Simulation method and system with controllable split-phase voltage drop time difference

Abstract

The invention provides a simulation method and system with controllable split-phase voltage drop time difference, and relates to the technical field of electric power testing, and the method comprises the steps: adjusting voltage through a three-phase voltage regulator, employing a bidirectional silicon controlled rectifier and IGBT hybrid switch structure to achieve voltage drop, collecting and processing a three-phase voltage signal through a digital signal processor, and extracting zero crossing point information; and a control pulse signal with phase compensation is generated, the conduction time sequence of the switch element is controlled to realize soft switching, and the industrial computer receives and processes the signal and displays a voltage drop waveform. The split-phase voltage drop time difference can be accurately controlled, and the simulation precision and the system reliability are improved.

Description

Technical Field

[0001] The present invention relates to power testing technology, and in particular to a simulation method and system for controllable time difference of phase-splitting voltage drop. Background Art

[0002] Voltage sags are a common power quality issue in power systems, severely impacting the normal operation of electrical equipment. Particularly in industrial production, voltage sags can cause sensitive equipment to shut down, control system malfunctions, and motor startup difficulties, resulting in production interruptions and economic losses. Therefore, simulating voltage sag events is crucial for studying electrical equipment immunity, assessing power system stability, and designing and testing related protection devices.

[0003] In voltage sag simulation technology, precisely controlling the depth, duration, and time difference between different phases of a voltage sag is a key requirement. Traditional voltage sag simulation methods primarily include electromagnetic voltage regulators, power amplifier drivers, and power electronic switches. Power electronic switches, due to their fast response speed and high control accuracy, have become a research hotspot.

[0004] However, existing voltage drop simulation technology has the following defects and deficiencies: Existing simulation methods often have large voltage transients and shocks during the switching process. Such mutations will introduce high-order harmonic interference and cannot accurately simulate the voltage drop process in the actual power system, affecting the reliability and accuracy of the test results.

[0005] Existing technologies make it difficult to achieve precise control of the time difference of voltage drops in different phases. Especially when simulating single-phase or two-phase voltage drops, they cannot accurately reflect the timing differences of voltage drops in each phase in the power system due to different fault properties, which limits the comprehensiveness and representativeness of the testing work.

[0006] The control system accuracy of existing voltage drop simulation equipment is insufficient, especially in phase voltage synchronization control and phase compensation. This leads to large differences between the simulated waveform and the actual grid fault waveform, reducing the practical value and reference value of the test results. Summary of the Invention

[0007] The embodiments of the present invention provide a simulation method and system with controllable phase voltage drop time difference, which can solve the problems in the prior art.

[0008] A first aspect of an embodiment of the present invention provides a simulation method for controllable phase voltage drop time difference, comprising: The voltage of the input three-phase AC power supply is regulated by a three-phase voltage regulator. The three-phase voltage regulator adopts a sliding contact structure and is driven and controlled by a DC motor to generate a three-phase voltage signal with a set voltage value; Inputting the three-phase voltage signals into the corresponding full-voltage power electronic switch and half-voltage power electronic switch, respectively, wherein the full-voltage power electronic switch and half-voltage power electronic switch of each phase adopt a hybrid switch structure of bidirectional thyristor and IGBT, the bidirectional thyristor is connected in series to the collector terminal of the IGBT, and the emitter terminal of the IGBT is connected in parallel with an RC snubber circuit; The three-phase voltage signals on the power supply side and the load side are collected and processed by a digital signal processor, the zero-crossing point information is extracted to calculate the phase voltage synchronization control signal, the phase voltage synchronization control signal is synchronized with the reference clock signal through a phase-locked loop, and a control pulse signal with phase compensation is generated; According to the control pulse signal, the bidirectional thyristor in the full-voltage power electronic switch is controlled to turn off, and after a preset delay, the IGBT in the full-voltage power electronic switch is controlled to turn off, and at the same time, the IGBT in the half-voltage power electronic switch is controlled to turn on. After the IGBT in the half-voltage power electronic switch is turned on stably, the bidirectional thyristor in the half-voltage power electronic switch is controlled to turn on, so as to realize soft switching during the voltage drop process; An industrial computer is used to receive the signal transmitted by the digital signal processor through a high-speed communication interface, and calculate and display the voltage drop waveform.

[0009] In an optional embodiment, The steps of collecting and processing three-phase voltage signals on the power supply side and the load side by a digital signal processor, extracting zero-crossing point information to calculate a phase voltage synchronization control signal, performing phase-locked loop synchronization processing on the phase voltage synchronization control signal and a reference clock signal to generate a control pulse signal with phase compensation include: The three-phase voltage signals on the power supply side and the load side are conditioned by a differential amplifier circuit, and the output of the differential amplifier circuit is subjected to a second-order Butterworth low-pass filter and then sampled by an analog-to-digital conversion circuit to obtain a digital voltage signal; Performing a three-point linear interpolation operation on the digital voltage signal, calculating a zero-crossing time based on the voltage values ​​of three adjacent sampling points and the corresponding sampling times, and setting a positive hysteresis threshold and a negative hysteresis threshold, determining that a zero-crossing time occurs when the digital voltage signal crosses the positive hysteresis threshold from negative to positive or crosses the negative hysteresis threshold from positive to negative; Calculating a phase voltage synchronization control signal based on a zero-crossing point, generating a reference clock signal through a phase accumulator, inputting the phase voltage synchronization control signal and the reference clock signal into a digital phase-locked loop circuit, the digital phase-locked loop circuit comprising a phase detector, a second-order loop filter, and a digitally controlled oscillator, wherein a proportional coefficient and an integral coefficient of the second-order loop filter are determined according to a system bandwidth and a damping ratio, and the digital phase-locked loop circuit outputs a synchronization phase signal; The synchronous phase signal is compensated according to the transmission delay characteristics of the switching device, including fixed delay compensation and dynamic characteristic compensation, and a control pulse signal is generated by adopting a space vector modulation algorithm according to the compensated phase angle.

[0010] In an optional embodiment, The step of compensating according to the transmission delay characteristics of the switching device includes: Inputting the synchronized phase signal into a dead zone compensation unit, calculating the dead zone time according to the turn-off delay time of the switching device, and introducing a transition level within the dead zone time to suppress voltage spikes during the switching process, thereby obtaining a phase angle based on fixed delay compensation; Inputting the phase angle based on the fixed delay compensation into the dynamic characteristic compensation unit, and calculating the compensation amount of the turn-on process and the turn-off process respectively based on the asymmetric characteristics of the rise time and fall time of the switching device to obtain the compensated phase angle; The output voltage distortion rate and switching loss corresponding to the control pulse signal generated based on the compensated phase angle are monitored in real time, and a comprehensive optimization objective function including the output voltage distortion rate and the switching loss is constructed. When the output voltage distortion rate exceeds a preset distortion threshold, the upper bridge arm dead time, the lower bridge arm dead time, the transition level threshold, the turn-on process compensation coefficient, and the turn-off process compensation coefficient are encoded into an optimization parameter vector, and the optimization parameter vector is iteratively optimized using a particle swarm algorithm with an adaptive inertia weight, wherein the adaptive inertia weight decreases with the square of the number of iterations. During the optimization process, an asymmetry constraint is imposed on the dead zone compensation parameter, a variation range constraint is imposed on the dynamic characteristic compensation amount, and a smoothing factor is used to perform a weighted combination of new and old compensation parameters to update the compensated phase angle.

