SVPWM dead zone accurate compensation method based on oscilloscope phase mark

Abstract

The invention relates to the technical field of power electronics and motor control, in particular to an accurate SVPWM (space vector pulse width modulation) dead zone compensation method based on oscilloscope phase marking, which comprises the following steps of: in frequency converter control software, setting a compensation angle to be 0, and inserting a zone bit of IO (input / output) conversion into a 0 phase of U-phase wave; operating the frequency converter, observing the position of the flag bit at the U-phase current through the oscilloscope, and obtaining a position angle, namely a compensation angle; the compensation angle is added to the software, and the obtained current compensation angle is the angle of the output voltage waveform + the angle of the output current + the compensation angle; and operating the frequency converter again, observing the position of the flag bit on the U-phase current through the oscilloscope, and judging whether the position is on the zero phase angle of the current or not. According to the method, the oscilloscope is utilized to visually find and determine the optimal compensation angle, so that accurate compensation of the SVPWM dead zone is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of power electronics and motor control, and in particular to a method for accurately compensating a SVPWM dead zone based on an oscilloscope phase marker. Background Art

[0002] With the continuous improvement of microprocessor computing power and IGBT / MOS switching device performance, PWM-based voltage source converters are widely used in scenarios such as frequency converters, servo drives, inverters, new energy vehicle traction, and rail transit traction. Among them, space vector modulation (SVPWM) has become the mainstream solution due to its high DC bus utilization and controllable harmonic spectrum distribution; the three-phase half-bridge controls the complementary drive of the upper and lower bridge arms. In order to avoid direct conduction of the same arm, a dead time must be introduced between the drive pulses; however, the dead time will introduce average output voltage vector deviation and zero displacement, causing the device that should be turned on to be temporarily not turned on and the device that should be turned off to be temporarily not turned off, especially at low frequency, light load and near current zero crossing, which manifests as phase current distortion and torque Problems such as pulsation and low-speed creep; the deviation is related to the device's on / off delay, reverse recovery, temperature, bus voltage, load inductance and drive circuit parasitic parameters, and has significant nonlinearity and time-varying properties; traditional compensation paths mainly include: current-type feedback compensation based on current polarity, voltage-type feedback compensation based on output voltage estimation, and indirect compensation of dead-zone errors extracted by high-frequency pulses; the above methods usually require the establishment of equivalent models and compensation transformation formulas containing device and load parameters, the algorithm chain is long and the parameters are highly sensitive; and the synchronous calibration of the compensation amount and compensation angle / phase is difficult. The actual wave transmission angle, sampling delay and zero-crossing drift will cause inaccurate selection of the compensation corner point, high debugging cost and poor reproducibility.

[0003] CN104065258A discloses a dead-zone compensation method for an inverter SVPWM waveform. This method concentrates the dead-zone impact on the upper bridge or lower bridge by determining the voltage space vector and current direction, and distinguishes between the parts that need to be compensated and those that do not by calculating the current polarity. A freewheeling diode is used to advance or delay the dead-zone time. Although this method is relatively simple to implement and has low requirements on chip computing speed, its compensation accuracy is limited by the judgment of current polarity. Under low speed and light load conditions, the current signal is weak and easily interfered by noise, resulting in poor compensation effect and inability to achieve precise compensation.

[0004] CN111697897A discloses a dead time compensation method for a motor controller based on predicted current. The method collects the instantaneous current of the three-phase motor and predicts it, converts the predicted current to the dq axis, and then calculates the dead time error voltage compensation duty cycle. Although this method improves the real-time performance of compensation by predicting the current, its algorithm is complex and the calculation amount is large, which puts high demands on the processing performance of the controller. More importantly, these existing technical solutions rely on complex mathematical models and internal parameter calculations, which are difficult to adjust and verify intuitively.

