Frequency-selective digital active filter
Patent Information
- Application Number
- EP2023706598
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2022-02-25
- Filing Date
- 2023-02-21
- Publication Date
- 2025-01-01
Smart Images
Figure 1.1
Abstract
Description
[0001]FREQUENCY ELECTIVE DIGITAL ACTIVE FILTER Technical Field The present invention relates to the technical field of interference signal suppression in power electronic systems, such as periodically switched inverters. In particular, the invention relates to a method aimed at suppressing such interference signals, as well as a device for carrying out such a method. Background of the Invention Power electronic systems can generate significant electromagnetic interference signals due to the underlying high-frequency switching processes, which can, for example, negatively affect other systems located in the vicinity of the system. Such interference signals are counteracted in the prior art by passive filters and shielding housings, which, however, are large,are heavy and expensive. Active filters (active EMI filters) have already been established to reduce the size of these components. In this process, analog or digital circuits are introduced that absorb the interference, transform it, and inject it back into the system. The aim is to achieve destructive interference between the interference and the counter-interference in order to reduce the interference emission. Previous active filters use analog circuit technology in controlled or regulated structures (feedforward and feedback types). The counter-interference signal is obtained directly from a measured quantity. The achievable interference suppression with these approaches is limited due to the time constants of the sensors, injectors, and amplifiers, due to delays and attenuations due to signal propagation times, and due to limited,frequency-dependent gain factors. The state of the art also proposes the use of synthesized counter-interference signals. If the interference is periodic over a sufficient period of time, stable harmonics and possibly sideband harmonics will occur in the interference signal in the frequency domain. To suppress these harmonics and sideband harmonics, a sinusoidal signal can be generated, the amplitude and phase of which are adjusted to achieve complete destructive interference. Injection delays can be compensated for by phase shifts. Compared to active filters, delayed signal paths can thus be systematically compensated.This can significantly improve the effectiveness of interference suppression. By superimposing the detected sinusoidal signals, a broadband counter-interference signal can be generated. The document DE 102018001051 A1 describes a method for reducing an electromagnetic interference signal from an interference source generated by an electronic system controlled at a frequency by active negative feedback. For this purpose, for any selection of harmonics of the clocked control, comprising at least a single harmonic, sinusoidal signals are synthesized at the frequency of the respective harmonic. Their amplitudes and phase positions are adjusted so thatthat by superimposing a counter-interference signal formed by the sinusoidal signals with the interference signal at a selected reference measuring point, destructive interference results in a substantial cancellation of the respective frequency components of the electromagnetic interference signal. DE 102018008224 A1 discloses a device for stabilizing electrical networks by means of a system for influencing the electrical energy consumption of larger consumers, e.g., for charging electric vehicles, comprising a connection network, a balancing, active network, and an adjacent, passive network, with higher-frequency measuring devices being provided for all occurring voltages and currents. It is provided that the time-varying energy consumption in the balancing, active network is modified in such a way thatthat by combining the measured impedance of connected energy consumers, network perturbations are reduced or network disturbances are compensated, wherein a measuring and analysis device is provided for measuring and analyzing a temporal profile of voltages and currents of a supply in the connection network and for characterizing the surrounding networks, wherein a computing unit is provided for determining a correction signal, wherein a modification device is provided for modifying a control signal for a power controller, wherein a power controller is provided which connects a power supply to an energy consumer, and wherein a bidirectional transmission device is provided for transmitting target parameters and fault messages,which is connected to a mains connection and is bidirectionally connected to the arithmetic unit. Description of the invention Based on the described prior art, it is an object of the present invention to further improve known devices and methods for reducing an electromagnetic interference signal. This object is achieved by the features of the independent main claims. Advantageous embodiments are specified in the subclaims. If technically possible, the teachings of the subclaims can be combined as desired with the teachings of the main and subclaims. According to a first aspect of the invention, the object is achieved by a method for suppressing an electromagnetic interference signal in a power electronic system by means of a digital counter-interference system, wherein a transmission path exists between the counter-interference