[0011] In an optional embodiment, The steps of controlling the bidirectional thyristor in the full-voltage power electronic switch to turn off according to the control pulse signal, controlling the IGBT in the full-voltage power electronic switch to turn off after a preset delay, and simultaneously controlling the IGBT in the half-voltage power electronic switch to turn on, and controlling the bidirectional thyristor in the half-voltage power electronic switch to turn on after the IGBT in the half-voltage power electronic switch is stably turned on, to achieve soft switching during voltage drop include: Obtaining the on-state voltage drop, on-state resistance, and through-current of the bidirectional thyristor to calculate the on-state characteristic parameters of the bidirectional thyristor; obtaining the collector-emitter voltage and collector current of the IGBT to calculate the switching loss parameters of the IGBT; and determining a preset delay time of the hybrid switch based on the on-state characteristic parameters and the switching loss parameters; Controlling the hybrid switch to perform a soft switching operation according to the control pulse signal, the soft switching operation comprising: controlling the bidirectional thyristor on the full-voltage side to turn off, controlling the IGBT on the full-voltage side to turn off after a preset delay time, and simultaneously controlling the IGBT on the half-voltage side to turn on, and controlling the bidirectional thyristor on the half-voltage side to turn on after the IGBT on the half-voltage side is stably turned on; During the soft switching process, a voltage stress index and a current stress index of the hybrid switch are calculated in real time, and a switch stress compensation angle is calculated, wherein the voltage stress index is obtained by the root mean square value of the collector-emitter voltage, and the current stress index is obtained by the root mean square value of the collector current; The control pulse signal is subjected to real-time dynamic phase compensation according to the switch stress compensation angle to generate a stress-optimized control pulse signal; the conduction timing of the hybrid switch is adjusted based on the stress-optimized control pulse signal to achieve stress-controlled soft switching control during voltage drop.

[0012] In an optional embodiment, During the soft switching process, the voltage stress index and the current stress index of the hybrid switch are calculated in real time. The steps of calculating the switch stress compensation angle include: A sliding time window is used to sample the collector-emitter voltage and collector current, and a real-time voltage stress index and a real-time current stress index are calculated based on the sampled data. The calculation period of the real-time voltage stress index and the real-time current stress index is an integer multiple of the switching period. Inputting the real-time voltage stress index and the real-time current stress index into an adaptive fuzzy neural network, wherein the adaptive fuzzy neural network includes a stress assessment layer and a compensation decision layer, wherein the stress assessment layer constructs a fuzzy rule base based on expert experience, and the compensation decision layer uses an online learning algorithm to update the network weight; Determining a reference value of a switch stress compensation angle according to the output of the adaptive fuzzy neural network and introducing a dynamic limiting mechanism; The compensation effect of the switch stress compensation angle is monitored in real time, and a comprehensive evaluation index including compensation accuracy and compensation smoothness is constructed. When the comprehensive evaluation index is lower than a preset evaluation threshold, the weight coefficient of the fuzzy rule is updated based on the recursive least squares method.

[0013] In an optional embodiment, The step of inputting the real-time voltage stress index and the real-time current stress index into the adaptive fuzzy neural network comprises: Inputting the real-time voltage stress index and the real-time current stress index into the stress assessment layer of the adaptive fuzzy neural network, the stress assessment layer maps the real-time voltage stress index and the real-time current stress index into multiple language variables, and constructing a fuzzy rule base based on expert experience, including a basic rule group and a dynamic rule group. The basic rule group includes multiple expert experience rules, and the dynamic rule group adaptively generates rules according to the operating status; Calculate the rule credibility based on historical operation data. The rule credibility is determined by the ratio of the number of successful rule applications to the total number of rule applications. The rule priority is calculated based on the rule credibility and the rule execution effect evaluation value. The rule priority is updated using exponential smoothing. Inputting the rule activation strength of the stress assessment layer into the compensation decision layer, the compensation decision layer adopts a variable structure neuron design, and updates the network weights through an online learning algorithm. The learning rate of the online learning algorithm decays exponentially with the number of iterations and a minimum learning rate is set; The fuzzy rule base is dynamically optimized, including: calculating rule strength, where the rule strength is determined based on the control error change rate and updated through a forgetting factor mechanism; merging rules when rule similarity exceeds a preset similarity threshold, increasing rule accuracy when the control error exceeds a preset error threshold, and deleting rules when the rule is not activated within a preset period and the rule strength is lower than a preset strength threshold.

[0014] In an optional embodiment, The dynamic limiting mechanism includes: collecting the switch voltage and switch current, calculating the root mean square value of the switch voltage change rate and the switch current change rate within the sliding sampling window to obtain the voltage stress characteristic and the current stress characteristic, calculating the thermal stress characteristic based on the switch junction temperature and the switch power loss, and constructing the voltage stress characteristic, the current stress characteristic, and the thermal stress characteristic into a stress characteristic vector; calculating a state transition relationship and a control action relationship based on a historical data sequence of the stress characteristic vector, predicting a stress characteristic vector of the next cycle based on the state transition relationship and the control action relationship, and calculating a stress overlimit probability based on the predicted stress characteristic vector; A clipping boundary curve is constructed based on a hyperbolic tangent function, wherein the amplitude coefficient of the clipping boundary curve is adaptively adjusted according to the stress over-limit probability, the slope coefficient of the clipping boundary curve is proportional to the second-order derivative of the compensation angle, and the current compensation angle is compared with the clipping boundary curve to obtain an initial clipping result; Constructing a stress balance evaluation function, wherein the stress balance evaluation function calculates an adjustment coefficient based on a balance degree of the three-dimensional stress characteristics, and correcting the initial clipping result according to the adjustment coefficient to obtain a clipping compensation angle; The limiting compensation angle is input into a dynamic buffer, the smoothing coefficient of the dynamic buffer is in an exponential decay relationship with the compensation angle change rate, and the original compensation angle and the buffer compensation angle are weightedly combined according to the smoothing coefficient to obtain the final compensation angle.

[0015] A second aspect of an embodiment of the present invention provides a simulation system with controllable phase voltage drop time difference, comprising: The first unit is used to regulate the voltage of the input three-phase AC power supply through a three-phase voltage regulator. The three-phase voltage regulator adopts a sliding contact structure and is driven and controlled by a DC motor to generate a three-phase voltage signal with a set voltage value; The second unit is used to input the three-phase voltage signals into the corresponding full-voltage power electronic switch and half-voltage power electronic switch, respectively. The full-voltage power electronic switch and half-voltage power electronic switch of each phase adopt a hybrid switch structure of bidirectional thyristor and IGBT, wherein the bidirectional thyristor is connected in series to the collector terminal of the IGBT, and the emitter terminal of the IGBT is connected in parallel with an RC snubber circuit; The third unit is used to collect and process the three-phase voltage signals on the power supply side and the load side through a digital signal processor, extract zero-crossing point information to calculate the phase voltage synchronization control signal, perform phase-locked loop synchronization processing on the phase voltage synchronization control signal and the reference clock signal, and generate a control pulse signal with phase compensation; The fourth unit is used for the digital signal processor to control the conduction timing of the bidirectional thyristor and the IGBT according to the control pulse signal, control the bidirectional thyristor in the full-voltage power electronic switch to turn off, control the IGBT in the full-voltage power electronic switch to turn off after a preset delay, and simultaneously control the IGBT in the half-voltage power electronic switch to turn on, and control the bidirectional thyristor in the half-voltage power electronic switch to turn on after the IGBT in the half-voltage power electronic switch is turned on stably, so as to realize soft switching during voltage drop; The fifth unit is used to use an industrial computer to receive the signal transmitted by the digital signal processor through a high-speed communication interface, and calculate and display the voltage drop waveform.

[0016] According to a third aspect of an embodiment of the present invention, an electronic device is provided, including: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0017] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0018] The present invention realizes the precise and controllable simulation of the phase voltage drop time difference, and solves the technical problem that the traditional method is difficult to accurately control the phase voltage drop time difference. By combining the digital signal processor synchronous control technology with the hybrid switch structure, it can realize soft switching during the grid voltage drop process, effectively suppress the transient impact generated during the voltage drop process, and improve the safety performance of the simulation system.

