A method for testing capacitance loss

Abstract

The invention belongs to the technical field of capacitor loss testing, and in particular relates to a capacitor loss testing method, comprising the following steps: obtaining an original data set related to electrolytic capacitor loss by performing a three-pulse cycle test on an electrolytic capacitor; calculating a charge qc flowing through an electrolytic capacitor under different test conditions according to the original data set and establishing a data set of electrolytic capacitor loss; training a BP neural network using the data set of electrolytic capacitor loss to obtain node parameters of the neural network; calculating the loss of a DC side electrolytic capacitor under different working conditions when a converter is actually running using the existing neural network parameters; during the test, comprehensively considering the influence of fundamental wave components and high-frequency components on capacitor loss, and keeping the temperature of the electrolytic capacitor almost unchanged, thereby eliminating the influence of temperature changes on electrolytic capacitor loss during the test, thereby achieving accurate evaluation of electrolytic capacitor loss and ensuring the accuracy of the calculation result.

Description

Technical Field

[0001] The invention belongs to the technical field of capacitor loss testing, and in particular relates to a capacitor loss testing method. Background Art

[0002] The reliability of electrolytic capacitors is usually measured by their lifespan, which can be estimated by the temperature of the capacitor. In order to accurately evaluate the temperature rise, an accurate loss test method must be used to build an accurate loss model. Therefore, accurately evaluating the power loss of electrolytic capacitors is crucial to improving system efficiency, thermal performance, and reliability.

[0003] Typically, the loss of an electrolytic capacitor is calculated by its equivalent series resistance (ESR), which is measured by the capacitor manufacturer under very low current, 100Hz or 120Hz sinusoidal current excitation. However, power electronic converters usually operate in a switching state to generate the required output voltage and current, which causes the actual excitation current of the electrolytic capacitor to present a complex square or triangular waveform with a high amplitude. In this case, the ESR is constantly changing, so it is inaccurate to use the ESR value provided in the data sheet to calculate the capacitor loss.

[0004] Another method to calculate capacitor loss is based on the loss calculation of the capacitor current spectrum. The capacitor current at different frequencies is obtained through Fourier decomposition, and the ESR value at the corresponding frequency is used to calculate the capacitor loss. This method usually requires the user to test the ESR value at different frequencies, and the influence of high-frequency components on the loss is often ignored during the calculation process.

[0005] Due to the limitations of the above two methods, there is an urgent need for a capacitor loss evaluation method that takes into account both the influence of the fundamental component and the influence of the high-frequency component, especially the component near the switching frequency. Summary of the invention

[0006] The invention discloses a capacitor loss test method, which mainly solves the problem that the current electrolytic capacitor loss calculation is inaccurate and the influence of high-frequency components on the loss is ignored.

[0007] To achieve the above object, the present invention provides a capacitance loss test method, comprising the following steps:

[0008] Raw data acquisition step: by performing a three-pulse cycle test on the electrolytic capacitor, the raw data set about the electrolytic capacitor loss is obtained;

[0009] The steps for establishing the data set of electrolytic capacitor loss are as follows: Calculate the charge qc flowing through the electrolytic capacitor under different test conditions based on the original data set, and calculate the energy E dissipated by the electrolytic capacitor under the corresponding test conditions. chargeand the energy dissipated by discharge E discharge , with a DC voltage V dc , the on-time of the switch device Q1 is T on , the off-time of the switch device Q1 is T off , inductor current I L , the energy dissipated by charging E charge , the energy dissipated by discharge E discharge Create a data set for the variable electrolytic capacitor loss;

[0010] Neural network node parameter acquisition step: using the electrolytic capacitor loss data set to train the BP neural network and obtain the node parameters of the neural network;

[0011] The steps for calculating the loss of electrolytic capacitors are as follows: using the existing neural network parameters, combined with the DC bus voltage obtained from the simulation when the converter is running continuously, the duration of the switch tube being turned on and off in each switching cycle, and the output current parameters of the converter during the duration of the switch tube being turned on and off, calculate the loss of the DC side electrolytic capacitor under different working conditions during the actual operation of the converter.

