure the current directly at the line.

It is an advantage of the proposed solution that an interruption can reliably be detected and that the detection can be performed rapidly, in particular much more rapidly than a detection performed at a basic frequency of an alternating current of the line. For example, it can be prevented that the controlling device effects an abrupt change in the operation of the converter due to a freezed (i.e. wrong and constant), missing (or zero) measurement signal of the line current.

Only one local minimum and one local maximum is needed for the evaluation. However, it is possible to increase the robustness of the evaluation by taking more extremes into account, for example by repeated evaluation of local extremes and/or by evaluating differences between more than two extremes. Also a faster detection by monitoring the minimum expected dl/dt (time derivative of the current I) between two samples/control cycles is possible, although this faster detection maybe more sensitive (e.g. due to noise on an interrupted line) and depends on the phase angles and the operating point of the system. In any case, the detection can be performed faster than a detection which is performed once in a period of an alternating current carried by the line. According to one embodiment, a difference between the minimum value and the maximum value is compared to a threshold value in order to detect whether there is a failure. The threshold value may be constant during the whole operation of the system. However, it is preferred to adapt the threshold value to the operating state of the system, to adapt it according to the control method currently used for the converter and/or to adapt it according to other criteria. For example if the power of the load (and a corresponding electric voltage) varies during operation, the typical differences between the local minimum and the local maximum vary correspondingly, so that the threshold value may be increased for higher powers and vice versa. As a result, the reliability of detection is not dependent on the instantaneous value of the electric load power.

Alternatively or in addition, a local increase or decrease of the measured current is calculated, wherein the increase or decrease is used to decide whether there is a failure of the system. In particular, the increase or decrease may be calculated using the difference between the local maximum and the local minimum and using information about a time interval between the minimum and the maximum or about another time interval.

For example, the time of the maximum and the time of the minimum may be known from the process of controlling the converter. More particularly, the controlling device may output control signals for switching electronic valves of the converter which result in the maximum or the minimum. However, in any case, the time interval may alternatively be a time interval of fixed length for repeatedly evaluating local minima and local maxima.

Generally speaking, the evaluation of the local minimum and of the local maximum and/or a sampling of the measured line current may be synchronised with or triggered by the switching actions which are controlled by the controlling device during operation of the converter. For example, time intervals may start at every local minimum or maximum, so that there is one of the local minima and one of the local maxima in each time interval. The evaluation may then be performed for each time interval. Alternatively, the evaluation may be performed for time intervals of constant length, which contain at least one of the maxima and one of the minima.

Speaking more generally, the converter may comprise a plurality of electronic valves which are switched on and off by the controlling device in order to operate the converter, so that the measured current comprises a plurality of local minima and a plurality of local maxima. A maximum value and a minimum value are repeatedly determined and an evaluation of the maximum value and the minimum value is repeatedly performed for time intervals comprising at least one of the local minima and comprising at least one of the local maxima.

The evaluation may be performed by using at least one of the maxima and at least one of the minima as well as a criterion for detecting the failure. However, according to a further embodiment of the invention, a model calculation of an operation of the load is performed, wherein the minimum value, the maximum value and the model calculation are used to decide whether there is a failure of the system.

In one example, the maximum value and the minimum value are input values of the model calculation. In this case, an output value of the model calculation may be compared to a known or predetermined parameter of the system or of the load. For example, a temporary increase or decrease of the line current may be calculated in the model calculation using the minimum value and the maximum value. Then, a value for an inductivity of the load (for example the inductivity of an electromagnetic coil of the load) is calculated using a further input value (which may be measured directly or indirectly), for example the electric voltage which corresponds to the line current, and the result is compared to a comparison value of the inductivity (e.g. the nominal inductivity value or a value which has been measured or calculated before). One advantage of such a model calculation is that the operating state (in the example the line voltage) is taken into account.

In another case, the model may be a model used by the controlling device for controlling the operation of the system. Such models are typically used in practice, particularly when the load is an asynchronous machine. For example, the controlling device may model an operation behaviour of the load using a model that is implemented by software. The measured phase currents and further input quantities (such as a voltage and the rotational speed of the machine), as well as parameters of the machine are used in the model. For example, DE 195 31 771 Al (inventor: Depenbrock) describes a method and a device for determining a rotational speed of a rotating field machine. The document discloses that signal processing comprises a complete machine model including a converter, an alternating current side of which is connected to the machine. In particular, such a model can be used to calculate a quantity or value (e.g. the increase or decrease of the line current) which is compared to a corresponding quantity or value calculated using the at least one maximum and at least one minimum value.