[0019] The bidirectional thyristor and IGBT hybrid switch structure adopted in the present invention, combined with the RC snubber circuit design, realizes soft start and soft shutdown during the voltage drop switching process, effectively reduces voltage waveform distortion, and significantly reduces electromagnetic interference during the switching process, ensuring the quality of the simulated voltage drop waveform and system stability.

[0020] The present invention combines a sliding contact three-phase voltage regulator with a digital control system to achieve flexible adjustment of voltage amplitude and drop time, accurately simulate various complex voltage drop scenarios, meet the power equipment testing requirements under different working conditions, and provide a reliable testing method for the safe and stable operation of the power system and the evaluation of the equipment's ability to resist voltage disturbances. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 Schematic diagram of the flow of a simulation method for controllable phase voltage drop time difference according to an embodiment of the present invention; Figure 2 The following is a comparison chart of output voltage distortion rate and switching loss of different compensation methods; Figure 3 The stress reduction rate comparison chart of three compensation methods during the soft switching process of hybrid switch. DETAILED DESCRIPTION

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0023] The following specific embodiments are used to describe the technical solution of the present invention in detail. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0024] Figure 1 FIG. 1 is a flow chart of a simulation method for controllable phase voltage drop time difference according to an embodiment of the present invention. Figure 1 As shown, the method includes: The voltage of the input three-phase AC power supply is regulated by a three-phase voltage regulator. The three-phase voltage regulator adopts a sliding contact structure and is driven and controlled by a DC motor to generate a three-phase voltage signal with a set voltage value; Inputting the three-phase voltage signals into the corresponding full-voltage power electronic switch and half-voltage power electronic switch, respectively, wherein the full-voltage power electronic switch and half-voltage power electronic switch of each phase adopt a hybrid switch structure of bidirectional thyristor and IGBT, the bidirectional thyristor is connected in series to the collector terminal of the IGBT, and the emitter terminal of the IGBT is connected in parallel with an RC snubber circuit; The three-phase voltage signals on the power supply side and the load side are collected and processed by a digital signal processor, the zero-crossing point information is extracted to calculate the phase voltage synchronization control signal, the phase voltage synchronization control signal is synchronized with the reference clock signal through a phase-locked loop, and a control pulse signal with phase compensation is generated; The digital signal processor controls the conduction timing of the bidirectional thyristor and the IGBT according to the control pulse signal, controls the bidirectional thyristor in the full-voltage power electronic switch to turn off, controls the IGBT in the full-voltage power electronic switch to turn off after a preset delay, and simultaneously controls the IGBT in the half-voltage power electronic switch to turn on, and controls the bidirectional thyristor in the half-voltage power electronic switch to turn on after the IGBT in the half-voltage power electronic switch is turned on stably, thereby realizing soft switching during voltage drop; An industrial computer is used to receive the signal transmitted by the digital signal processor through a high-speed communication interface, and calculate and display the voltage drop waveform.

[0025] In an optional embodiment, the steps of collecting and processing three-phase voltage signals on the power supply side and the load side by a digital signal processor, extracting zero-crossing information to calculate a phase voltage synchronization control signal, performing phase-locked loop synchronization processing on the phase voltage synchronization control signal and a reference clock signal to generate a control pulse signal with phase compensation include: The three-phase voltage signals on the power supply side and the load side are conditioned by a differential amplifier circuit, and the output of the differential amplifier circuit is subjected to a second-order Butterworth low-pass filter and then sampled by an analog-to-digital conversion circuit to obtain a digital voltage signal; performing a three-point linear interpolation operation on the digital voltage signal, calculating a zero-crossing time based on the voltage values ​​of three adjacent sampling points and the corresponding sampling times, and setting a positive hysteresis threshold and a negative hysteresis threshold, determining a zero-crossing point when the digital voltage signal crosses the positive hysteresis threshold from negative to positive or crosses the negative hysteresis threshold from positive to negative, wherein the absolute values ​​of the positive hysteresis threshold and the negative hysteresis threshold are both 0.5% of the rated voltage; Calculating a phase voltage synchronization control signal based on a zero-crossing point, generating a reference clock signal through a phase accumulator, inputting the phase voltage synchronization control signal and the reference clock signal into a digital phase-locked loop circuit, the digital phase-locked loop circuit comprising a phase detector, a second-order loop filter, and a digitally controlled oscillator, wherein a proportional coefficient and an integral coefficient of the second-order loop filter are determined according to a system bandwidth and a damping ratio, and the digital phase-locked loop circuit outputs a synchronization phase signal; The synchronous phase signal is compensated according to the transmission delay characteristics of the switching device, including fixed delay compensation and dynamic characteristic compensation, and a control pulse signal is generated by adopting a space vector modulation algorithm according to the compensated phase angle.

[0026] For example, in the signal acquisition process, the differential amplifier circuit uses the high-precision operational amplifier LT1167, with a differential gain set to 0.01 and a common-mode rejection ratio greater than 90dB, effectively suppressing common-mode interference signals. The output signal of the differential amplifier circuit is processed by a second-order Butterworth low-pass filter with a cutoff frequency set to 1kHz to fully filter out high-frequency noise while retaining the basic waveform characteristics of the signal. The filtered signal is sampled by the analog-to-digital conversion circuit, using a 16-bit AD converter AD7656 with a sampling frequency set to 10kHz to ensure that at least 200 sampling points are obtained within each power frequency cycle, meeting the requirements of high-precision phase detection. The analog-to-digital conversion uses bipolar conversion with a reference voltage of ±10V, corresponding to a digital value range of -32768 to 32767. The conversion result is transmitted to the digital signal processor via the SPI bus.

[0027] In the zero-crossing detection phase, a three-point linear interpolation operation is performed on the digital voltage signal to calculate the precise zero-crossing instant. When a sign change is detected between adjacent sampling points, a zero-crossing is determined. Assume three consecutive sampling points are (t1, V1), (t2, V2), and (t3, V3), where V1 < 0 and V2 > 0, indicating a voltage crossing zero from negative to positive. Three-point linear interpolation can be used to obtain a more precise zero-crossing instant t0. In actual operation, under a 10kHz sampling frequency and a 50Hz power grid, test results show that this method can achieve zero-crossing detection accuracy within ±5μs. To enhance interference immunity, positive and negative hysteresis thresholds are set, each with an absolute value of 0.5% of the rated voltage, or ±0.5V (for a 220V rated voltage system). A valid zero-crossing is determined when the voltage crosses the +0.5V threshold from negative to positive or the -0.5V threshold from positive to negative.

[0028] In the phase synchronization control phase, the phase voltage synchronization control signal is calculated based on the detected zero-crossing time. Simultaneously, a reference clock signal is generated via a phase accumulator. The phase increment of the phase accumulator is set to 0.00314 rad, corresponding to a 50 Hz power frequency reference signal. The phase voltage synchronization control signal and the reference clock signal are input into a digital phase-locked loop circuit for synchronization. This digital phase-locked loop circuit comprises three components: a phase detector, a second-order loop filter, and a digitally controlled oscillator. The phase detector is implemented using phase difference multiplication with an output range of ±π. The second-order loop filter uses a proportional-integral structure, with the proportional coefficient kp set to 0.2 and the integral coefficient ki set to 0.02. These parameters are determined based on the system bandwidth of 10 Hz and the damping ratio of 0.707 to ensure a balance between system response speed and stability. The digitally controlled oscillator is implemented using a 32-bit phase accumulator with a frequency resolution of 0.001 Hz.