[0012] Furthermore, a three-pulse cycling test was performed using a full-bridge circuit based on GaN HEMT devices.

[0013] Furthermore, a three-pulse cycle test is performed to simulate the energy flow of the electrolytic capacitor in one switching cycle when the single-phase converter is bipolar modulated;

[0014] Specifically: When performing a three-pulse test, a set of driving signals is given to switch device 1 Q1 and switch device 4 Q4. The signal consists of multiple pulses, and the high level duration of the first pulse is T 1+ , low level duration is T 1- , and the duration of all subsequent high-level pulses is T 2+ , low level duration is T 2- , where T 1- =T 2+ =T 2- , T 1+ The time is determined by the inductor current I L Determine, the specific calculation is T 1+ =I L ∙L / V dc , the driving signals of switch device 2 Q2 and switch device 3 Q3 are complementary to the driving signals of switch device 1 Q1 and switch device 4 Q4, and the DC voltage V is recorded. dc , the voltage across the electrolytic capacitor V C , DC current I on the left side of the electrolytic capacitor C_power , DC current I on the right side of the electrolytic capacitor C_bridge, the on-time of the switch device Q1 is T on , the off-time of the switch device Q1 is T off , inductor current I L .

[0015] Furthermore, the VQ hysteresis loop of the electrolytic capacitor is drawn, and the energy E dissipated by charging is calculated based on the VQ hysteresis loop. charge and the energy dissipated by discharge E discharge .

[0016] Furthermore, the variables that need to be detected during the collection of the original data set are: DC side voltage V dc , inductor current and different pulse times T 1+ , T 1- , T 2+ , T 2- ; In the loop test, one of the above variables is changed during the loop, and other variables are kept unchanged to obtain the test data set;

[0017] The test range of the variables must cover the actual operating range of the converter, such as the DC side voltage and output current need to be tested from the minimum value to the rated value, T 2+ It is necessary to cover the device on-time in each switching cycle during actual operation.

[0018] Furthermore, the charge qc of the electrolytic capacitor is calculated as follows:

[0019] ;

[0020] I C_power is the DC current on the left side of the electrolytic capacitor, I C_bridge为 DC current on the right side of an electrolytic capacitor.

[0021] Furthermore, the energy E dissipated by the charging charge The calculation formula is:

[0022] ;

[0023] V max The voltage at the end of charging of the capacitor; V min The voltage at the beginning of charging the capacitor; It is the change in the charge of the capacitor during the charging or discharging process.

[0024] Furthermore, the energy dissipated by the discharge E discharge With E charge is calculated in the same way.

[0025] Furthermore, in the process of training the BP neural network, the input variable used is the DC voltage Vdc , the on-time of the switch device Q1 is T on , the turn-off time T of the switch device Q1 off , inductor current I L The output variable used is the energy E dissipated by charging the electrolytic capacitor. charge , the energy dissipated by the discharge of the electrolytic capacitor E discharge .

[0026] Furthermore, the calculated DC side electrolytic capacitor loss under different operating conditions during the actual operation of the converter refers to the dissipated energy E of the electrolytic capacitor charging and discharging in each switching cycle. charge 、E discharge The loss of the electrolytic capacitor on the DC side during the entire fundamental wave cycle needs to be obtained by adding the dissipated energy of the electrolytic capacitor charging and discharging in each switching cycle.

[0027] The technical solution provided by the present invention has at least the following technical effects:

[0028] The present invention discloses a method for testing capacitance loss, which makes the excitation of the electrolytic capacitor under test closer to the excitation of the electrolytic capacitor on the DC side of the converter under actual operation. The duration of the test process is only a few hundred milliseconds, so during the test, the temperature of the electrolytic capacitor remains almost unchanged, thereby eliminating the influence of temperature changes on the loss of the electrolytic capacitor during the test, thereby achieving accurate evaluation of the loss of the electrolytic capacitor and ensuring the accuracy of the calculation results.