The method according to the invention is particularly useful for systems, wherein the line comprises a plurality of (in particular three) alternating current phases. In this case, the method is preferably performed for each measured phase current. More generally speaking, the maximum value and the minimum value are determined for at least two of the phases of such a line.

In such a system, a further optional step may be performed: the measured current values of the at least two phases are repeatedly evaluated for equal points in time and it is decided that there is a failure of the system, if pairs of values of the measured currents do not differ for a plurality of the points in time. This step is based on the fact that the phase currents of different phases are different due to the phase shift, except at the two intersection points in each period. If there is no significant difference between two of the phase currents at a plurality of the points in time, there is a failure. This step may be performed to confirm a failure detection result which has been obtained according to the method of evaluating the maximum value and the minimum value.

When the evaluation of the maximum and the minimum value is performed, it may include a special procedure for deciding whether a failure and/or malfunction exists. This special procedure can be performed in order to increase the reliability of making the decision and is referred to as "Fuzzy-evaluation" in this description.

When the evaluation is performed, it is proposed to repeatedly determine a first plausibility value, wherein the first plausibility value is an instantaneous measure of a degree of plausibility and/or non-plausibility. Further, a second plausibility value is derived from a plurality of the first plausibility values, wherein each of the first plausibility values can influence the second plausibility value correspondingly to its degree of plausibility and/or non-plausibility. An appropriate action (such as switching the converter to zero voltage) is taken, if the second plausibility value fulfils a predetermined criterion.

In the Fuzzy-evaluation the second plausibility value reflects the plausibility of a plurality of first plausibility values and, therefore, a faulty first plausibility value is less likely to result in an interruption of the converter operation, for example. The effect of the Fuzzy-evaluation can be compared with the effect of an intelligent filter. Furthermore, it can easily be implemented in hardware and/or software, for example by using a counter, wherein the counter value is the second plausibility value. The term "counter" is not limited to counting integer numbers. In the example, a first plausibility value with a lower degree of plausibility results in a greater increase of the counter number than a first plausibility value with a higher degree of plausibility. Nice versa, the counter number can be decreased, if the degree of plausibility of the first plausibility value is high.

There are several features which can be combined with this basic approach, independently from each other or in combination of at least some of the features. First, the degree of plausibility of the first plausibility value can be weighted and the second plausibility value is influenced correspondingly to the weighted first plausibility value. In the simplest case, a weight factor of 1 can be used for all degrees of plausibility of the first plausibility value. However, it is also possible to decrease the weight of the first plausibility values according to an increase of the degree of plausibility.

Secondly, it is possible to decide whether the degree of plausibility of the first plausibility value matches a lower threshold value and/or is smaller than the threshold value. If this is the case, the second plausibility value is influenced to show the decreased plausibility. If this is not the case, the second plausibility value is not amended or is amended to show an increased total plausibility (e.g. the counter number is decreased).

Thirdly, it may be that only the effect of a limited number of most recent first plausibility values influences the second plausibility value. This means that first plausibility values, which belong to incidents that took place too long ago, do not affect the second plausibility value. Alternatively, if a counter is used for the second plausibility value, both decreasing and increasing the counter number is possible, depending on the degree of plausibility of the first plausibility value. As a result, the counter number may be decreased to zero and, in this case, incidents that took place in the past do no longer affect the second plausibility value.

In particular, the first plausibility value is the result of a comparison (e.g. a deviation) between a threshold value and the difference between the local mimmum value and the local maximum value of the line current. In addition or alternatively, the "degree of plausibility" means that a higher deviation to a normal, an expected or a desired result has a lower degree of plausibility, for example. In the case of the result of the comparison being the first plausibility value, the degree of plausibility is lower for higher deviations. For example, the predetermined criterion may be that a threshold value of the second plausibility value is matched and/or exceeded. Generally, the Fuzzy-evaluation or one of its embodiments has the following advantages:

Small and/or rare deviations to a normal, an expected or a desired state do not necessarily result in an interruption of the current converter operation.

Large and/or many deviations to a normal, an expected or a desired state will result in a fast and reliable detection of the failure and/or malfunction.

The procedure can easily be implemented. The parameters of the procedure, like the weight and the threshold values, can easily be adapted.

- A process which might work at a small repetition frequency, but might be part of a higher-level control routine, can reliably check the result of the Fuzzy-evaluation, in particular check the counter number.