[0029] During phase compensation and pulse generation, the synchronous phase signal is compensated for the transmission delay characteristics of the switching devices. This compensation is divided into two parts: fixed delay compensation and dynamic characteristic compensation. The fixed delay compensation value is 4μs, which is used to compensate for the turn-on delay of the IGBT device. Dynamic characteristic compensation is dynamically adjusted based on the switching current, with a compensation value range of 0-2μs. The compensation curve is generated using a lookup table, with the compensation value increasing by 0.2μs for every 10A increase in current. The compensated phase angle is used to generate the control pulse signal using the space vector modulation algorithm. Space vector modulation uses a seven-segment modulation scheme, with a carrier frequency set to 10kHz and a dead time of 2μs. Minimum harmonic output is achieved by evenly dividing the effective vector action time.

[0030] This invention processes voltage signals through a differential amplifier circuit and a Butterworth filter, combining three-point linear interpolation with a hysteresis threshold mechanism to accurately identify zero-crossing points, significantly improving the accuracy of phase voltage synchronization control signals. The introduction of a digital phase-locked loop circuit ensures phase synchronization stability in the presence of grid disturbances and waveform distortion, avoiding the phase tracking errors of traditional synchronization methods under non-ideal grid conditions. Phase compensation based on switching characteristics combined with a space vector modulation algorithm effectively reduces timing errors in control pulses.

[0031] In an optional implementation, the step of compensating according to the transmission delay characteristics of the switching device includes: Inputting the synchronized phase signal into a dead zone compensation unit, calculating the dead zone time according to the turn-off delay time of the switching device, and introducing a transition level within the dead zone time to suppress voltage spikes during the switching process, thereby obtaining a phase angle based on fixed delay compensation; Inputting the phase angle based on the fixed delay compensation into the dynamic characteristic compensation unit, and calculating the compensation amount of the turn-on process and the turn-off process respectively based on the asymmetric characteristics of the rise time and fall time of the switching device to obtain the compensated phase angle; The output voltage distortion rate and switching loss corresponding to the control pulse signal generated based on the compensated phase angle are monitored in real time, and a comprehensive optimization objective function including the output voltage distortion rate and the switching loss is constructed, wherein the output voltage distortion rate is calculated based on the effective value of the fundamental component and the effective value of each harmonic component, and the switching loss is calculated based on the turn-on energy loss and turn-off energy loss of the switching device. When the output voltage distortion rate exceeds a preset distortion threshold, the upper arm dead time, the lower arm dead time, the transition level threshold, the turn-on process compensation coefficient, and the turn-off process compensation coefficient are encoded into an optimization parameter vector, and the optimization parameter vector is iteratively optimized using a particle swarm algorithm with an adaptive inertia weight, wherein the adaptive inertia weight decreases with the square of the number of iterations. During the optimization process, an asymmetry constraint is imposed on the dead zone compensation parameter, a variation range constraint is imposed on the dynamic characteristic compensation amount, and a smoothing factor is used to perform a weighted combination of new and old compensation parameters to suppress parameter mutations, thereby updating the compensated phase angle.

[0032] Exemplarily, a synchronous phase signal is received as input in the dead-band compensation unit. The dead-band time is calculated based on the turn-off delay time of the switching device. This turn-off delay time can be obtained through actual measurement. For example, for a specific model of MOSFET switching device, its turn-off delay time is 120ns. The dead-band time is typically set to 1.2 to 1.5 times the turn-off delay time, and in this embodiment, it is set to 150ns. Within the dead-band time, a transition level between a high level and a low level is introduced. The initial transition level threshold can be set to 50% of the power supply voltage. For example, when the power supply voltage is 400V, the transition level threshold is set to 200V. The introduction of this transition level can effectively suppress voltage spikes caused by rapid voltage changes during the switching process. Actual measurements show that after introducing the transition level, the voltage spike can be reduced from 650V to 480V, a reduction of approximately 26%. Through this step, a phase angle based on fixed delay compensation can be obtained. The initial compensation amount is typically set between 2° and 5°, and in this embodiment, it is set to 3.5°.

[0033] The phase angle calculated based on the fixed delay compensation is then input into the dynamic characteristics compensation unit. Due to the asymmetry between the rise and fall times of the switching device, compensation must be calculated separately for the turn-on and turn-off processes. For example, a certain MOSFET model has a rise time of 80ns and a fall time of 60ns. This asymmetry can cause distortion in the actual output waveform. The initial value for the turn-on compensation can be set to the phase angle corresponding to 30% of the rise time, approximately 0.86°; the initial value for the turn-off compensation can be set to the phase angle corresponding to 25% of the fall time, approximately 0.54°. These compensations yield a compensated phase angle that more accurately reflects the actual switching behavior of the switching device.

[0034] Real-time monitoring of the performance indicators of the control pulse signal generated based on the compensated phase angle. The output voltage distortion rate is a key indicator, which is calculated based on the effective value of the fundamental component and the effective value of each harmonic component. Specifically, assuming that the measured effective value of the fundamental component is 380V, the effective value of the third harmonic is 15V, the effective value of the fifth harmonic is 8V, and the effective value of the seventh harmonic is 4V, the total harmonic distortion rate is approximately 4.5%. Another important indicator is switching loss, which includes turn-on energy loss and turn-off energy loss. For example, at a switching frequency of 1kHz, the energy loss for each turn-on is approximately 0.25mJ, the energy loss for each turn-off is approximately 0.18mJ, and the total switching loss is 0.43W.

[0035] When the output voltage distortion exceeds a preset threshold (e.g., 4%), the compensation parameter optimization process is initiated. The optimized parameter vector includes five key parameters: upper-arm dead time, lower-arm dead time, transition level threshold, turn-on compensation coefficient, and turn-off compensation coefficient. Initial parameters can be set as follows: upper-arm dead time 150ns, lower-arm dead time 150ns, transition level threshold 50%, turn-on compensation coefficient 0.3, and turn-off compensation coefficient 0.25.

[0036] The optimization process uses a particle swarm algorithm with adaptive inertia weight. The initial value of the adaptive inertia weight is set to 0.9, which decreases with the square of the number of iterations and can be reduced to 0.4. For example, at the 10th iteration, the inertia weight is 0.9-0.5×(10 / 100) 2 =0.85; at the 50th iteration, the inertia weight drops to 0.65. In the algorithm, the number of particles is set to 20 and the maximum number of iterations is 100. Each particle represents a set of possible compensation parameter combinations.

[0037] During the optimization process, parameter constraints were imposed to ensure system stability. A dead-band compensation parameter asymmetry constraint ensured that the ratio of the upper and lower dead-band times remained between 0.8 and 1.2. A dynamic compensation range constraint limited the compensation coefficient during the turn-on phase to between 0.2 and 0.4, and the compensation coefficient during the turn-off phase to between 0.15 and 0.35. To prevent system instability caused by sudden parameter changes, a smoothing factor α = 0.8 was used to weight the old and new compensation parameters: the final parameter = α × old parameter + (1-α) × new parameter.

[0038] After optimization, the parameters are adjusted to: upper arm dead time 168ns, lower arm dead time 142ns, transition level threshold 45%, turn-on process compensation coefficient 0.32, and turn-off process compensation coefficient 0.28.

[0039] Figure 2 The figure shows a comparison of the output voltage distortion rate and switching loss of different compensation methods. The horizontal axis of the graph represents the switching frequency (kHz), the vertical axis on the left represents the voltage distortion rate (%), and the vertical axis on the right represents the switching loss (W). In terms of voltage distortion rate, the method of the present invention shows significant advantages, and this advantage becomes more obvious as the switching frequency increases. In terms of switching loss, the present invention also shows excellent performance. In the full frequency range from 1kHz to 11kHz, the method of the present invention maintains obvious advantages in both voltage distortion rate and switching loss. As the switching frequency increases, the advantages of the method of the present invention over the other two methods become more significant, indicating that it has better adaptability in high-frequency application scenarios.