system and the power electronic system,The method comprises the steps of a) detecting an interference signal present at the counter-interference system, b) identifying several interference-relevant frequencies in the detected interference signal, c) determining a transfer function of the transmission path for each interference-relevant frequency, d) synthesizing a narrowband counter-interference signal for each interference-relevant frequency, e) determining the influence of the counter-interference signal on the interference signal based on the transfer function for each interference-relevant frequency, wherein the synthesis of the narrowband counter-interference signals is carried out in a cascade, wherein the synthesis of a respective narrowband counter-interference signal is carried out based on the sum of the interference signal and the influences of all previously synthesized counter-interference signals on the interference signal, and wherein the counter-interference signals are summed to form an overall counter-interference signal and coupled into the power electronic system. It is preferred that the sequence of method steps,unless technically required in an explicit order. However, the aforementioned order of the method steps is particularly preferred. Advantageous aspects of the claimed invention are explained below, and furthermore, preferred modified embodiments of the invention are described below. Explanations, in particular regarding advantages and definitions of features, are essentially descriptive and preferred, but not limiting, examples. If an explanation is limiting, this is expressly stated. Power electronic systems can, for example, be inverters or comprise such inverters. Such systems have, for example, high-frequency switching line semiconductors that generate interference signals, with an interference signal being particularly pronounced at the clock frequency and at harmonics of the clock frequency. If such power electronic systems are periodically clocked,For example, by means of pulse width modulation (PWM), in particular by means of sine pulse width modulation (SPWM), or have power factor correction filters, sideband harmonics also occur in addition to the clock frequency and the harmonics. A transmission path is understood to be a circuit of components that is designed, in particular, for reciprocal signal transmission between the power electronic system and the counter-interference system. In particular, such a transmission path has a first sub-path for detecting an interference signal at the power electronic system, converting the detected interference signal, for example, in an analog-to-digital converter (A / D converter), and transmitting the signal to the counter-interference system or for coupling the signal to the counter-interference system. The transmission path then further has a second sub-path for transmitting a counter-interference signal from the counter-interference system to the power electronic system.For example, with means for converting the counter-interference signal, for example, in a digital-to-analog converter (D / A converter), and with means for coupling the counter-interference signal into the power electronic system. The components of the transmission path influence the amplitudes and phases of the signals during the transmission and / or conversion of the interference signal, as well as during the transmission and / or conversion of the counter-interference signal. This influence can be mathematically described by the transfer function. The transmission path, or the transfer function, also includes an influence that the counter-interference system itself has on the processed signals, as well as an influence that the power electronic system has on the processed signals.Whereby these influences can be assigned purely by definition to the first or second sub-path, to both sub-paths, or to none of the sub-paths. In particular, the transmission path changes the amplitude of the signals by a factor α and shifts the phase by an amount Δφ. Insofar as the influence of the counter-interference signal on the interference signal is determined using the transfer function for each interference-relevant frequency, the influence of the entire transmission path with both sub-paths is determined, i.e., the influence of the counter-interference signal emitted by the counter-interference system on the residual interference signal subsequently detected by the counter-interference system. Insofar as an interference signal is detected, it is detected in particular by means of an analog sensor,i.e. measured. The interference signal can also be detected directly or indirectly in other ways. In the case of an analog sensor, an A / D converter is preferably provided between the sensor and the counter-interference system in the transmission path. The interference-relevant frequencies are identified in particular with the aid of a Fourier analysis and means for their evaluation. In particular, the identification is carried out using the counter-interference system. A frequency is recognized as interference-relevant, for example, if an amplitude is present at this frequency that lies above a threshold value. For example, the frequency lies above a limit value from which it can be concluded that the amplitude can or could lead to interference with another, for example, neighboring system. It can also be providedthat only a certain number or a certain proportion of the interference-relevant frequencies are selected and taken