[0029] When evaluating the loss of electrolytic capacitors, this method comprehensively considers the impact of fundamental components and high-frequency components on capacitor losses. This technology is not only applicable to the evaluation of DC side electrolytic capacitor losses, but also to the evaluation of losses of different types of capacitors such as DC side film capacitors and ceramic capacitors. Loss estimation only involves time domain variables and electrical variables, aiming to simplify the user's use process. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0031] Figure 1 A flowchart of a capacitance loss test method according to an embodiment of the present invention;

[0032] Figure 2 This is a topological diagram of a test platform according to an embodiment of the present invention;

[0033] Figure 3 A schematic diagram of a half-loop of a VQ hysteresis loop when charging an electrolytic capacitor according to an embodiment of the present invention;

[0034] Figure 4 A schematic diagram of a half-loop of a VQ hysteresis loop when an electrolytic capacitor is discharged according to an embodiment of the present invention;

[0035] Figure 5 This is a pulse test schematic diagram of Embodiment 3 of the present invention;

[0036] Figure 6 This is a pulse test experiment waveform diagram of Example 3 of the present invention;

[0037] Figure 7 Embodiment V of the present invention dc This is the electrolytic capacitor loss data set at 20V;

[0038] Figure 8 Embodiment V of the present invention dc This is a comparison chart between the loss evaluation results and the actual loss results at 20V;

[0039] Fig. 9 The figure is a structural block diagram of the test platform according to an embodiment of the present invention. DETAILED DESCRIPTION

[0040] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the embodiments of the present invention, and should not be construed as limiting the present invention.

[0041] According to an embodiment of the present invention

[0042] Reference Figure 1 This embodiment discloses a capacitor loss test method, which is mainly used for loss detection of electrolytic capacitors. Fig. 9 The structure diagram of the test platform of this embodiment is shown as follows Figure 2 The topological structure diagram of the test platform of this embodiment is shown in FIG. The specific operation steps are as follows: First, a full-bridge circuit based on a gallium nitride high electron mobility transistor (GaN HEMT) device is used to perform a three-pulse cycle test on an electrolytic capacitor. Figure 6 The single three-pulse test experimental waveform is shown to obtain the original data set related to the electrolytic capacitor loss. The three-pulse cycle test simulates the energy flow of the electrolytic capacitor in one switching cycle under the bipolar modulation of the single-phase converter. Figure 5 A schematic diagram showing the test, Figure 6Specifically, when performing the three-pulse test, a set of driving signals consisting of 3 to 6 pulses is provided to the switch device Q1 and the switch device Q4. The high level duration of the first pulse is T 1+ , the low level duration is T 1- , and the high level duration of all subsequent pulses is T 2+ The low level duration is T 2- Among them, T 1- Equal to T 2+ Equal to T 2- , while T 1+ The duration of the inductor current I L The specific calculation formula is T 1+ =I L ∙L / V dc The driving signals of the switch device 2 Q2 and the switch device 3 Q3 are complementary to the driving signals of the switch device 1 Q1 and the switch device 4 Q4. The recorded data includes the DC voltage V dc , the voltage across the electrolytic capacitor V C , DC current I on the left side of the electrolytic capacitor C_power , DC current I on the right side of the electrolytic capacitor C_bridge , the conduction duration T of the switch device Q1 on And the off time duration T of the switch device Q1 off , inductor current I L .

[0043] In the three-pulse cycle test method described, the parameters to be monitored include the DC link voltage V dc , inductor current and different pulse times T 1+ , T 1- , T 2+ , T 2- During the cyclic test, one of the above parameters will be changed in each cycle while keeping the other parameters unchanged to obtain the test data set. In addition, the test variation range of the parameters must cover the actual operating range of the converter. For example, the DC side voltage and output current should be tested from the minimum value to the rated value, while T 2+ The device on-time in each switching cycle during actual operation should be covered.