[0040] The dual compensation strategy adopted by the present invention successfully solves the problems caused by the non-ideal characteristics of switching devices during the actual switching process. By adaptively optimizing the compensation parameters, the method of the present invention can reduce voltage distortion while reducing switching losses, achieving an excellent balance between output waveform quality and system efficiency, and has important practical value for improving the performance of power electronic switching systems. The dynamic characteristic compensation unit compensates for the asymmetric characteristics of the turn-on and turn-off processes respectively, reducing waveform distortion at the switching moment. The comprehensive optimization objective function considers the voltage distortion rate and switching losses in a unified manner, and the adaptive particle swarm algorithm can dynamically optimize the compensation parameters during operation, achieving a balance between output waveform quality and system efficiency.

[0041] In an optional embodiment, according to the control pulse signal, the bidirectional thyristor in the full-voltage power electronic switch is controlled to turn off, the IGBT in the full-voltage power electronic switch is controlled to turn off after a preset delay, and the IGBT in the half-voltage power electronic switch is controlled to turn on at the same time, and after the IGBT in the half-voltage power electronic switch is stably turned on, the bidirectional thyristor in the half-voltage power electronic switch is controlled to turn on to achieve soft switching during voltage drop, including: Obtaining the on-state voltage drop, on-state resistance, and through-current of the bidirectional thyristor to calculate the on-state characteristic parameters of the bidirectional thyristor; obtaining the collector-emitter voltage and collector current of the IGBT to calculate the switching loss parameters of the IGBT; and determining a preset delay time of the hybrid switch based on the on-state characteristic parameters and the switching loss parameters; Controlling the hybrid switch to perform a soft switching operation according to the control pulse signal, the soft switching operation comprising: controlling the bidirectional thyristor on the full-voltage side to turn off, controlling the IGBT on the full-voltage side to turn off after a preset delay time, and simultaneously controlling the IGBT on the half-voltage side to turn on, and controlling the bidirectional thyristor on the half-voltage side to turn on after the IGBT on the half-voltage side is stably turned on; During the soft switching process, a voltage stress index and a current stress index of the hybrid switch are calculated in real time, and a switch stress compensation angle is calculated, wherein the voltage stress index is obtained by the root mean square value of the collector-emitter voltage, and the current stress index is obtained by the root mean square value of the collector current; The control pulse signal is subjected to real-time dynamic phase compensation according to the switch stress compensation angle to generate a stress-optimized control pulse signal; the conduction timing of the hybrid switch is adjusted based on the stress-optimized control pulse signal to achieve stress-controlled soft switching control during voltage drop.

[0042] For example, the voltage drop value of the bidirectional thyristor in the on-state is measured. In actual application scenarios, the voltage drop value is generally between 0.8V and 1.2V. At the same time, the on-resistance at both ends of the bidirectional thyristor is measured, and the typical value is 0.01 ohm to 0.05 ohm. The current value flowing through the bidirectional thyristor is recorded, which may vary between 10A and 100A depending on the application scenario. Based on these collected data, the conduction characteristic parameters of the bidirectional thyristor are calculated, including the power loss coefficient and temperature coefficient. In actual applications, the power loss coefficient is generally 0.01 to 0.03W / A, and the temperature coefficient is 0.005 to 0.01Ω / ℃.

[0043] The IGBT's operating parameters are obtained by measuring the collector-emitter voltage, which typically varies between 0 and 600V during the switching process. The collector current is also recorded, which is typically 5A to 50A in applications. Based on these measured data, the IGBT's switching loss parameters, including turn-on and turn-off losses, are calculated. In practical applications, typical turn-on losses are 0.5mJ to 2.0mJ, and typical turn-off losses are 1.0mJ to 3.0mJ.

[0044] Based on the conduction characteristic parameters and switching loss parameters obtained above, the preset delay time of the hybrid switch is determined. This delay time must be long enough to ensure that the full-voltage side bidirectional thyristor is completely turned off, but not too long to cause response delay. Through analysis of experimental data, under operating conditions of 380V voltage and 30A current, the optimal preset delay time is 50μs to 100μs. In practice, this delay time can be dynamically adjusted between 40μs and 150μs for different load conditions.

[0045] After receiving the control pulse signal, soft switching is performed. First, the bidirectional thyristor (TRIAC) on the full-voltage side is turned off by applying a reverse control signal to the control terminal of the TRIAC. After a preset delay (e.g., 80μs), the IGBT on the full-voltage side is turned off by reducing the IGBT gate voltage from 15V to -5V. Simultaneously, the IGBT on the half-voltage side is turned on by increasing the gate voltage from -5V to 15V. The collector current and collector-emitter voltage of the half-voltage side IGBT are monitored. When the collector current stabilizes above 95% of the set value and the collector-emitter voltage drops below 2V, the IGBT is considered stable. The bidirectional thyristor on the half-voltage side is then turned on, completing the entire soft switching process. This process typically completes within 200μs to 400μs.

[0046] During soft switching, the voltage and current stress indicators of the hybrid switch are calculated in real time. The collector-emitter voltage of the hybrid switch is sampled using a sliding time window technique at a sampling frequency of 50kHz. The sliding window length is set to 10 switching cycles, with 100 data points collected per switching cycle. Based on this sampled data, the RMS value of the collector-emitter voltage is calculated as the voltage stress indicator. For example, at a 380V operating voltage, the voltage stress indicator typically ranges from 420V to 460V. The current stress indicator is obtained by sampling the IGBT collector current and calculating its RMS value. Similarly, current data is collected every 10μs, and the RMS value is calculated from 100 sampling points. Under a 30A load condition, the current stress indicator typically ranges from 33A to 37A.

[0047] Based on the calculated voltage and current stress indicators, the switch stress compensation angle is calculated. This compensation angle is proportional to the amount by which the stress exceeds its threshold. For example, when the voltage stress indicator exceeds the preset threshold of 420V to reach 450V (exceeding 30V), the compensation angle is approximately 1.5 degrees, calculated using a proportionality factor of 0.05 degrees / volt. When the current stress indicator exceeds the preset threshold of 32A to reach 35A (exceeding 3A), the compensation angle is approximately 0.3 degrees, calculated using a proportionality factor of 0.1 degrees / ampere. If both stresses exceed their thresholds simultaneously, the larger calculated compensation angle is used.

[0048] The control pulse signal is dynamically phase-compensated in real time based on the calculated switch stress compensation angle, generating a stress-optimized control pulse signal. For a 50Hz power frequency, a 1-degree phase angle corresponds to approximately 55.6μs of adjustment time. For a 2-degree compensation angle, the control pulse signal must be issued approximately 111.2μs earlier. A phase-shifting circuit achieves precise phase adjustment with an accuracy of 0.1 degree, or approximately 5.6μs.

[0049] Finally, the hybrid switch's turn-on timing is adjusted based on the stress-optimized control pulse signal. In actual operation, the full-voltage-side bidirectional thyristor's turn-off time is advanced by 2 to 5 degrees, and the IGBT switching timing is adjusted accordingly. The turn-on timing of the half-voltage-side components is also optimized to ensure a smooth transition. This stress-controlled soft switching strategy results in a smoother voltage drop, effectively controls switch stress, and significantly improves reliability.

[0050] The soft-switching method of this invention leverages the low-loss characteristics of triacs and the fast switching capabilities of IGBTs, significantly reducing switching losses during voltage dips. By precisely calculating the conduction characteristic parameters of the triac and the switching loss parameters of the IGBT, it optimizes the preset delay time of the hybrid switch, ensuring the reliability and stability of the soft-switching process.

