Simplified method for the impedance analysis of electrical components

Abstract

The invention relates to a method for the impedance analysis of an electrical component, wherein the component is periodically excited by means of an electrical excitation signal with a preselected signal shape and with a predetermined repetition frequency, and an electrical current profile and an electrical voltage profile at the component are measured as a function of time over at least one repetition period. The measured time-dependent current and voltage profiles are transformed into a frequency-dependent current and voltage spectrum using a mathematical method, wherein the frequency range comprises at least the repetition frequency and at least one integer multiple of the repetition frequency. Frequency-dependent impedance values are calculated on the basis of the respective transformed current and voltage spectra.

Description

[0001] Description

[0002] Simplified method for impedance analysis of electrical components

[0003] The invention relates to a method for impedance analysis of electrical components.

[0004] For a more in-depth analysis of the condition of electrical components, impedance spectroscopy is one of the techniques used. This method is used, for example, for electrochemical devices (e.g., lithium-ion batteries, fuel cells, or electrochemical electrolysis cells).

[0005] In impedance spectroscopy, the component is periodically excited externally during operation. This excitation can be a cyclic increase and decrease of the output current (e.g., via a frequency-modulated load) or an excitation voltage superimposed on the operating voltage.

[0006] Typically, a frequency generator is used for this purpose, producing a fully symmetrical signal (e.g., a sine wave), which is transmitted to the component via a voltage source (called "potentiostatic") or an electrical source or sink (called "galvanostatic"). The resulting response signal (voltage fluctuation in the case of galvanostatic excitation, or current signal fluctuation in the case of potentiostatic excitation) is measured on the component under test. A Fourier analysis of the resulting phase shift from excitation to response signal at different excitation frequencies provides information about the real and imaginary parts of the impedance. The basic function and process of impedance spectroscopy can be found in numerous books and publications.

[0007] Technically, this method is complex and susceptible to interference for several reasons: Generating the periodic excitation signals typically requires a frequency generator capable of producing a wide spectrum of different frequencies. Typically, many frequencies are scanned per measurement to capture a complex, frequency-dependent impedance pattern. Furthermore, the signals must be transmitted cleanly and with signal fidelity to the component using suitable voltage sources / current sinks, etc. Interference from cables and lines or the measurement setup regarding inductances and capacitances must be laboriously circumvented or measured out using calibration in order to derive a "clean" diagnosis of the component under test from the analysis.

[0008] Ideally, a setup should be identical for all (repeated) measurements (e.g., temporal aging measurements or after component damage due to temperature influences, etc.), i.e., it should be repeatable in terms of location and / or setup. Likewise, external disturbances on the signal path to be examined / excited must be avoided (e.g., load steps during a measurement cycle). The signals to be examined must be resolved very precisely (both the signal strength and the temporal correlation to the excitation). The procedure—particularly for a broad frequency spectrum (depending on the observed characteristics of the component under investigation)—requires a certain amount of measurement time. This is difficult to implement, particularly in mobile applications (e.g., online analysis in a vehicle). The described restrictions regarding the required hardware severely limit implementation in mobile or non-stationary applications.In addition, the required equipment and construction variants are comparatively cost-intensive.

[0009] The invention is therefore based on the object of providing a simplified method for diagnosing electrical components which avoids the difficulties mentioned above.

[0010] The object is achieved according to the invention by a method according to the independent claim. Advantageous further developments are specified in the dependent claims.

[0011] According to the inventive method for impedance analysis of an electrical component, the component is periodically excited by means of an electrical excitation signal with a preselected signal shape and at a predetermined repetition frequency. The course of the electrical current and the electrical voltage across the component is measured as a function of time (if necessary with a preselected sampling rate) over at least one repetition period. Theoretically, measurement over a single repetition period is sufficient. In practice, however, it is advisable to perform the measurement over at least two repetition periods. Three or four repetition periods have proven particularly advantageous for the measurement. However, measurement can also be carried out over five or ten repetition periods. There is fundamentally no upper limit here, but multiple repetition periods extend the measurement duration.