[0005] Therefore, the existing technology has the following problems: the algorithm is complex, the calculation amount is large, and the hardware resources are high; the compensation accuracy is limited by the quality of the current signal, especially under low-speed and light-load conditions, the compensation effect is poor; there is a lack of intuitive debugging and verification means, and it is difficult to perform accurate parameter correction in practical applications; the present invention addresses the problems of complex calculations, inaccurate compensation and lack of intuitive debugging means in the existing dead zone compensation methods, and proposes an SVPWM dead zone precise compensation method based on oscilloscope phase marking, which aims to use the oscilloscope to intuitively find and determine the optimal compensation angle, thereby achieving accurate compensation for the SVPWM dead zone. Summary of the Invention

[0006] The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract of the specification and the title of the invention of this application to avoid blurring the purpose of this section, the abstract of the specification and the title of the invention, and such simplifications or omissions cannot be used to limit the scope of the invention.

[0007] In view of the above existing problems, the present invention is proposed.

[0008] To solve the above technical problems, the present invention provides the following technical solutions: Step S1: in software control, the compensation angle is set to 0, and the IO conversion flag is inserted into the 0 phase of the U-phase wave; Step S2: Run the frequency converter and observe the position of the U-phase current through an oscilloscope. The position angle obtained is the compensation angle. Step S3, adding the compensation angle to the software to obtain a current compensation angle of: the angle of the output voltage waveform + the output current angle + the compensation angle; Step S4: Run the inverter again, observe the position of the flag at the U-phase current through the oscilloscope, and determine whether the position is at the 0 phase angle of the current: If so, the position at the 0 phase angle of the current is the dead zone compensation angle; If not, continue adjusting the current compensation angle value and repeat step S3 until the flag is at the zero-point phase of the current waveform.

[0009] The beneficial effects of the present invention are as follows: the present invention realizes direct, rapid and accurate calibration of the dead zone compensation angle: on the one hand, the traditional compensation problem requiring sophisticated models and multi-parameter identification is transformed into a visual phase alignment problem, significantly reducing the computational complexity and parameter sensitivity; on the other hand, the compensation angle is closed-loop calibrated under real working conditions, which can fully absorb the combined errors of the device dynamic characteristics and the control link delay, thereby reducing the current THD in the low-frequency and light-load areas, reducing the torque pulsation, reducing the risk of zero-crossing oscillation and DC bias, and improving the controllability of the weak magnetic area and low-speed crawling conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort. Among them: Figure 1 Schematic diagram of the flow of the SVPWM dead zone accurate compensation method based on oscilloscope phase marking shown in the present invention; Figure 2 This is a current waveform diagram without dead zone compensation shown in the present invention; Figure 3 The waveform diagram of the upper and lower bridge arms under the ideal state shown in the present invention; Figure 4 Schematic diagram of the Clark transformation process shown in the present invention; Figure 5 Schematic diagram of the process of converting three-phase current into controllable d / d current shown in the present invention; Figure 6 This is a waveform diagram of the output current of the insertion mark position when the insertion angle is -90° as shown in the present invention; Figure 7 This is an output current waveform diagram of the present invention when the dead zone compensation angle is 0 and a flag is inserted; Figure 8 This is a schematic diagram of no-load current output before 5Hz compensation shown in the present invention; Figure 9 This is the output current waveform diagram after dead zone compensation shown in the present invention; Figure 10 This is a waveform diagram of the rated 50Hz output current after dead zone compensation shown in the present invention. DETAILED DESCRIPTION

[0011] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0012] Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without making any creative work should fall within the scope of protection of the present invention.

[0013] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0014] It should be noted that the purpose of the dead time is that when the MOS or IGBT is in the on state, the dead time is added to close it in advance; when the MOS or IGBT is in the off state, the dead time is added to delay the opening, which effectively avoids the upper and lower bridge arms being directly connected due to the delay problem when the MOS or IGBT is turned on or off; However, due to the addition of the dead time, the tube that should be turned on is not turned on, and the tube that should be turned off is not turned off, which will inevitably lead to a larger distortion rate of the current waveform when the motor is driven or the inverter load is light; such as Figure 2 , is the current waveform of the EV520 series inverter at 5Hz without dead zone compensation, Figure 3 These are the waveforms of the upper and lower bridge arms under theoretical and ideal conditions.