[0051] In an optional embodiment, the steps of calculating the voltage stress index and the current stress index of the hybrid switch in real time during the soft switching process and calculating the switch stress compensation angle include: A sliding time window is used to sample the collector-emitter voltage and collector current, and a real-time voltage stress index and a real-time current stress index are calculated based on the sampled data. The voltage stress index is obtained by the root mean square value of the collector-emitter voltage, and the current stress index is obtained by the root mean square value of the collector current. The calculation period of the real-time voltage stress index and the real-time current stress index is an integer multiple of the switching period. comparing the real-time voltage stress indicator and the real-time current stress indicator with respective preset thresholds, and performing compensation angle calculation when the voltage stress indicator or the current stress indicator exceeds the respective preset thresholds; Inputting the real-time voltage stress index and the real-time current stress index into an adaptive fuzzy neural network, wherein the adaptive fuzzy neural network includes a stress assessment layer and a compensation decision layer, wherein the stress assessment layer constructs a fuzzy rule base based on expert experience, and the compensation decision layer uses an online learning algorithm to update the network weight; Determining a reference value of the switch stress compensation angle according to the output of the adaptive fuzzy neural network, the reference value calculation follows the principle that the compensation angle is proportional to the stress excess, and a dynamic limiting mechanism is introduced; The compensation effect of the switch stress compensation angle is monitored in real time, and a comprehensive evaluation index including compensation accuracy and compensation smoothness is constructed. When the comprehensive evaluation index is lower than the preset evaluation threshold, the weight coefficient of the fuzzy rule is updated based on the recursive least squares method, the exponential forgetting factor is used to reduce the influence of historical data, and the upper limit of the weight update step is set to ensure system stability.

[0052] Exemplarily, a sliding time window technique is used to sample the collector-emitter voltage and collector current of the hybrid switch. The length of the sliding window is set to 10 switching cycles, the sampling frequency is 50 kHz, and 100 data points are collected in each switching cycle. The sliding window is moved forward by 1 switching cycle each time, and the data overlap of 9 cycles is maintained to ensure the continuity and smoothness of the calculation. Based on the collected data, the real-time voltage stress index and the real-time current stress index are calculated. The voltage stress index is obtained by the root mean square value of the collector-emitter voltage. Specifically, for the n voltage sampling points vi (i=1, 2, ..., n) in the sliding window, the voltage root mean square value Vrms is calculated. Similarly, for the n current sampling points ii (i=1, 2, ..., n) in the sliding window, the current root mean square value Irms is calculated as the current stress index. The real-time calculation period is set to 10 times the switching period, that is, the stress index is updated every 10 switching cycles.

[0053] The preset threshold for the voltage stress indicator is set at 1.2 times the rated voltage. For example, for a switch device with a rated voltage of 600V, the voltage stress threshold is set at 720V. The preset threshold for the current stress indicator is set at 1.1 times the rated current. For example, for a switch device with a rated current of 50A, the current stress threshold is set at 55A. When the real-time voltage stress indicator exceeds 720V or the real-time current stress indicator exceeds 55A, the compensation angle calculation process is immediately triggered.

[0054] The adaptive fuzzy neural network consists of five layers: input layer, fuzzification layer, rule layer, defuzzification layer, and output layer. The input layer receives standardized voltage and current stress indicators and normalizes the input range to the interval [0, 1]. The fuzzification layer uses a Gaussian membership function to convert the input into a fuzzy set. Each input variable is classified into three linguistic variables: "low," "medium," and "high," with a total of nine fuzzy rules. The rule layer uses a rule base built based on expert experience. For example, "If voltage stress is high and current stress is low, then the compensation angle is medium." The initial rule weight is set to 1.0 and dynamically adjusted during system operation. The defuzzification layer uses a weighted average method to synthesize the outputs of multiple rules. The output layer generates the final compensation angle reference value.

[0055] The compensation angle is calculated in direct proportion to the amount of stress exceeded. Phase compensation is performed when the voltage stress indicator exceeds 120% of the rated value or the current stress indicator exceeds 110% of the rated value. When the voltage stress exceeds the threshold, the compensation angle is proportional to the amount exceeded, with the initial proportional coefficient set at 0.05 degrees per volt. When the current stress exceeds the threshold, the compensation angle is proportional to the amount exceeded, with the initial proportional coefficient set at 0.1 degrees per ampere. A dynamic limiting mechanism is introduced, with the upper limit of the compensation angle set at 5 degrees and the lower limit at 0 degrees to prevent overcompensation and system instability. In actual application, when the voltage stress indicator is detected at 750V, exceeding the threshold by 30V, the initial compensation angle is 30 × 0.05 = 1.5 degrees.

[0056] The compensation effect of the switch stress compensation angle is monitored in real time, and a comprehensive evaluation index J is constructed, encompassing compensation accuracy and compensation smoothness. Compensation accuracy is quantified by the proximity of the stress index to the threshold, while compensation smoothness is quantified by the rate of change of the continuous compensation angle. The compensation accuracy weight coefficient is set to 0.7, and the compensation smoothness weight coefficient is set to 0.3. When the comprehensive evaluation index J falls below the preset evaluation threshold of 0.8, the fuzzy rule weight adjustment mechanism is activated, updating the fuzzy rule weight coefficients using a recursive least squares method. This adjustment mechanism uses an exponential forgetting factor λ = 0.95 to mitigate the influence of historical data and sets an upper limit of 0.1 on the weight update step to ensure system stability. For example, if the rule "If voltage stress is high and current stress is medium, then the compensation angle is high" is detected to be ineffective, the system adjusts the weight of this rule from its initial value of 1.0 to 0.95, reducing its influence in decision-making.

[0057] For example, operating under the conditions of rated voltage 600V and rated current 50A, the initial detected voltage stress index is 735V and the current stress index is 52A, both exceeding the preset thresholds. The adaptive fuzzy neural network outputs a compensation angle baseline value of 1.8 degrees. After applying this compensation angle, the voltage stress index is reduced to 705V and the current stress index is reduced to 51A. After five calculation cycles of adaptive adjustment, the compensation angle is optimized to 2.3 degrees, the voltage stress index is further reduced to 695V, and the current stress index is reduced to 49A, both below the preset thresholds.

[0058] Figure 3The following figure compares the stress reduction rates of three compensation methods during the soft switching of a hybrid switch. In scenarios where voltage stress exceeds the specified value, the proposed method achieves a stress reduction rate of 31.0%, significantly higher than fixed compensation (11.0%) and fuzzy compensation (21.0%). In scenarios where current stress exceeds the specified value, the proposed method achieves a stress reduction rate of 36.0%, outperforming fixed compensation (15.0%) and fuzzy compensation (26.0%). Under complex operating conditions where both current and voltage exceed the specified value, the proposed method achieves even greater effectiveness, with a reduction rate of 51.0%, far exceeding fixed compensation (19.0%) and fuzzy compensation (33.0%). These results demonstrate that the proposed compensation method, based on an adaptive fuzzy neural network, can monitor switch stress in real time and dynamically adjust the compensation strategy, demonstrating significant advantages across all tested scenarios. By utilizing sliding time window technology and an online learning algorithm, this method effectively reduces the voltage and current stresses of the hybrid switch, improving system reliability and efficiency, and providing reliable technical support for high-frequency power conversion systems. The proposed method enables real-time monitoring and adaptive compensation of voltage and current stresses during the soft switching of the hybrid switch, effectively reducing switching losses and improving system reliability and efficiency. This method is applicable to various high-frequency power conversion systems and has broad application prospects.