Furthermore, the present invention includes: a computer loadable data structure that is adapted to perform the method according to one of the embodiments described in this description, while being executed on a computer, - a computer program, wherein the computer program is adapted to perform the method according to one of the embodiments described in this description, while being executed on a computer, a computer program comprising program means for performing the method according to one of the embodiments described in this description while the computer program is being executed on a computer or on a computer network, a computer program comprising program means according to the preceding item, wherein the program means are stored on a storage medium readable to a computer, a storage medium, wherein a data structure is stored on the storage medium and wherein the data structure is adapted to perform the method according to one of the embodiments described in this description after having been loaded into a main and/or working storage of a computer or of a computer network, and a computer program product having program code means, wherein the program code means can be stored or are stored on a storage medium, for performing the method according to one of the embodiments described in this description, if the program code means are executed on a computer or on a computer network.

In particular, the computer or part of the computer network mentioned in one of the paragraphs before may be realised by using a processing unit, in particular a central processing unit of the controlling device. At least the determination of the minimum and the maximum value and the evaluation can be performed by the computer or computer network. The computer program product is to be understood as a product, which is a sellable or tradable good.

Furthermore, it is proposed to provide an arrangement which is adapted to perform the method according to at least one of the embodiments described before. The arrangement may be part of the converter, in particular part of a controlling device of the converter. It may be realised by hardware and/or software.

In particular, the following is proposed: An arrangement for detecting a failure in a system which comprises a converter and a load, in particular a driving motor of a railroad traction vehicle, wherein the converter is controlled by a controlling device, wherein the load is connected to the converter via a line, in particular an alternating current line having three phases, and wherein the arrangement comprises: - a measurement processing device for receiving and processing measured values of at least one electric current carried by the line, an extreme-determining device which is connected to the measurement processing device and which is adapted to determine at least one local minimum value and at least one local maximum value of the measured current, - an evaluating device which is connected to the extreme-determining device and which is adapted to decide whether there is a failure of the system by evaluating the at least one the minimum and the at least one maximum value. Specific embodiments of the present invention will be described in the following by way of example and with reference to the accompanying drawing. The figures of the drawing schematically show:

Fig. 1 an arrangement with a converter, an alternating current side of which is connected to an electric load, in particular to an asynchronous motor; Fig. 2 devices and units used to process a sensor signal of a sensor shown in Fig. 1; Fig. 3 an arrangement of units and devices for detecting a failure; Fig. 4 a flow chart illustrating an evaluation of minimum and maximum values; Fig. 5 a flow chart illustrating the Fuzzy-evaluation;

Fig. 6 a diagram illustrating the behaviour of a line current as a function of time;

Fig. 7 a diagram illustrating a first method of obtaining maximum and minimum values of a line current; and Fig. 8 a diagram illustrating a second method of obtaining maximum and minimum values of a line current.

Fig. 1 shows a system 1 comprising a converter 11 and a load 13, in particular an asynchronous machine. The converter 11 may be constructed as known from prior art. For example, the converter may be a DC/ AC converter (inverter) and may comprise three parallel paths, which connect a first and a second DC connection line of a DC intermediate circuit (not shown in the figure). Each of the paths may comprise one bridge having two electronic valves (for example Gate Turn-Off thyristors or Insulated Gate Bipolar Transistors, IGBTs), which are connected in series to each other. Each phase 5a, 5b, 5c of an AC connection to a machine may be connected to a connecting point between the two electronic valves of one path.

Input signal lines of the electronic valves for receiving switching signals can be connected to a control signal input 10 of the converter. The control signal input 10 is connected to a control signal output of a controlling device 4 via a connection 9 of the arrangement shown in Fig. 1. The DC side of the converter may be connected to a second converter, which is adapted to output a DC current to the converter. An input side of the second converter may be connected to a power supply network, for example a single-phase alternating current network of a railway system. However, other configurations and operations are possible, such as feeding back electric energy from the converter to the second converter. As shown in Fig. 1 the three phases 5 a, 5b, 5 c of the AC connection connect the converter 11 to the electric load 13, here to the asynchronous motor, which may be the driving motor of a railway traction vehicle.

A first 5a and a second 5b of the phases 5a, 5b, 5c are combined, in each case with only one current sensor 3a, 3b for measuring the phase current. Since the electric load 13 is balanced with respect to the three phases 5 a, 5b, 5c, the third phase 5 c is not combined with a current sensor. The current sensors 3 a, 3b are connected to the controlling device 4 via in each case one sensor signal connection 7a, 7b.