[0015] Traditional dead-band compensation methods include current-type feedback dead-band compensation, voltage-type feedback dead-band compensation, and high-frequency pulse extraction dead-band compensation. These dead-band compensation methods often involve complex compensation conversion theory formulas, making the entire compensation calculation difficult to understand. Furthermore, due to the selection of the compensation angle point and the actual PWM wave angle, it is difficult to accurately find an appropriate compensation angle, making dead-band compensation extremely difficult. Therefore, the core of dead-band compensation lies in determining the compensation angle and compensation amount.

[0016] According to an embodiment of the present invention, Figure 1 The flowchart shown is a method for accurately compensating SVPWM dead zone based on oscilloscope phase marking, which specifically includes the following steps: Step S1: In software control, set the compensation angle to 0 and insert the IO conversion flag at the 0th phase of the U-phase wave. In the inverter control software, initialize the dead zone compensation angle parameter to ; In the SVPWM algorithm, the zero phase point of the U-phase output voltage modulation waveform is located, which is recorded as ; exist At the corresponding interrupt moment, a digital I / O signal that can be captured externally is output. The I / O signal switches from a high level to a low level or from a low level to a high level at the 0 phase point, which serves as a flag that can be triggered by an oscilloscope.

[0017] Furthermore, the flags include: Set the idle general-purpose I / O pins of the inverter to digital output mode; In the main loop of the SVPWM algorithm, the U-phase output voltage modulation waveform is determined Phase angle Is it satisfied or ; When the trigger condition is met, the level state of the I / O pin is flipped and the preset pulse width is maintained. , and then returns to the initial level state to form a narrow pulse.

[0018] In a preferred embodiment, in the main loop of the SVPWM algorithm, the U-phase output voltage modulation waveform is determined. Phase angle The methods for determining whether the trigger conditions are met include: right The voltage reference component in the coordinate system is smoothed and preprocessed to obtain , the smoothing uses exponential moving average:

[0019] in, is the smoothing coefficient; Calculate phase and amplitude:

[0020]

[0021] and will Normalized to ; Set the amplitude threshold only when It is allowed to enter the phase judgment when It is the preset ratio of the DC bus per unit value; Set the angle window with hysteresis, Tolerance for triggering, To release the tolerance and meet , respectively define:

[0022] when When the trigger state is set to 0, Release 0 trigger state; when When π is set to trigger state, When π is released, the trigger state is released; Setting the rotation direction and continuity criterion: Calculation ,when The confirmation is made only when the direction of rotation is consistent with the expected direction and Q sampling cycles are continuously satisfied within the window. or ; When there is a fixed delay between sampling and updating When , feedforward compensation is used to make judgment correction:

[0023] and Alternative Participate in the above steps; In confirmation or After the event, the I / O pin is flipped in the current control interrupt and the preset pulse width is maintained. , as a flag output that can be triggered by an oscilloscope; in, , for The voltage reference component at the kth sampling moment in the coordinate system, , For , The smoothing component after the exponential moving average (EMA), k is the discrete sampling / control period index, The reference voltage vector is The instantaneous electrical phase angle in the coordinate system, is the reference voltage vector magnitude, is the angle normalization function, , is the phase deviation, which represents the angular difference relative to 0 and relative to π, , is the trigger / release threshold of the angle hysteresis window, is the discrete phase increment, Q is the sample number threshold of the persistence criterion, is the electrical frequency, is the electrical angular velocity, To consider the prediction phase with fixed delay, is the reference amplitude threshold, It is the I / O flag pulse width.