[0059] In an optional embodiment, the step of inputting the real-time voltage stress index and the real-time current stress index into the adaptive fuzzy neural network includes: Inputting the real-time voltage stress index and the real-time current stress index into the stress assessment layer of the adaptive fuzzy neural network, the stress assessment layer maps the real-time voltage stress index and the real-time current stress index into multiple language variables, and constructing a fuzzy rule base based on expert experience, including a basic rule group and a dynamic rule group. The basic rule group includes multiple expert experience rules, and the dynamic rule group adaptively generates rules according to the operating status; Calculate the rule credibility based on historical operation data. The rule credibility is determined by the ratio of the number of successful rule applications to the total number of rule applications. The rule priority is calculated based on the rule credibility and the rule execution effect evaluation value. The rule priority is updated using exponential smoothing. Inputting the rule activation strength of the stress assessment layer into the compensation decision layer, the compensation decision layer adopts a variable structure neuron design, and updates the network weights through an online learning algorithm. The learning rate of the online learning algorithm decays exponentially with the number of iterations and a minimum learning rate is set; The fuzzy rule base is dynamically optimized, including: calculating rule strength, where the rule strength is determined based on the control error change rate and updated through a forgetting factor mechanism; merging rules when rule similarity exceeds a preset similarity threshold, increasing rule accuracy when the control error exceeds a preset error threshold, and deleting rules when the rule is not activated within a preset period and the rule strength is lower than a preset strength threshold.

[0060] For example, after preprocessing the collected voltage and current data, real-time voltage and current stress indices are calculated. These two indices are fed as input variables into the stress assessment layer of the adaptive fuzzy neural network. The stress assessment layer performs fuzzy mapping on the input real-time voltage stress indices, mapping a voltage stress index value of 0.85, for example, to two linguistic variables, "medium-low" and "medium," with membership degrees of 0.3 and 0.7, respectively. Similarly, a real-time current stress index value of 0.92 is mapped to two linguistic variables, "medium" and "medium-high," with membership degrees of 0.4 and 0.6, respectively. This fuzzy mapping process is implemented using a triangular membership function, with each indicator categorized into seven levels: "very low," "low," "medium-low," "medium," "medium-high," "high," and "very high."

[0061] The fuzzy rule base consists of a basic rule group and a dynamic rule group. The basic rule group contains empirical rules summarized by power system experts, such as "If voltage stress is high and current stress is high, then the system stress state is severe." In actual applications, the basic rule group contains 25 similar rules, covering common operating conditions. The dynamic rule group is adaptively generated based on the actual system operating conditions. For example, if a special condition of 0.82 voltage stress and 0.95 current stress is detected, if the existing rules cannot effectively evaluate it, the system will generate a new rule, "If voltage stress is medium-low and current stress is high, then the system stress state is relatively severe," and add it to the dynamic rule group.

[0062] To calculate rule credibility, the application history of each rule is recorded. For example, if 85 of the past 100 rule applications resulted in evaluations that matched the actual system state, the rule's credibility is 0.85. Rule priority is calculated by combining rule credibility with the rule execution effectiveness evaluation value. The rule execution effectiveness evaluation value is determined based on the control deviation and ranges from 0 to 1, with smaller deviations giving higher evaluation values. If the current execution effectiveness evaluation value of the rule is 0.9, the rule priority is updated using exponential smoothing: new priority = 0.7 × original priority + 0.3 × (0.85 × 0.9) = 0.7 × 0.75 + 0.3 × 0.765 = 0.755, where 0.7 is the smoothing factor. This update mechanism ensures smooth changes in rule priority, avoiding drastic changes caused by temporary fluctuations.

[0063] The rule activation strength of the stress assessment layer is transmitted to the compensation decision layer, which uses a variable structure neuron design. Variable structure neurons are characterized by their ability to dynamically adjust the network structure based on input data characteristics. For example, when the system detects new stress characteristics, it can automatically add neuron nodes. The initial network consists of an input layer, a hidden layer, and an output layer. The input layer corresponds to the rule activation strength, and the output layer corresponds to the compensation control signal. To adapt to the complex and changing operating environment of the power system, the compensation decision layer uses an online learning algorithm to update the network weights. The learning rate is designed to decay exponentially with the number of iterations. The initial learning rate is 0.05, and the learning rate is reduced to 80% of the original value every 100 iterations, but it does not fall below the minimum learning rate of 0.001. This learning rate adjustment strategy allows for larger weight adjustments in the initial system phase for rapid adaptation, while making smaller adjustments to fine-tune performance after the system stabilizes.

[0064] The dynamic optimization mechanism of the fuzzy rule base includes rule strength calculation, rule merging, rule precision enhancement, and rule deletion. Rule strength is determined based on the rate of change of control error. For example, after applying a rule, the system stress drops from 0.85 to 0.65, and the rate of change of control error is 23.5%, indicating that the rule is effective and the rule strength has increased. Rule strength is updated using a forgetting factor mechanism: the new rule strength = 0.8 × original rule strength + 0.2 × current effect, where 0.8 is the forgetting factor, which causes the system to gradually forget historical performance and focus more on recent results.

[0065] When rule similarity exceeds a preset similarity threshold of 0.9, rules are merged. For example, if the similarity between the two rules "If voltage stress is medium-high and current stress is high, then the system stress state is severe" and "If voltage stress is high and current stress is high, then the system stress state is severe" is calculated to be 0.92, exceeding the threshold, the system will merge them into "If voltage stress is medium-high or high and current stress is high, then the system stress state is severe." When the control error exceeds the preset error threshold of 0.15, the system increases rule precision by subdividing the fuzzy set. For example, "medium" voltage stress can be subdivided into two more precise linguistic variants: "medium-low" and "medium-high." If a rule has not been activated within 30 days and its rule strength falls below the preset strength threshold of 0.3, the system removes it from the dynamic rule group to maintain a streamlined and efficient rule base.

[0066] Taking the operating data of a distribution network as an example, when monitoring the voltage stress index of area A at 0.78 and the current stress index at 0.89, the stress assessment layer maps the voltage stress to "medium-low" (0.6) and "medium" (0.4), and the current stress to "medium" (0.2) and "medium-high" (0.8). The activated fuzzy rule includes "If the voltage is medium-low and the current is medium-high, then the system stress is high." The rule activation strength is calculated as 0.6 × 0.8 = 0.48. Based on this activation strength and combined with the network weights, the compensation decision layer outputs a compensation control signal of 0.42, indicating that moderate regulation measures are required. After 20 applications of similar scenarios, the confidence level of this rule increased from the initial 0.75 to 0.82, and the rule priority also increased to 0.79, indicating that the system is gradually adapting to this operating mode and optimizing its decision-making performance.

[0067] This invention combines a basic rule set constructed from expert experience with a dynamic rule set adaptively generated based on operating status, significantly improving the system's adaptability to complex power environments. The calculation mechanism for rule credibility and priority makes the decision-making process more consistent with actual operating experience, improving system reliability. The compensation decision layer designed with variable structure neurons achieves dynamic optimization of network weights through an online learning algorithm. The dynamic optimization strategy of the fuzzy rule base keeps the rule base streamlined and efficient, improving decision-making speed and reducing computing resource consumption.