In high-power applications, such as in railway traction vehicles, the current sensor function may be based on the principle of detecting a current by evaluating the magnetic field produced by the current. An example of such a current sensor 3 and a corresponding arrangement for processing the sensor signal is shown in Fig. 2. The current sensor 3 is attached to one phase 5 and generates a current signal which corresponds to the current carried by the phase 5. A signal line for outputting the sensor signal is connected to a current/voltage converter 41 (e.g. a shunt resistor) for converting the current signal to a voltage signal. The current/voltage converter 41 is optionally connected to a filter 43 for filtering the voltage signal in order to eliminate signal parts, which are the result of transient interferences, and in order to eliminate noise. An output of the filter 43 is connected to an input of an amplifier 45 for amplifying the voltage signal. If the filter 43 is not provided, the output of the current/voltage converter 41 may directly be connected to the amplifier 45.

An output of the amplifier 45 is connected to an analogue to digital (A/D) converter 47 for digitising the voltage signal. The digitised signal can be used for digital data processing, in particular performed by a computer of the controlling device 4 shown in Fig. 1. For example, the devices 41, 43, 45, 47 of the arrangement shown in Fig. 2 may be arranged between the current sensor 3 a or 3b and a signal input of the controlling device 4. Alternatively, at least some of the devices 41, 43, 45, 47 may be part of the controlling device 4.

During operation of the converter 11, the controlling device 4 for controlling the converter 11 uses phase current values, which are based on the measurement performed by the current sensors 3a, 3b. For example, model calculations for modelling the operation of the electric load 13 and the evaluation of the minimum and maximum values can be performed by a central processing unit of the controlling device 4.

The arrangement shown in Fig. 3 comprises a measurement processing device 51 for receiving values of the measured phase current from the A/D converter 47 and for processing the measured values. On an input side, the measurement processing device 51 is connected to the A/D converter 47. On an output side, it is connected to an extreme-determining device 53 for determining local minimum values and a local maximum values of the measured current, The extreme-determining device 53 is connected to an evaluating device 55 which is adapted to decide whether there is a failure of the system by evaluating the minimum and the maximum values. The devices 51, 53, 55 may be part of a device or unit 50 and can be realised by software and/or hardware. The unit 50 may be part of the controlling device 4 of Fig. 1.

During operation, the A D converter 47 converts the analogue signals of the measured current to digital signals at a sampling rate which is greater (in particular at least by a factor of 5, preferably at least by a factor of 10) than an average switching rate of the electronic valves of the converter 11. Typically, an asynchronous motor which is used as a driving motor of a railway traction vehicle rotates at a rotor frequency of 0 to 150 Hz or higher (which is, for example, nearly equal to a basic frequency of the line current), the switching frequency of the converter is in the range of 250 Hz to 800 Hz and the sampling frequency of sampling the measured phase current values is even higher, namely in the range of some kHz, e.g. 2 to 4 kHz. The average switching rate may, for example, be the rate of pulse width modulation signals generated by the controller 4 in order to control the switching of the converter 11. Consequently, it is possible to reliably detect local mimmum values and local maximum values of the measured current which are caused by the switching actions (see Figures 6 to 8 for a typical behaviour of the line or phase current as a function of time).

The sampled values are transferred to the measurement processing device 51 which may perform several tasks such as preparing or operating the values for a model calculation

(simulation) of the system's operation behaviour. At least, it receives the sampled values and transfers them to the extreme-determining device 53 which determines the local minimum and maximum values of the current. Corresponding information (such as the current values of at least one minimum and at least one maximum per operation cycle of the device 50 or per time interval of evaluation) is then transferred to the evaluating device 55 which performs the evaluation of the extremes, as will be described in more detail by way of example in the following.

According to the flow chart shown in Fig. 4 at least one local minimum and at least one local maximum of the measured current is determined in step S10. In step SI 1 an increase or decrease of the current is calculated using the at least one local minimum and at least one local maximum and using time information, such as the length of a constant time interval of evaluation, or of a time of switching the converter's valves. An inductivity of the load is calculated in step S12 using the increase or decrease, as well as further information. In step S 13 the inductivity is compared to a comparison value.

Alternatively, instead of calculating the inductivity, the difference of the local maximum (or of the highest local maximum within the evaluation interval) to the local minimum (or of the lowest local minimum within the evaluation interval) can be calculated in step S 11 and can be compared to a threshold value in step SI 3. In this case step S12 can be omitted. According to a further alternative, step S12 can be omitted and the increase or decrease calculated in step SI 1 can be compared to a threshold value in step S13.