[0024] Step S2: Run the inverter and observe the position of the U-phase current through the oscilloscope. The position angle is the compensation angle. Connect oscilloscope channel CH1 to the U-phase current sensor output to observe the U-phase current. , connect channel CH2 to the I / O flag signal; Set CH2 edge trigger as external trigger source and collect at least one electrical cycle; Measure the CH2 flag triggering moment The moment when the U phase current passes through zero most recently The time difference between ; According to the output frequency of the inverter , calculate the time difference The corresponding phase angle is used as the initial compensation angle :

[0025] And normalize the function by angle Constrain it to :

[0026] Will As the initial compensation angle measurement value For subsequent compensation; in, is the instantaneous current of phase U, is the electrical frequency.

[0027] Step S3: Add the compensation angle to the software to obtain the current compensation angle as follows: the angle of the output voltage waveform + the output current angle + the compensation angle. Will Write control software parameters ; In each control cycle, the instantaneous voltage is obtained through Clarke / Park transformation With current ,calculate ; The current compensation angle is obtained according to the following formula and used to correct the SVPWM reference phase or sampling / update timing:

[0028] When there is a fixed delay When introducing To make compensation:

[0029] in, is the current compensation angle, is the angle normalization function, is the instantaneous phase angle of the output voltage reference, is the four-quadrant inverse tangent function, is the instantaneous phase angle of the output current, is the dead zone compensation angle, is the phase quantization of a fixed equivalent delay.

[0030] Step S4: Run the inverter again and use the oscilloscope to observe the position of the flag at the U-phase current to determine whether the position is at the 0 phase angle of the current: The current compensation angle obtained by loading , and re-operate the inverter; Use the oscilloscope again to observe the I / O flag signal and the U-phase current waveform at the same time; By observing the alignment of the trigger point of the I / O flag signal and the zero crossing point of the U-phase current waveform, and calculating the new time difference between the two ; like ≤preset error threshold , the compensation is successful, This is the final dead zone compensation angle, and adjustment stops; like >Error threshold , the compensation is not accurate and needs to be fine-tuned, that is, according to the new time difference Calculate correction angle , the calculation formula is:

[0031] in, is the electrical frequency, is the new time difference; In the inverter control software, the original dead zone compensation angle parameter Add or subtract correction angle , to get the new compensation angle , the direction of addition or subtraction depends on the new time difference The positive or negative value of the error is then re-run and the above iterative process is repeated until the error threshold is met. In a preferred embodiment, the frequency converter is based on the microprocessor clock frequency used. and SVPWM modulation frequency ,Will Set to meet a SVPWM minimum regulation period T min , where the minimum adjustment period T minDefined as 1 / and 1 / Integer multiple relationship.

[0032] In a preferred embodiment, the method for determining the addition and subtraction directions is: When the trigger point of the flag signal observed on the oscilloscope is ahead of the zero crossing point of the U phase current, it indicates that the compensation amount is too large and needs to be reduced. The correction angle is changed from the original dead zone compensation angle parameter to minus; When the trigger point of the flag signal observed on the oscilloscope lags behind the zero crossing point of the U phase current, it indicates that the compensation amount is insufficient and needs to be increased. Add.

[0033] In a preferred embodiment, the iterative process is specifically as follows: after obtaining a new compensation angle After that, the current compensation angle obtained by loading is returned. In this step, re-run the inverter and observe until the new time difference is determined. Meet the error threshold The iteration stops when the value is set, and the final compensation angle is locked and saved in the inverter parameters.

[0034] Reference Figure 4 and Figure 5 , is a schematic diagram of the process of converting the three-phase current U, V, W into controllable d, q currents. The first is the Clark transformation, which converts the three-phase U, V, W currents into α and β, as shown in Figure 4 The process from a) to b) can be expressed mathematically as follows:

[0035] in, , , The corresponding output is the current of U, V, W. is the peak current, is the output waveform angle / electrical angle; Then, by performing park transformation on α and β, we can get the dq axis current, which is Figure 4 From b) to c), the conversion process is as follows Figure 5 As shown; Park's transformation process can be expressed mathematically as follows:

[0036]

[0037] in, and The angle between the two is the current angle that needs to be obtained. and The inverse trigonometric function calculation can be used to obtain the angle of the output current. :

[0038] From the above, we can see that dead zone compensation includes electrical angle, lagging current phase angle and compensation angle.