[0012] The measured time-dependent current and voltage waveforms are then transformed into a frequency-dependent current and voltage spectrum using a mathematical method. The frequency range over which the current and voltage spectrum extends comprises at least the repetition frequency and at least one integer multiple of the repetition frequency. Preferably, the frequency range comprises several integer multiples of the repetition frequency, for example, 10, 20, 30, 50, or more integer multiples of the repetition frequency.

[0013] Based on the current and voltage spectra thus obtained, frequency-dependent impedance values ​​are then calculated according to the invention. The frequency-dependent impedance values ​​can then advantageously be used to obtain information about the properties or condition of the electrical component.

[0014] The electrical excitation signal can be a current signal or a voltage signal, which corresponds to a potentiostatic or galvanostatic excitation as described above. Accordingly, according to the method according to the invention, the time-dependent course of the current and voltage at the component is measured. Depending on whether the excitation signal is a current signal or a voltage signal, the response signal is a voltage signal or a current signal. Thus, according to the invention, an excitation signal and a corresponding response signal in the form of a current signal and a voltage signal are always measured.

[0015] The excitation signal itself is not necessarily periodic, like a sine signal, for example, but is merely periodically repeated at a preselected repetition frequency. For example, the signal shape of the excitation signal can at least approximately correspond to a pulse, step, or step signal. For the method according to the invention, different pulse shapes or signal shapes are particularly advantageous, each of which can be represented by the superposition of different frequencies. These can, for example, correspond to a pulse excitation, step or step function or disturbance. A rectangular pulse, for example, can be represented as the superposition of an infinite number of sine waves. However, the signal structure (course of the excitation pulse) should not be viewed as a periodic excitation. In contrast, the excitation itself (e.g. the pulse / disturbance) is cyclically impressed at specific time intervals, and its temporal course is measured.This can in turn be used to subject the excitation and response signals to a Fourier analysis.

[0016] The excitation by means of pulse functions, step, jump or other disturbance signals can be implemented much more easily in applications than the generation of symmetrical, inherently periodic sine or step functions, which generally have to be generated by means of signal generators.

[0017] According to a particularly advantageous development of the method according to the invention, the excitation signal can be generated by a passive component. Such a passive component can be, for example, a capacitor or a coil, or even an electrical resistor. The excitation signal is then, for example, a charge or discharge signal or a decay pulse of the passive component. The signal shape is then determined by the specific charge or discharge signal of the passive component.

[0018] It is particularly advantageous to generate the excitation signal using a capacitor that is connected upstream of the component in the electrical circuit. This capacitor can be periodically charged and discharged, thus periodically imparting an excitation signal to the electrical component. In this case, the electrical excitation signal is a current pulse. The corresponding response signal is then a voltage signal. A corresponding excitation method using a passive component is described, for example, in German patent application DE 102023204236.3, which was not yet published at the time of application.

[0019] According to an advantageous development of the method according to the invention, the impedance values ​​are not calculated for a continuous spectrum across the entire frequency range, but rather suitable frequencies are advantageously selected. In particular, frequencies corresponding to integer multiples of the repetition frequency are selected. Such suitable frequencies are selected, for example, using a window function.

[0020] Preferably, a filter function can also be applied. This allows, for example, only areas of the current and voltage spectrum whose amplitude lies above a threshold value to be selected. Frequencies whose current or voltage amplitude lies below the threshold value are preferably ignored. This advantageously only considers measured values ​​that are significantly above noise. The impedance values ​​can then advantageously be calculated only for the selected frequencies of the current and voltage spectrum. This saves computational effort. In principle, the selection of suitable frequencies can also be made after the impedance values ​​have been calculated.

[0021] The measured time-dependent current and voltage values ​​are transformed in a known manner using a mathematical method, such as a Fourier transformation or a Laplace transformation. A dynamic Fourier transformation (DFT) or a fast Fourier transformation (FFT) can be used advantageously.

[0022] When applying the Fast Fourier Transform (FFT), it is particularly important that sufficient measurement points are acquired per excitation signal pulse for the time-dependent measurement of current and voltage. Therefore, a sampling rate that ensures sufficient measurement points per repetition period should preferably be selected. Accordingly, the sampling rate should preferably be selected based on the repetition frequency.