In the following step S 14 it is decided whether a deviation to the threshold value indicates a failure of the system. If so, an appropriate action is taken in step SI 5. If not, the procedure returns to step S10 after resetting a stored maximum and a stored minimum value to the last sampled value. Then the procedure is repeated for the next time interval or for other extremes. By resetting the stored maximum and the stored minimum value to the same (last sampled) value the difference is set to zero.

It is preferred to perform the procedure continuously for consecutive time intervals (e.g. for every switching cycle of switching the electronic valves connected to the line or phase on and off). However, the procedure may be performed discontinuously as well. Preferably, it is performed for each phase current which is measured in the system. It is possible to perform the procedure subsequently for the phases (saving time and resources), or to perform the procedures for the different phases at the same time or in overlapping time intervals. An exemplary way of performing the Fuzzy-evaluation is described in the following with reference to Fig. 5. In step SI, the difference between a maximum and a minimum value is calculated. In step SI a, a corresponding comparison value of the length is obtained and/or processed. The calculated value and the comparison value are transferred to a comparator, which compares the two values in step S2. In particular, the comparator calculates the difference between the two values. The difference is compared in step S3 with a threshold value, e.g. by threshold comparing means. If the threshold value is exceeded, the procedure continues with step S4, wherein a counter value is increased by the amount by which the difference exceeds the threshold value. If the threshold value is not exceeded, the procedure continues with step S5, wherein the counter value is decreased by a constant amount. After step S5, the procedure returns to the beginning and continues with step SI.

After step S4, the procedure continues with step S6, wherein a decision is made whether the counter value matches a second threshold value or whether it matches or exceeds the second threshold value. If this is the case, it is decided that there is a failure, in particular an interruption, and an appropriate action is taken in step S7. If the second threshold value is not exceeded, the procedure returns to the beginning and continues with step S 1.

The procedure described above can be implemented by software and/or hardware. Furthermore, it is possible to modify the procedure. For example, the counter value may be decreased in step S 5 by an amount that depends on the operating state of the system or of the load.

Fig. 6 shows a current carried by a phase of the line as a function of time t. The continuously decreasing and - in the right half of the diagram - increasing line corresponds to a desired or effective current ISET- The actual phase current I comprises the zigzag or saw-tooth- like behaviour shown in the figure, wherein the falling flank of each tooth intersects the continuous line at half the difference between the corresponding maximum and minimum. The intersection times are indicated by thin vertical lines which cut the pulse signals A (shown in Fig. 6) in two equal halves. The pulse signals correspond to pulse width modulation signals, generated by the controlling device in order to control the operation of the converter. At the rising edge of the signals A, the switching of the valves of the converter results in a local maximum of the current I. At the falling edge, the current I comprises a local minimum.

Shortly before the minimum of the effective current ISET_> the time interval between time ti and time t , the current I does not comprise a tooth with a significant peak or maximum value, since the pulse width is too small. Therefore, any evaluation procedure should not detect a failure in this interval. For example, the evaluation procedure may be interrupted, if the pulse width is smaller than a threshold value, and/or the Fuzzy-evaluation may be applied.

In Fig. 7, a first procedure of detecting local minima and local maxima of the current I is illustrated for two cases, the normal operation of the system (at the top of the figure) and the situation of a system failure (at the bottom of the figure, e.g. the current I is zero and/or there is noise). According to the first procedure, the time of measuring the extremes is synchronised with the action of switching the valves. Consequently, an extreme value is measured or determined for each of the switching times. Pairs of- in each case - one minimum MUST and one maximum MAX (in Fig. 7 located at or between two vertical lines) are then evaluated. As can be seen by comparing the two cases (top and bottom curve) the difference between the extremes of the pair reliably indicates the situation (failure or not).

In Fig. 8 there is a fixed length of time intervals in which the highest maximum value MAX and the lowest minimum value MIN is detected. The evaluation procedure for the corresponding pair of extremes may be the same as for the procedure of Fig. 7. Preferably, the length of the time interval is chosen in such a way that at least one minimum and one maximum lies within each time interval, independent from the operating state. In the example of pulse width modulation, the widest possible pulse width defines the smallest possible length of the time interval.