[0039] It should be noted that dead zone compensation requires the insertion of 0° and 180° phase angles of the current waveform. Therefore, in the embodiment of the present invention, by making the dead zone compensation angle, the flag IO outputs a high level when it is 0 degrees and outputs a low level at other times. Because the waveform of the control output has a range, this embodiment adopts a range within ±2 degrees. The output current waveform of the inserted flag is as follows Figure 6 As shown, through Figure 6 From the observation of the current waveform, it can be seen that the insertion angle is at -90 degrees.

[0040] When the dead zone compensation angle is 0, the flag is inserted. The output current waveform of the inserted flag is as follows: Figure 7 As shown, through Figure 7 From the observation, it can be seen that the flag is exactly at the 0-point phase of the output current waveform, and this point is the required dead zone compensation angle; if the flag is not at the 0-point phase of the current waveform, continue to adjust the compensation value until the flag is at the 0-point phase of the current waveform.

[0041] It is not difficult to understand that when a phase of the sine wave output passes through the zero point, the current is 0, that is, it should be in the off state. During normal output, the current should increase or decrease according to the sinusoidal law in the non-zero state. Therefore, the waveform outside the dead time should compensate for the IGBT or MOS turn-on loss time caused by the dead time. Within the dead time, it should be compensated according to the sinusoidal conversion. Because the time is very short, it can also be approximately compensated according to a straight line.

[0042] Further, Figure 8 It is the no-load current output of EV520 inverter before 5Hz compensation; Figure 9 is the output current waveform after dead zone compensation; Figure 10 This is the current waveform of the rated 50Hz output after dead zone compensation. It can be seen intuitively that the method of the present invention uses an oscilloscope to intuitively find and determine the optimal compensation angle, thereby achieving accurate compensation for the SVPWM dead zone, reducing the current THD in the low-frequency and light-load areas, reducing torque ripple, and reducing the risks of zero-crossing oscillation and DC bias.

[0043] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for accurate dead zone compensation of SVPWM based on oscilloscope phase marking, characterized in that: include: Step S1: In the inverter control software, set the compensation angle to 0 and insert the IO conversion flag at the 0th phase of the U-phase wave; Step S2: Run the frequency converter and observe the position of the U-phase current through an oscilloscope. The position angle obtained is the compensation angle. Step S3, adding the compensation angle to the software to obtain a current compensation angle of: the angle of the output voltage waveform + the output current angle + the compensation angle; Step S4: Run the inverter again, observe the position of the flag at the U-phase current through the oscilloscope, and determine whether the position is at the 0 phase angle of the current: If so, the position at the 0 phase angle of the current is the dead zone compensation angle; If not, continue adjusting the current compensation angle value and repeat step S3 until the flag is at the zero-point phase of the current waveform.

2. The SVPWM dead zone accurate compensation method based on oscilloscope phase marking according to claim 1, characterized in that: The step S1 comprises: In the inverter control software, initialize the dead zone compensation angle parameter to ; In the SVPWM algorithm, the zero phase point of the U-phase output voltage modulation waveform is located, which is recorded as ; exist At the corresponding interrupt moment, a digital I / O signal that can be captured externally is output. The I / O signal switches from a high level to a low level or from a low level to a high level at the 0 phase point to serve as a flag that can be triggered by an oscilloscope.

3. The SVPWM dead zone accurate compensation method based on oscilloscope phase marking according to claim 2, characterized in that: The flag bits include: Setting the idle general I / O pins of the inverter to digital output mode; In the main loop of the SVPWM algorithm, the U-phase output voltage modulation waveform is determined Phase angle Is it satisfied or ; When the trigger condition is met, the level state of the I / O pin is flipped and the preset pulse width is maintained. , and then returns to the initial level state to form a narrow pulse.