into account in the further process. The determination of a transfer function can be carried out, for example, by estimation, by calculation, or based on a test signal using optimization methods. In particular, numerical calculation methods can be used to determine the transfer function. The transfer functions differ depending on the frequency and are therefore preferably determined separately for each identified interference-relevant frequency. Using the aforementioned method, it is advantageous to achieve an overall counter-interference signal that is particularly well-matched to the interference signal and suitable for suppressing an interference signal across a broad frequency band. In particular, individual frequencies of the interference signal are not excessively attenuated, i.e., beyond the amplitude of the interference signal.Because the individual counter-interference signals are generated depending on their respective influence on the power electronic system, the interference signal and the overall counter-interference signal cancel each other out as precisely as possible. When identifying the interference-relevant frequencies, the method is also completely independent of whether these interference-relevant frequencies are related to one another or whether these frequencies are more or less far apart in the frequency band. Identification is possible over any frequency band. In particular, the method can be used to cancel out a clock frequency, several harmonics related to the clock frequency, and several sideband harmonics over a wide bandwidth. The method is therefore suitable for power electronic systems controlled by pulse width modulation.Particularly well-suited for suppressing interference signals. Furthermore, by determining the influence of the counter-interference signal on the interference signal using the transfer function for each interference-relevant frequency, a precise prediction can be made as to how the counter-interference signal will reach the power electronic system, or how the interference signal influenced by the counter-interference signal will in turn reach the counter-interference system.so that the counter-interference signals synthesized downstream in the cascade can be synthesized with particular precision. The influence of the transmission path is thus continuously and predictively considered during the synthesis of the counter-interference signals. In other words, after the synthesis of a counter-interference signal, the influence of the counter-interference signal on the interference signal actually detected by the counter-interference system is determined. With increasingly accurate determination of the transfer function, an increasingly precise counter-interference signal is achieved for each interference-relevant frequency, or ultimately, an increasingly precise overall counter-interference signal. Each synthesized counter-interference signal is not actually coupled into the power electronics system to determine its influence on the interference signal; rather, this is done computationally based on the transfer function. In this respect, it can be referred to as a pseudo-cascade.from which only the result – i.e., the total counter-interference signal – is actually coupled into the power electronics system. In addition to the influence of the transfer function or the transfer path on a respective counter-interference signal, further influences can also be taken into account for each counter-interference signal or in relation to the total counter-interference signal, such as the speed of adaptation of the electrical components involved in the synthesis of the counter-interference signals. Preferably, the method is repeatedly carried out as a control method for minimizing the interference signal, comprising at least steps a), d), and e). The method is thus repeated in a loop, whereby a renewed identification of interference-relevant frequencies and the determination of a transfer function for each interference-relevant frequency can be dispensed with, at least as long as the interference-relevant frequencies do not change, which is particularly the case as long ashow the excitation signal of the power electronic system remains constant in its frequency response. Within the loop-like control process, the process adapts itself to changing amplitudes and the respective phase of the interference signal for each interference-relevant frequency and forms a counter-interference signal that interferes with the interference signal. If the synthesized total counter-interference signal does not yet completely cancel the interference signal during the control process, a residual interference is present at the input of the synthesis cascade, which leads to a correspondingly modified synthesis of the counter-interference signals.so that the overall counter-interference signal is ultimately adapted to completely cancel the interference signal. In a preferred embodiment, the transfer function of the transmission path comprises at least one amplitude change α and one phase shift Δφ for each interference-relevant frequency. In this way, the actual change in the interference signal in the power electronic system due to the respectively generated counter-interference signal, as well as the subsequent detection of the resulting interference signal by the counter-interference system, can be mathematically mapped with sufficient accuracy.to achieve a precise determination of the influence of the transmission path on the