[0044] Then, based on the original data set, the charge qc flowing through the electrolytic capacitor under different test conditions is calculated, and the voltage-charge VQ hysteresis loop of the electrolytic capacitor is plotted. The VQ hysteresis loop half loop of the electrolytic capacitor during charging and discharging is shown in the figure below. Figure 3 , Figure 4 As shown, based on the VQ hysteresis loop, the energy E dissipated during the charging process of the electrolytic capacitor under the corresponding test conditions is calculated.charge And the energy dissipated during the discharge process E discharge . Then, with a DC voltage V dc , the conduction duration T of the switch device Q1 on , the off-time duration T of the switching device Q1 off , inductor current I L , charging dissipation energy E charge and discharge dissipation energy E discharge As a variable, a data set of electrolytic capacitor losses is constructed.

[0045] The charge qc flowing through the electrolytic capacitor is calculated as follows:

[0046] ;

[0047] I C_power is the DC current on the left side of the electrolytic capacitor, I C_bridge is the DC current on the right side of the electrolytic capacitor.

[0048] Energy dissipated by charging the electrolytic capacitor E charge Calculation, such as Figure 3 As shown, V max is the voltage at the end of capacitor charging, V min The voltage at the beginning of charging the capacitor; It is the change of the capacitor charge during the charging or discharging process. The specific calculation is as follows:

[0049] ;

[0050] The energy dissipated by the discharge of the electrolytic capacitor is E discharge Calculation and E charge The calculation is the same.

[0051] Then, the BP neural network is trained using the electrolytic capacitor loss data set to obtain the node parameters of the neural network. In the process of training the BP neural network, the input variable used is the DC voltage V dc , the on-time of the switch device Q1 is T on , the turn-off time T of the switch device Q1 off , inductor current I L The output variable used is the energy E dissipated by charging the electrolytic capacitor. charge , the energy dissipated by the discharge of the electrolytic capacitor E discharge .

[0052] Finally, by applying the existing neural network parameters and combining the DC bus voltage data of the converter in continuous operation, the on and off time of the switch tube in each switching cycle, and the output current parameters of the converter during the on and off period of the switch tube, we were able to calculate the DC side electrolytic capacitor loss of the converter under different working conditions in actual operation. Specifically, this involves the charge and discharge dissipation energy E of the electrolytic capacitor in each switching cycle. charge With E discharge The total loss of the electrolytic capacitor on the DC side during the entire fundamental cycle is calculated by accumulating the energy dissipated by the charge and discharge of the electrolytic capacitor in each switching cycle. Figure 7 As shown, V dc This is the electrolytic capacitor consumption data set at 20V. Figure 8 The comparison between the loss calculation result and the actual loss result is shown when the DC side voltage is 20V and the AC side output current peak is 0.7A. It can be observed from the figure that with the increase of switching frequency, the loss of the electrolytic capacitor on the DC side is reduced accordingly, and the calculation error is controlled within 9%, indicating that the calculation result has a high accuracy.

[0053] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should all be included in the protection scope of the present invention.