4. The SVPWM dead zone accurate compensation method based on oscilloscope phase marking according to claim 3, characterized in that: The position angle obtained by observing the position of the U-phase current through the oscilloscope is the compensation angle, including: Connect oscilloscope channel CH1 to the U-phase current sensor output to observe the U-phase current. , connect channel CH2 to the I / O flag signal; Set CH2 edge trigger as external trigger source and collect at least one electrical cycle; Measure the CH2 flag triggering moment The moment when the U phase current passes through zero most recently The time difference between ; According to the output frequency of the inverter operation , calculate the time difference The corresponding phase angle is used as the initial compensation angle : And normalize the function by angle Constrain it to : Will As the initial compensation angle measurement value For subsequent compensation; in, is the instantaneous current of phase U, is the electrical frequency.

5. The SVPWM dead zone accurate compensation method based on oscilloscope phase marking according to claim 4, characterized in that: Adding the compensation angle to the software to obtain the current compensation angle includes: Will Write control software parameters ; In each control cycle, the instantaneous voltage is obtained through Clarke / Park transformation With current ,calculate ; The current compensation angle is obtained according to the following formula and used to correct the SVPWM reference phase or sampling / update timing: When there is a fixed delay When introducing To make compensation: in, is the current compensation angle, is the angle normalization function, is the instantaneous phase angle of the output voltage reference, is the four-quadrant inverse tangent function, is the instantaneous phase angle of the output current, is the dead zone compensation angle, is the phase quantization of a fixed equivalent delay.

6. The SVPWM dead zone accurate compensation method based on oscilloscope phase marking according to claim 5, characterized in that: Observing the position of the flag at the U-phase current through the oscilloscope and determining whether the position is at the 0 phase angle of the current includes: The current compensation angle obtained by loading , and re-operate the inverter; Using the oscilloscope again, simultaneously observing the I / O flag signal and the U-phase current waveform; By observing the alignment of the trigger point of the I / O flag signal and the zero-crossing point of the U-phase current waveform, and calculating the new time difference between the two ; If the ≤preset error threshold , the compensation is successful, the This is the final dead zone compensation angle, and adjustment stops; If the >The error threshold , the compensation is not accurate and needs to be fine-tuned, that is, according to the new time difference Calculate correction angle , the calculation formula is: in, is the electrical frequency, is the new time difference; In the inverter control software, the original dead zone compensation angle parameter Add or subtract the correction angle , to get the new compensation angle , the direction of addition and subtraction depends on the new time difference The positive or negative value of the error is then determined, and the frequency converter step is re-run, and the above iterative process is repeated until the error threshold is met and the iteration is stopped.

7. The SVPWM dead zone accurate compensation method based on oscilloscope phase marking according to claim 6, characterized in that: The preset error threshold ,include: According to the clock frequency of the microprocessor used by the inverter and SVPWM modulation frequency , the Set to meet a SVPWM minimum regulation period T min , wherein the minimum adjustment period T min Defined as 1 / and 1 / Integer multiple relationship.

8. The SVPWM dead zone accurate compensation method based on oscilloscope phase marking according to claim 6, characterized in that: The method for determining the addition and subtraction directions is as follows: When the trigger point of the flag signal observed on the oscilloscope is ahead of the zero crossing point of the U-phase current, it indicates that the compensation amount is too large and needs to be reduced. The correction angle is changed from the original dead zone compensation angle parameter minus; When the trigger point of the flag signal observed on the oscilloscope lags behind the zero crossing point of the U-phase current, it indicates that the compensation amount is insufficient and needs to be increased. The correction angle is compared with the original dead zone compensation angle parameter. Add.

9. The SVPWM dead zone accurate compensation method based on oscilloscope phase marking according to claim 6, characterized in that: The iterative process is specifically as follows: Getting a new compensation angle After that, the current compensation angle obtained by loading is returned. In this step, re-run the inverter and observe until the new time difference is determined. Satisfy the error threshold The iteration stops when the value is set, and the final compensation angle is locked and saved in the inverter parameters.

Images