counter-interference signal and the interference signal. The influence of the signals through the transmission path, both from the counter-interference system to the power electronic system and from the power electronic system to the counter-interference system, is taken into account by the transfer function. Particularly preferably, the counter-interference system generates a harmonic test signal for each interference-relevant frequency to determine the transfer function of the transmission path, and the transfer function is determined based on the test signal of the interference-relevant frequency. The counter-interference system is configured toInstead of the complex broadband counter-interference signal, a harmonic test signal of exactly one frequency is generated simultaneously. For each interference-relevant frequency, the associated transfer function is then determined using a harmonic test signal of that frequency. For this purpose, the parameters of the transfer function are determined iteratively, for example, using the test signal, and adjusted so that a counter-interference signal dependent on the transfer function is achieved to completely and precisely cancel the test signal. The transfer function is preferably determined using a least-mean-square method (LMS method) based on the test signal, which represents a particularly stable and mathematically simple iterative procedure. For the iterative determination of the transfer function, the transfer function is preferably defined as an orthogonal system depending on weighting variables s0, s1, from which the parameters to be determined can then be calculated.transformed. Starting from a transfer function H(fk) at a disturbance-relevant frequency fk, which can be represented as follows (, 1) Using the addition theorem, the signal can be represented as a superposition of weighted cosine and sine signals. ( 2) The weighting variables s0, s1 are then (3) (4) The amplitude change α and the phase shift ∆φ = 2πf κ Δt with Δt as the time shift when passing through the transmission path can then be calculated with known weighting variables s0, s1 as follows. (5) ( 6) The terms sin(2πfκt) and cos(2πfκt) form the aforementioned orthogonal system due to their phase shift of 90° to each other. By iteratively optimizing the weighting variables s0, s1, in particular using a least-mean-square method, based on the test signal, the amplitude change α and the phase shift Δφ of the transfer function can be determined frequency-specifically. Such a method is explained in more detail, for example, in the following description of the exemplary embodiments illustrated in the figures. In a further preferred embodiment, the narrowband counter-interference signals are synthesized for each interference-relevant frequency using a least-mean-square method. For example, the synthesis is carried out using a filtered-x-least-mean-square method (FxLMS method), as explained in more detail in the following description of the exemplary embodiments illustrated in the figures.Particularly preferably, the synthesis of a respective narrowband counter-interference signal is adapted in such a way that the influence of the transmission path is already taken into account in its influence on the interference signal. The object is further achieved by a method for determining a transfer function of the transmission path at an interference-relevant frequency for a method according to one of the preceding claims, - wherein the transfer function comprises at least one amplitude change and a phase shift, and - wherein the counter-interference system generates a harmonic test signal for each interference-relevant frequency to determine the transfer function of the transmission path, and the transfer function is determined using a least-mean-square method based on the test signal of the interference-relevant frequency.Such a method has already been described above in connection with the method for suppressing an electromagnetic interference signal in a power electronic system and has the advantages already described. In particular, such a method can be used to determine a transfer function with sufficient accuracy to precisely determine the influence of a counter-interference signal on an interference signal based on the determined transfer function. Using a transfer function determined in this way for each interference-relevant frequency, an overall counter-interference signal that is very precisely tuned to the interference signal can then be achieved in a previously described method. The least mean square method is particularly preferably aimed at optimizing weighting variables s0, s1 of an orthogonal system describing the relationship between the amplitude changes α and the phase shift Δφ.The object is further achieved by a device for suppressing electromagnetic interference signals in a power electronics system, comprising a digital counter-interference system and a transmission path formed between the counter-interference system and the power electronics system, wherein the counter-interference system is designed to carry out a previously described method for suppressing an electromagnetic interference signal in a power electronics system. The device allows the advantages described above with regard to the method to be achieved accordingly. In particular, a particularly precise overall counter-interference signal can be generated and coupled into the power electronics system, by means of which the interference signal present in the power electronics