Claims

1. A capacitance loss test method, characterized in that: The steps include: Raw data acquisition step: by performing a three-pulse cycle test on the electrolytic capacitor, obtaining the raw data set related to the electrolytic capacitor loss; The test platform used for the test includes an electrolytic capacitor, an inductor L, a switch device Q1, a switch device Q2, a switch device Q3, and a switch device Q4; the switch device Q1 and the switch device Q2 are connected in series and then connected in parallel with the electrolytic capacitor, the switch device Q3 and the switch device Q4 are connected in series and then connected in parallel with the electrolytic capacitor, one end of the inductor L is connected to the connection point of the switch device Q1 and the switch device Q2, and the other end is connected to the connection point of the switch device Q3 and the switch device Q4; The three-pulse cycle test is used to simulate the energy flow of electrolytic capacitors in a switching cycle when the single-phase converter is bipolar modulated; Specifically: When performing a three-pulse test, a set of driving signals is given to switch device 1 Q1 and switch device 4 Q4. The signal consists of multiple pulses, and the high level duration of the first pulse is T 1+ , low level duration is T 1- , and the duration of all subsequent high-level pulses is T 2+ , low level duration is T 2- , where T 1- =T 2+ =T 2- , T 1+ The time is determined by the inductor current I L Determine, the specific calculation is T 1+ =I L ∙L / V dc , the driving signals of switch device 2 Q2 and switch device 3 Q3 are complementary to the driving signals of switch device 1 Q1 and switch device 4 Q4, and the DC voltage V is recorded. dc , the voltage across the electrolytic capacitor V C , DC current IC on the left side of the electrolytic capacitor _power , DC current I on the right side of the electrolytic capacitor C_bridge , the on-time of the switch device Q1 is T on , the off-time of the switch device Q1 is T off , inductor current I L ; The steps for establishing the data set of electrolytic capacitor loss are as follows: Calculate the charge qc flowing through the electrolytic capacitor under different test conditions based on the original data set, and calculate the energy E dissipated by the electrolytic capacitor under the corresponding test conditions. charge and the energy dissipated by discharge E discharge , with a DC voltage V dc , the on-time of the switch device Q1 is T on , the off-time of the switch device Q1 is T off , inductor current I L , the energy dissipated by charging E charge , the energy dissipated by discharge E discharge Create a data set for the variable electrolytic capacitor loss; Neural network node parameter acquisition step: using the electrolytic capacitor loss data set to train the BP neural network and obtain the node parameters of the neural network; The steps for calculating the loss of electrolytic capacitors are as follows: using the existing neural network parameters, combined with the DC bus voltage obtained from the simulation when the converter is running continuously, the duration of the switch tube being turned on and off in each switching cycle, and the output current parameters of the converter during the duration of the switch tube being turned on and off, calculate the loss of the DC side electrolytic capacitor under different working conditions during the actual operation of the converter.

2. The capacitance loss test method according to claim 1, characterized in that: The three-pulse cycling test uses a full-bridge circuit based on GaN HEMT devices.

3. The capacitance loss test method according to claim 1, characterized in that: Draw the VQ hysteresis loop of the electrolytic capacitor and calculate the energy E dissipated by charging based on the VQ hysteresis loop charge and the energy dissipated by discharge E discharge .

4. The capacitance loss test method according to claim 1, characterized in that: The variables that need to be detected during the collection of the original data set are: DC side voltage V dc , inductor current and different pulse times T 1+ , T 1- , T 2+ , T 2- ; In the loop test, one of the above variables is changed during the loop, and other variables are kept unchanged to obtain the test data set; The DC side voltage and output current need to be tested from the minimum value to the rated value, T 2+ It is necessary to cover the device on-time in each switching cycle during actual operation.

5. The capacitance loss test method according to claim 1, characterized in that: The charge qc calculation formula of the electrolytic capacitor is: ； I C_power is the DC current on the left side of the electrolytic capacitor, I C_bridge is the DC current on the right side of the electrolytic capacitor.

6. The capacitance loss test method according to claim 1, characterized in that: The energy E dissipated by the charging charge The calculation formula is: ； V max The voltage at the end of capacitor charging; V min The voltage at the beginning of charging the capacitor; ∆q c It is the change in the charge of the capacitor during the charging or discharging process.

7. The capacitance loss test method according to claim 6, characterized in that: The energy dissipated by the discharge is E discharge With E charge is calculated in the same way.

8. The capacitance loss test method according to claim 1, characterized in that: The calculated DC side electrolytic capacitor loss under different operating conditions during the actual operation of the converter refers to the dissipated energy E of the electrolytic capacitor charging and discharging in each switching cycle. charge 、E discharge The loss of the electrolytic capacitor on the DC side during the entire fundamental wave cycle needs to be obtained by adding the dissipated energy of the electrolytic capacitor charging and discharging in each switching cycle.

Images