Claims

1. A method for detecting a failure in a system (1) which comprises a converter (11) and a load (13), in particular a driving motor of a railroad traction vehicle, wherein the load (13) is connected to the converter (11) via a line (5), in particular an alternating current line having three phases (5a, 5b, 5c), at least one electric current carried by the line (5) is measured, the converter (11) is controlled by a controlling device (4), the controlling device (4) uses values of the measured current for operating the system (1), a local minimum value and a local maximum value of the measured current are determined and the local minimum value and the local maximum value are used to decide whether there is a failure of the system.

2. The method of claim 1, wherein a difference between the mimmum value and the maximum value is compared to a threshold value.

3. The method of claim 2, wherein the threshold value is adapted in dependence on an operating state of the system and/or in dependence on a method of controlling the operation of the converter (11).

4. The method of one of claims 1 to 3, wherein a local increase or decrease of the measured current is calculated and wherein the increase or decrease is used to decide whether there is a failure of the system.

5. The method of one of claims 1 to 4, wherein a model calculation of an operation of the load is performed and wherein the minimum value, the maximum value and the model calculation are used to decide whether there is a failure of the system.

6. The method of the preceding claim, wherein the maximum value and the minimum value are input values of the model calculation.

7. The method of one of claims 1 to 6, wherein the converter (11) comprises a plurality of electronic valves which are switched on and off by the controlling device (4) in order to operate the converter, so that the measured current comprises a plurality of local minima and a plurality of local maxima, wherein a maximum value and a minimum value are repeatedly determined and wherein an evaluation of the maximum value and the minimum value is repeatedly performed for time intervals comprising at least one of the local minima and comprising at least one of the local maxima.

8. The method of one of claims 1 to 6, wherein the line (5) comprises a plurality of alternating current phases (5a, 5b, 5c) and wherein the maximum value and the minimum value are determined for at least two of the phases (5a, 5b, 5c).

9. The method of the preceding claim, wherein the measured current values of the at least two phases (5a, 5b, 5c) are repeatedly evaluated for equal points in time and wherein it is decided that there is a failure of the system, if pairs of values of the measured currents do not differ for a plurality of the points in time.

10. The method of one of claims 1 to 9, wherein, when the local minimum value and the local maximum value are used to decide whether there is a failure of the system, a first plausibility value is repeatedly determined, wherein the first plausibility value is an instantaneous measure of a degree of plausibility and/or non-plausibility, wherein a second plausibility value is derived from a plurality of the first plausibility values, wherein each of the first plausibility values can influence the second plausibility value correspondingly to its degree of plausibility and/or non-plausibility and wherein an action is taken, if the second plausibility value fulfils a predetermined criterion.

11. A computer loadable data structure, that is adapted to perform the method according to one of the preceding method claims while the data structure is being executed on a computer, in particular on a computer of a controlling device for controlling an operation of the converter (11).

12. A computer program, wherein the computer program is adapted to perform the method according to one of the preceding method claims while the computer program is being executed on a computer, in particular on a computer of a controlling device for controlling an operation of the converter (11).

13. A computer program comprising program means for performing the method according to one of the preceding method claims while the computer program is being executed on a computer, in particular on a computer of a controlling device for controlling an operation of the converter (11), or on a computer network.

14. A computer program comprising program means according to the preceding claim, wherein the program means are stored on a storage medium readable to a computer.

15. A storage medium, wherein a data structure is stored on the storage medium and wherein the data structure is adapted to perform the method according to one of the preceding method claims after having been loaded into a main and/or working storage of a computer, in particular on a computer of a controlling device for controlling an operation of the converter (11), or of a computer network.

16. A computer program product having program code means, wherein the program code means can be stored or are stored on a storage medium, for performing the method of one of the preceding method claims, if the program code means are executed on a computer, in particular on a computer of a controlling device for controlling an operation of the converter (11), or on a computer network.

17. An arrangement for detecting a failure in a system (1) which comprises a converter (11) and a load (13), in particular a driving motor of a railroad traction vehicle, wherein the converter (11) is controlled by a controlling device (4), wherein the load (13) is connected to the converter (11) via a line (5), in particular an alternating current line having three phases (5a, 5b, 5c), and wherein the arrangement comprises: - a measurement processing device (51) for receiving and processing measured values of at least one electric current carried by the line (5), - an extreme-determining device (53) which is connected to the measurement processing device (51) and which is adapted to determine at least one local minimum value and at least one local maximum value of the measured current, an evaluating device (55) which is connected to the extreme-determining device (53) and which is adapted to decide whether there is a failure of the system by evaluating the at least one the minimum and the at least one maximum value.

18. The converter (11) comprising the arrangement of the prec eding claim.