system can be canceled out with the greatest possible precision.In one embodiment of the device, the transmission path has a first sub-path for transmitting a signal from the power electronic system to the counter-interference system and a second sub-path for transmitting a signal from the counter-interference system to the power electronic system. A transfer function is then described for the entire transmission path, i.e., for the sum of the first sub-path and the second sub-path. The transmission path or the transfer function further includes the influence of the signals by the counter-interference system itself and by the power electronic system. In one embodiment, the transmission path has at least one D / A converter, at least one injector for coupling the overall counter-interference signal into the power electronic system, at least one sensor for measuring the interference signal present in the power electronic system, and at least one A / D converter.The transmission path is therefore designed to convert an interference signal measured in analog form on the power electronics system into a digital signal and transmit it to the counter-interference system, and to convert a digital total counter-interference signal into an analog total counter-interference signal and couple it into the power electronics system. In a further embodiment, the counter-interference system is designed to determine the transfer function for each interference-relevant frequency using the harmonic test signal. In another preferred embodiment, the counter-interference system is designed as a field programmable gate array (FPGA), DSP, or microcontroller in conjunction with analog-to-digital and digital-to-analog converters. Brief description of the drawings The invention is explained in more detail below with reference to the attached drawings using preferred exemplary embodiments. The wording "figure" in the drawings is denoted by Fig.abbreviated. In the drawings: Fig. 1 shows a graphic representation of an excitation signal generated by means of pulse width modulation in a power electronic system; Fig. 2 shows a graphic representation of an interference signal of a power electronic system; Fig. 3 shows a schematic view of an apparatus for carrying out a method for suppressing an electromagnetic interference signal in a power electronic system according to one aspect of the invention; Fig. 4 shows a process diagram illustrating a method for suppressing an electromagnetic interference signal in a power electronic system according to one aspect of the invention; Fig. 5 shows a graphic representation of a residual interference spectrum resulting from a superposition of an interference signal and a total counter-interference signal generated according to Figure 4; Fig.6 shows a further graphical representation of a residual interference spectrum resulting from the superposition of an interference signal and a total counter-interference signal generated according to FIG. 4; FIG. 7 shows a process diagram illustrating a method for determining a transfer function of the transmission path; and FIG. 8 shows a process diagram illustrating a method for determining a transfer function of the transmission path and a modified method for suppressing an electromagnetic interference signal in a power electronic system at only a single frequency. Detailed Description of the Embodiments The described embodiments are merely examples that can be modified and / or supplemented in many ways within the scope of the claims.Each feature described for a specific embodiment can be used independently or in combination with other features in any other embodiment. Each feature described for an embodiment of a specific claim category can also be used correspondingly in an embodiment of a different claim category. Figure 1 shows, in a lower diagram, an excitation signal 1 for a power electronics system, which is generated by pulse width modulation shown in an upper diagram. During pulse width modulation, a sine signal 2 is superimposed with a sawtooth signal 3, wherein the comparison of the sine signal 2 and the sawtooth signal 3 is used to switch between two states that result in the excitation signal 1 formed from rectangular pulses.The superposition results in excitation signal 1 with different pulse widths, which repeat periodically at the frequency fref of the sinusoidal signal 2. The actual pulse frequency is the frequency f. PWM of the sawtooth signal 3. The excitation signal 2 shown in Figure 1 generates an interference signal 4 shown in Figure 2 over a broad frequency spectrum. The interference signal 4 has a pulse frequency f PWM and harmonics 2∙f PWM , 3∙f PWM , …, 10∙f PWM significant amplitude peaks 5, as shown in an upper diagram of Figure 2. A detailed view shown in a lower diagram of Figure 2 also shows that immediately adjacent to the pulse frequency f PWM and the harmonic 2∙f PWM , 3∙f PWM , …, 10∙f PWM in the case of sideband harmonics, significant amplitude peaks 6 are also formed in the interference signal 4, here using the example of the harmonic 10∙f PWMThe amplitude peaks 6 formed in the sideband harmonics arise from the pulse width modulation or the different pulse widths in the excitation signal 1. The amplitude is plotted logarithmically on the y-axis in both diagrams in Figure 2. Figure 3 shows a device 7 for carrying out a method for suppressing the above-described interference signal 4. The device 7 is arranged on a power electronics system 8, which is essentially formed