[0023] The method according to the invention makes it possible to calculate impedance which, however, is not related to a specific excitation frequency of the interference signal, but only the sampling rate and the repetition rate or.

[0024] Repetition frequency is taken into account. The spurious frequency amplitudes (also called image frequencies or spectral frequencies) discovered during the Fourier analysis can be filtered, and an impedance value for a large number of these spurious frequencies can be evaluated. This produces an impedance spectrum that can be superimposed on a conventionally generated impedance spectrum. Measurement intervals and signal properties can be adjusted to ensure a high degree of spectral agreement. This significantly reduces the excitation measurement time, the power required for excitation, and the costs for the required components / hardware compared to conventional impedance analysis methods.The method according to the invention allows impedance spectra to be generated (regardless of the signal structure) that examine a multitude of frequencies, even though the repetition frequency of the excitation is several times lower. In principle, one could thus impose an excitation using a pulse function at an exemplary frequency of 2 Hz, but process an impedance spectrum for 100, 1000, or even 10,000 Hz. Compared to known methods for the impedance analysis of electrical components, the invention thus enables a simplified method in which excitation does not have to be performed over a large frequency range; instead, a periodic repetition of a simple excitation signal is sufficient to obtain an impedance analysis across a broad frequency spectrum.

[0025] In the following, the invention will be explained in more detail using the drawings 1 - 3 as examples.

[0026] They show schematically:

[0027] Figure 1 : a series of current excitation pulses with a repetition frequency of approximately 10 Hz;

[0028] Figure 2: a series of corresponding voltage response signals to the excitation signal according to Figure 1; and

[0029] Figure 3: a frequency-dependent current spectrum generated by mathematical transformation from the excitation pulses shown in Figure 1;

[0030] Figure 4: Frequency-dependent impedance values ​​determined according to the method according to the invention compared to impedance values ​​determined using a conventional method for impedance analysis; and

[0031] Figure 5: an exemplary measuring arrangement for carrying out the method according to the invention according to an advantageous embodiment.

[0032] Figure 1 shows, by way of example, a series of excitation pulses which, according to an embodiment of the method according to the invention, can be impressed onto an electrical component 20 to be analyzed at a repetition frequency 12 of 10 Hz. Figure 1 shows, by way of example, current pulses which represent, for example, discharge pulses of an electrical capacitor. The capacitor is charged and discharged at a repetition frequency 12 of 10 Hz. The repetition period 15 is accordingly 0.1 s. According to the invention, this periodic excitation signal 13 of asymmetric current pulses and the corresponding response signal 14 in the form of a voltage signal are measured at the component 20. A corresponding voltage signal 14 is shown as an example in Figure 2.

[0033] According to another embodiment of the method according to the invention, excitation by means of periodically executed voltage pulses is also possible. In this case, a current signal 13 would be measured as the response signal. In any case, the excitation signal and the response signal together form a periodic current signal 13 and a periodic voltage signal 14 with the same repetition frequency 12. The current signal 13 and the voltage signal 14 are each measured at a sampling rate that ensures sufficient measurement points per period to perform a mathematical transformation, such as a fast Fourier transform (FFT).

[0034] Through mathematical transformation, the measured time-dependent current 13 and voltage signals 14 are each converted into a frequency-dependent current or voltage spectrum 16 according to the invention. Figure 3 shows an example of a frequency-dependent current spectrum 16. The absolute values ​​of the transformed complex current values ​​are shown. The excitation frequency according to the exemplary embodiment shown in Figures 1 and 2 is 10 Hz. Consequently, in the frequency spectrum from Figure 3, current peaks, or generally amplitude peaks, can be seen at the repetition frequency 12 of 10 Hz as well as at integer multiples of the repetition frequency 12. The calculated spectrum here covers a frequency range from the repetition frequency (10 Hz) to over 60 times (600 Hz) the repetition frequency. The integer multiples of the repetition frequency 12 are also called secondary frequencies or image frequencies.To evaluate the amplitude spectrum, the amplitude peaks at the repetition frequency 12 as well as at the secondary or image frequencies are preferably taken into account. This can be achieved by applying a window function to the obtained amplitude spectra 16 (current and voltage spectrum). Secondary frequencies with a small amplitude below the noise level should preferably be ignored. Amplitude values ​​below a threshold value 18, represented in Figure 3 by a horizontal solid line, are also ignored, for example, by applying a filter function. This ensures that only measured values ​​that are significantly above the noise level are used to determine the impedance.