by an inverter 10 connected to a load 11 and a direct current source 12. The device 7 is arranged between the inverter 10 and the direct current source 12 and has a sensor 14 for measuring the interference signal 4 present in the power electronics system 8. The device further comprises an A / D converter 15 by means of which the signal detected by the sensor 14 is converted in order to provide it to a digital counter-interference system 16.The sensor 14 and the A / D converter 15 form a first sub-path 17.1. The device 7 further comprises a D / A converter 18, by means of which a digital total counter-interference signal generated by the counter-interference system 16 is converted into an analog total counter-interference signal in order to couple it into the line electronics system 8 by means of an injector 19. The D / A converter 18 and the injector 19 together form a second sub-path 17.2. The first sub-path 17.1 and the second sub-path 17.2, as well as the counter-interference system 16 and the part of the power electronics system 8 located between the injector 19 and the sensor 14 together form a transmission path 17, through which signals in the device 7 pass between the counter-interference system 16 and the power electronics system 8. The transmission path 17 influences the signals passing through it in such a way that their amplitudes and phases are changed by amounts α, Δφ.Figure 4 shows a method for suppressing a previously described interference signal 4, for example by means of a device according to Figure 3. The interference signal 4 is present at an input node 20, which corresponds to the input of the digital counter-interference system 16, and is summed with a total counter-interference signal 21 to form a residual interference signal 4.1. At the beginning of the method, when no total counter-interference signal 21 has yet been generated, the residual interference signal 4.1 corresponds to the interference signal 4. Insofar as an ideal total counter-interference signal 21 has been generated, the residual interference signal 4.1 is largely or approximately zero. In the interference signal 4, or the residual interference signal 4.1, several interference-relevant frequencies are first identified in a step not shown in detail. In a further step, not shown in detail, but described below with reference to Fig. 7 and Fig.8, a predetermined transfer function of the transmission path is determined for each of these interference-relevant frequencies. In the part of the method illustrated in Figure 4, a first narrowband counter-interference signal is then generated for the residual interference signal 4.1 in a first synthesis 22.1 at a first of the interference-relevant frequencies, for example, using a least-mean-square method described below. Such a synthesis is carried out, for example, in a narrowband filter. Subsequently, in a first influence determination 23.1, the influence of this first counter-interference signal on the residual interference signal 4.1 is determined based on the transfer function for the relevant frequency. The determination is performed computationally, and at this point in time, no signal is transmitted to the power electronic system 8. The determined influence is then combined with the residual interference signal 4.1 at a first interference signal node 24.1 is summed. The thus modified residual interference signal 4.1 is then used as the basis for a second synthesis 23.2 of a second counter-interference signal at a second interference-relevant frequency. The method is continued in this way until the last interference-relevant frequency is reached at a final synthesis 22.n. The influence determination 23 is preferably carried out exactly m times and summed at m interference signal nodes with the previously modified residual interference signal 4.1, where m=n-1. Furthermore, individual influence determinations can also be omitted, so that several syntheses are based on the same residual interference signal 4.1 in order to achieve certain desired characteristics in the interference suppression. The counter-interference signals are summed at counter-interference signal nodes 25.1 to 25.n and ultimately result in a preliminary total counter-interference signal 21.1. The preliminary total counter-interference signal 21.1 is then coupled into the power electronic system 8 via the transmission path 17 and in turn transmitted to the counter-interference system 16. The part of the process illustrated in Figure 4 is repeated as a control process in loop form and can thus permanently and adaptively suppress an interference signal 4. From the above, it follows that both the interference signal 4 or residual interference signal 4.1, as well as the total counter-interference signal 21, differ at different points on the transmission path 17 depending on the influence of the transmission path 17. The interference signal 4 or residual interference signal 4.1 is therefore present at the sensor 14 with a different amplitude and a different phase than at the input of the counter-interference system 16, at the output of the counter-interference system 16, or the injector 19.The method described above aims to generate a total counter-interference signal 21 that is as accurate as possible to cancel out the interference signal 4 by determining this influence and taking it into account frequency-wise during the synthesis of counter-interference signals. For the sake of simplicity, when considering the total counter-interference signal 21 and interference signal 4, the point at which they are compared along the transmission path is not differentiated in more detail below. Figure 5 shows the effect of the total counter-interference signal 21 on the interference signal 4. In an upper diagram, an interference signal 4 corresponding to the representation in Figure 2, but with absolute frequency plotted on the x-axis in kHz, as well as a superposition 30 of the interference signal 4, which corresponds to the residual interference signal 4.1 after running through the process shown in Figure 4, with the total counter-interference signal 21.It can be seen that the amplitudes of the superposition 30 plotted in dBµV on the y-axis in Fig. 5 largely result in zero, both at the amplitude peak 5 and at the amplitude peaks 6, as well as in the vicinity of the amplitude peaks 5, 6. The lower diagram shows the associated frequency-dependent interference suppression 26, which has particularly high reductions around the interference-relevant frequencies f1, f2, f3. Figure 6 shows, in the same way as the upper diagram in Fig. 5, an interference signal 4 and a superposition 30 resulting from a total counter-interference signal 21 (not shown), with the frequency in kHz being plotted on the x-axis and the amplitude in dBµV on the y-axis. Here, the counter-interference signal is not generated for each interference-relevant frequency, but a synthesis 22.1, …, 22.n, as well as the influence determination 23.1, …, 23.m is only carried out for five frequencies marked by dashed lines.As can be seen, the cascading during the synthesis of the counter-interference signals results in a favorable superposition 30 with just five syntheses 22.1, ..., 22.n, in which the interference signal 4 can be largely suppressed. Fig. 7 shows an illustration of a method for determining a transfer function of the transmission path, as used for determining influences 23.1, ..., 23.m. In this case, a harmonic test signal 31 is generated by means of the counter-interference system 16, which simultaneously passes through the transmission path 17 and a model 32 of the transmission path 17 integrated in the counter-interference system 16. An interference signal 4 present in the power electronic system 8 is set to zero (d(t)=0), or the power electronic system 8 is decoupled from the counter-interference system 16.The test signal 31 is decomposed in the counter-interference system 16 into an orthogonal system consisting of the terms x0(n)=cos(2πfκtn) and x1(n)=sin(2πfκtn), where fk is the frequency of the harmonic test signal 31 and n is the number of steps of the iterative procedure. Weighting variables s0, s1 are then determined from these terms, which, when summed, form the transfer function. The transfer function is then compared with the actual transfer behavior of the transmission path, and the result ε(n) of this comparison is minimized using a least-mean-square method 33 by optimizing the weighting variables s0, s1.Figure 8 shows the overall functionality of the counter-interference system 16 in simplified form, wherein several switches 34 can be used to switch between a method described in Figure 7 for determining a transfer function of the transmission path 17 and a method for synthesizing a counter-interference signal. Here, the overall counter-interference signal 21 is only generated at one frequency and therefore not in a cascaded form. However, it becomes clear how the modeled influence of the transmission path 17 on the counter-interference signal is already taken into account during the synthesis of the counter-interference signal. As soon as the transfer function has been stabilized in the method described with reference to Figure 7, the switches 34 are flipped so that a single narrowband filter 35, in which the previously described synthesis 22.1, ..., 22.n also takes place, is now connected to the transmission path and the interference signal 4 is present in the power electronic system 8 (d(t)≠0).In a method step 36, the modeled weighting variables s0, s1 are transmitted to this narrowband filter 35, whereupon a counter-interference signal is synthesized using a filtered-×-least-mean-square method 37, incorporating the weighting variables s0, s1. The method shown in Figure 8 leads to the described cascaded method by replacing the narrowband filter 35 with several narrowband filters 35 for synthesizing 22.1, ..., 22.n a respective counter-interference signal for each interference-relevant frequency f1, f2, f3, with influence determinations 23.1, ..., 23.m being performed between each synthesis based on the transfer function. List of reference symbols 1 Excitation signal 2 Sinusoidal signal 3 Sawtooth signal 4 Interference signal 4.1 Residual interference signal 5 Amplitude peak 6 Amplitude peak 7 Device 8 Power electronic system 10 Inverter 11 Load 12 DC power source 14 Sensor 15 A / D converter 16 Counter-interference system 17 Transmission path 17.1 First partial path 17.2 second partial path 18 D / A converter 19 injector 20 input node 21 total interference signal 21.1 preliminary total interference signal 22.1 synthesis … 22.n synthesis 23.1 influence determination … 23.m influence determination 24.1 interference signal node … 24.m interference signal node 25.1 interference signal node … 25.n interference signal node 26 interference suppression 30 superposition 31 test signal 32 model of the transmission path 33 least mean square method 34 switch 35 narrowband filter 36 process step 37 filtered x least mean square method.