[0035] By applying the filter and / or window function, suitable frequencies 19 are selected from the calculated amplitude spectra. For these selected frequencies 19, which each represent, for example, a multiple of the repetition frequency 12, impedance values ​​17 are calculated in a known manner from the respective current and voltage values ​​at the respective frequencies 19.

[0036] These impedance values ​​17 are calculated as examples for the repetition frequency 12 of 10 Hz as well as for integer multiples up to 10 times the repetition frequency at 100 Hz and are plotted as discrete points (crosses) in Figure 4. The solid line corresponds to a simulation of impedance values ​​obtained using a conventional, sinusoidal excitation signal. The combined representation in Figure 4 shows that the impedance values ​​17 determined using the method according to the invention agree well with an impedance curve determined using a conventional method.

[0037] The method according to the invention thus enables a method for impedance analysis that is greatly simplified compared to the conventional method.

[0038] Figure 5 shows an example of a measuring arrangement 21 with which excitation can be realized according to an advantageous embodiment of the method according to the invention. The electrical component 20 to be tested is connected to a load 36 in an electrical circuit 21. A first measuring device 46 serves to measure the electrical voltage applied to the component 20 to be tested. A second measuring device 47 is provided to measure the electrical current flowing in the circuit. A first passive component 22, for example a capacitor, can be connected in the circuit in parallel with the component 20 to be tested by means of a first switch 40. The first passive component 22 can be connected to a (further) voltage source 45 via a second switch 44 in order to be charged with a voltage.Thus, the passive component can be periodically connected to the voltage source 45 and brought to a defined potential before it applies its charging or discharging current as an excitation signal to the electrical component 20. This process is repeated periodically at the repetition frequency.

Claims

Claims 1. Method for impedance analysis of an electrical component (20) comprising the steps: - periodically exciting the component by means of an electrical excitation signal (13) with a preselected signal shape and with a predetermined repetition frequency (12), - measuring an electrical current profile (13) and an electrical voltage profile (14) on the component (20) as a function of time over at least one repetition period (15), - Transforming the measured time-dependent current (13) and voltage (14) curves by means of a mathematical method into a frequency-dependent current and voltage spectrum (16) over a frequency range which includes at least the repetition frequency (12) and at least an integer multiple of the repetition frequency, - Calculating frequency-dependent impedance values (17) based on the respective current and voltage spectra (16).

2. Method for impedance analysis of an electrical component (20) according to claim 1, wherein the electrical excitation signal is a current signal (13) or a voltage signal (14).

3. Method for impedance analysis of an electrical component (20) according to claim 1 or 2, wherein the signal shape of the excitation signal (13) corresponds at least approximately to a pulse, step or jump signal.

4. Method for impedance analysis of an electrical component (20) according to claim 1 or 2, wherein the excitation signal (13) is generated by a passive component (22).

5. Method for impedance analysis of an electrical component (20) according to one of the preceding claims, with the further step: - Selecting suitable frequencies (19) and calculating the impedance values based on the voltage and current spectrum (16) for the selected frequency values (19).

6. Method for impedance analysis of an electrical component (20) according to claim 5, wherein suitable frequencies (19) each comprise the repetition frequency (12) or an integer multiple of the repetition frequency (12).

7. Method for the impedance analysis of an electrical component (20) according to one of the preceding claims, wherein the mathematical method for transformation comprises a Fourier transform, a Dynamic Fourier Transform (DFT), a Fast Fourier Transform (FFT), or a Laplace Transform.

8. Method for the impedance analysis of an electrical component (20) according to one of the preceding claims, wherein a sampling rate is selected as a function of the repetition period (15) for measuring the electrical current profile (13) and the electrical voltage profile (14) at the component (20).

Images