Claims
Claims 1. A method for suppressing an electromagnetic interference signal (4, 4.1) in a power electronic system (8) by means of a digital anti-interference system (16), wherein a transmission path (17) exists between the anti-interference system (16) and the power electronic system (8), the method comprising the steps a) detecting an interference signal (4, 4.1) present at the anti-interference system (16), b) identifying several interference-relevant frequencies (f1, f2, f3) in the detected interference signal (4, 4.1), c) determining a transfer function of the transmission path (17) for each interference-relevant frequency (f1, f2, f3), d) synthesizing (22.1, …, 22.n) a narrowband anti-interference signal for each interference-relevant frequency (f1, f2, f3), e) determining (23.1, …, 23.m) the influence of the counter-noise signal on the disturbance signal (4, 4.1) based on the transfer function for each disturbance-relevant frequency (f1, f2, f3), where the synthesis (22.1, …, 22.n) the narrowband anti-noise signals are cascaded, wherein the synthesis (22.1, …, 22.n) of each narrowband anti-noise signal is performed based on the sum of the noise signal (4, 4.1) and the effects of all previously synthesized anti-noise signals on the noise signal (4, 4.1), and wherein the anti-noise signals are summed to form a total anti-noise signal (21, 21.1) and coupled into the power electronic system (8).
2. Method according to claim 1, wherein the method is repeatedly performed as a control method to minimize the noise signal (4, 4.1), at least comprising steps a), d), and e).
3. A method according to claim 1 or 2, wherein the transfer function of the transmission path (17) comprises at least one amplitude change α and one phase shift Δφ for each interference-relevant frequency (f1, f2, f3).
4. A method according to any one of the preceding claims, wherein the counter-interference system (16) generates a harmonic test signal (31) for each interference-relevant frequency (f1, f2, f3) and the transfer function is determined from the test signal (31) of the interference-relevant frequency (f1, f2, f3).
5. A method according to claim 4, wherein the transfer function is determined from the test signal (31) using a least-mean-squares method (33).
6. The method of claim 5, wherein the least-mean-squares method (33) is directed to the optimization of weighting variables s0, s1 of an orthogonal system describing the relationship between the amplitude changes α and the phase shift Δφ. 7.A method according to any one of the preceding claims, wherein the synthesis (22.1, …, 22.n) of the narrowband counter-noise signals for each interference-relevant frequency (f1, f2, f3) is performed using a least-mean-squares method (37).
8. A method for determining a transfer function of the transmission path (17) at an interference-relevant frequency (f1, f2, f3) for a method according to any one of the preceding claims, wherein the transfer function comprises at least an amplitude change α and a phase shift Δφ, and wherein the counter-noise system (16) for determining the transfer function of the transmission path (17) for each interference-relevant frequency (f1, f2, f3) generates a harmonic test signal (31) and the transfer function is determined using a least-mean-squares method (33) based on the test signal (31) of the interference-relevant frequency (f1, f2, f3). 9.Method according to claim 8, wherein the Least-Mean-Squares method (33) is applied to the optimization of weighting variables s0, s1 of a relationship.
10. Device (7) for suppressing electromagnetic interference signals (4, 4.1) in a power electronic system (8), comprising a digital counter-interference system (16) and a transmission path (17) formed between the counter-interference system (16) and the power electronic system (8), wherein the counter-interference system (16) is configured to perform a method according to any one of claims 1 to 7.
11. Device (7) according to claim 10, wherein the transmission path (17) comprises a first sub-path (17.1) for transmitting a signal from the power electronic system (8) to the counter-interference system and a second sub-path (17.2) for transmitting a signal from the counter-interference system (16) to the power electronic system (8). 12.Device (7) according to claim 10 or 11, wherein the transmission path (17) comprises at least one D / A converter (18), at least one injector (19) for coupling the total anti-interference signal (21) into the power electronic system (8), at least one sensor (14) for measuring the interference signal (4, 4.1) present in the power electronic system (8), and at least one A / D converter (15).
13. Device (7) according to any one of claims 10 to 12, wherein the anti-interference system (16) is configured to determine the transfer function for each interference-relevant frequency (f1, f2, f3) using a harmonic test signal (31).
14. Device (7) according to one of claims 10 to 13, wherein the counter-interference system (16) is designed as a Field Programmable Gate Array, DSP or Microcontroller in conjunction with analog-to-digital and digital-to-analog converters.