[0068] In an optional implementation, the dynamic limiting mechanism includes: collecting the switch voltage and switch current, calculating the root mean square value of the switch voltage change rate and the switch current change rate within the sliding sampling window to obtain the voltage stress characteristic and the current stress characteristic, calculating the thermal stress characteristic based on the switch junction temperature and the switch power loss, and constructing the voltage stress characteristic, the current stress characteristic, and the thermal stress characteristic into a stress characteristic vector; calculating a state transition relationship and a control action relationship based on a historical data sequence of the stress characteristic vector, predicting a stress characteristic vector of the next cycle based on the state transition relationship and the control action relationship, and calculating a stress overlimit probability based on the predicted stress characteristic vector; A clipping boundary curve is constructed based on a hyperbolic tangent function, wherein the amplitude coefficient of the clipping boundary curve is adaptively adjusted according to the stress over-limit probability, the slope coefficient of the clipping boundary curve is proportional to the second-order derivative of the compensation angle, and the current compensation angle is compared with the clipping boundary curve to obtain an initial clipping result; Constructing a stress balance evaluation function, wherein the stress balance evaluation function calculates an adjustment coefficient based on a balance degree of the three-dimensional stress characteristics, and correcting the initial clipping result according to the adjustment coefficient to obtain a clipping compensation angle; The limiting compensation angle is input into a dynamic buffer, the smoothing coefficient of the dynamic buffer is in an exponential decay relationship with the compensation angle change rate, and the original compensation angle and the buffer compensation angle are weightedly combined according to the smoothing coefficient to obtain the final compensation angle.

[0069] For example, the first step in the dynamic limiting mechanism is to construct a stress characteristic vector. The switching voltage and current of the power switching device are collected with a sampling frequency of 50kHz and a sampling accuracy of 12 bits. The collected voltage and current data are processed within a 200μs sliding sampling window. Specifically, the switching voltage change rate is calculated as the difference between the voltage values ​​at two adjacent sampling points divided by the sampling interval. The RMS value of all voltage change rates within the sliding window is then calculated to obtain the voltage stress characteristic. For example, if the voltage change rate data within the window is 5V / μs, 6V / μs, 4V / μs, and 7V / μs, the calculated voltage stress characteristic is approximately 5.59V / μs. Similarly, the RMS value of the switching current change rate is calculated to obtain the current stress characteristic. The thermal stress characteristic is calculated using a direct measurement method, using a temperature sensor to monitor the junction temperature of the power switching device in real time and simultaneously calculate the instantaneous power loss of the switching device. The thermal stress characteristic value can be simply expressed as the product of the junction temperature and the power loss, expressed in W·°C. For example, when the measured junction temperature is 85°C and the instantaneous power dissipation is 20W, the thermal stress characteristic value is 85°C × 20W = 1700W·°C. This characteristic value reflects the degree of thermal burden on the switching device during operation, with larger values ​​indicating higher thermal stress. By comparing it with a preset safety threshold (such as 2000W·°C), it can be determined whether the current thermal stress is within the safe range. The three characteristics are combined into a three-dimensional vector [5.59V / μs, 3.25A / μs, 1700W·°C], which serves as the current stress characteristic vector.

[0070] The second step is stress prediction and calculation of the probability of exceeding limits. A historical data queue of stress eigenvectors, containing nearly 100 cycles, is maintained. Based on this data, a recursive least squares algorithm is used to calculate the state transition relationship matrix and the control action relationship matrix. Specifically, a third-order state-space model is established, using the stress eigenvectors and corresponding control inputs (compensation angles) for 10 consecutive cycles as training data. After model training, the current stress state and control input are substituted to predict the stress eigenvector for the next cycle. For example, if the stress eigenvector for the current cycle is [5.59 V / μs, 3.25 A / μs, 1700 W·°C], the predicted stress eigenvector for the next cycle is [6.02 V / μs, 3.40 A / μs, 1820 W·°C]. The stress safety threshold is set to [7 V / μs, 4 A / μs, 2000 W·°C]. Based on the statistical distribution characteristics of the historical data, the probability of the predicted stress vector exceeding the safety threshold is calculated. In this example, the kernel density estimation method calculates the probability of stress exceeding the limit to 0.15.

[0071] The third step is to construct the limiting boundary curve. This curve, constructed using a hyperbolic tangent function, defines the feasible range of the compensation angle. The amplitude coefficient, K, is adaptively adjusted based on the probability of stress overshoot, p: when p is 0.15, K is 0.85; when p increases to 0.3, K decreases to 0.7, tightening the limit. The slope coefficient of the limiting boundary curve is proportional to the second-order derivative of the compensation angle, reflecting its dynamic characteristics. For example, when the second-order derivative of the compensation angle is 0.2, the slope coefficient is set to 0.6; when the second-order derivative increases to 0.5, the slope coefficient increases to 1.5. Comparing the current compensation angle of 15° with the limiting boundary curve, if it exceeds the boundary (assuming the boundary is ±12°), the initial limiting result is 12°.

[0072] Stress balance assessment is the fourth step. A balance evaluation function for three-dimensional stress characteristics was constructed. This function characterizes the degree of balance by calculating the variance between the normalized three-dimensional stress characteristics. The higher the balance, the smaller the variance, indicating a more uniform distribution of the three stresses. For example, if the normalized stress characteristics are [0.8, 0.81, 0.85] with a variance of 0.0007, the balance is high, and the adjustment coefficient is set to 0.95; if it is [0.6, 0.8, 0.9] with a variance of 0.023, the balance is low, and the adjustment coefficient is set to 0.8. Multiplying the initial clipping result of 12° by the adjustment coefficient of 0.95 yields a clipping compensation angle of 11.4°.

[0073] Dynamic buffering is the final step. A dynamic buffer is designed, and its smoothing coefficient α decays exponentially with the rate of change of the compensation angle. When the compensation angle changes at a rate of 5° / ms, α is 0.7; when the rate increases to 10° / ms, α decreases to 0.5. The buffer processes the compensation angle as follows: Assuming the buffer compensation angle in the previous cycle was 10°, the current raw compensation angle is 11.4°, and the smoothing coefficient α is 0.7, the current buffer compensation angle is calculated as 0.7 × 10° + 0.3 × 11.4° = 10.42°. This processing ensures smooth changes in the compensation angle and avoids oscillations caused by sudden changes.

[0074] The present invention achieves a comprehensive evaluation of switch stress, taking into account not only voltage stress and current stress but also thermal stress characteristics, ensuring balanced protection of the switch device in various stress dimensions. The limiting boundary curve based on the hyperbolic tangent function provides a smooth limiting characteristic, avoiding the control discontinuity problem caused by traditional hard limiting. The stress balance evaluation function ensures balanced processing between multi-dimensional stresses and prevents the deterioration of other stresses caused by single stress optimization. The smoothing coefficient design of the dynamic buffer makes the compensation angle adjustment smoother, improving the stability and controllability during the voltage drop simulation process.

[0075] A second aspect of an embodiment of the present invention provides a simulation system with controllable phase voltage drop time difference, comprising: The first unit is used to regulate the voltage of the input three-phase AC power supply through a three-phase voltage regulator. The three-phase voltage regulator adopts a sliding contact structure and is driven and controlled by a DC motor to generate a three-phase voltage signal with a set voltage value; The second unit is used to input the three-phase voltage signals into the corresponding full-voltage power electronic switch and half-voltage power electronic switch, respectively. The full-voltage power electronic switch and half-voltage power electronic switch of each phase adopt a hybrid switch structure of bidirectional thyristor and IGBT, wherein the bidirectional thyristor is connected in series to the collector terminal of the IGBT, and the emitter terminal of the IGBT is connected in parallel with an RC snubber circuit; The third unit is used to collect and process the three-phase voltage signals on the power supply side and the load side through a digital signal processor, extract zero-crossing point information to calculate the phase voltage synchronization control signal, perform phase-locked loop synchronization processing on the phase voltage synchronization control signal and the reference clock signal, and generate a control pulse signal with phase compensation; The fourth unit is used for the digital signal processor to control the conduction timing of the bidirectional thyristor and the IGBT according to the control pulse signal, control the bidirectional thyristor in the full-voltage power electronic switch to turn off, control the IGBT in the full-voltage power electronic switch to turn off after a preset delay, and simultaneously control the IGBT in the half-voltage power electronic switch to turn on, and control the bidirectional thyristor in the half-voltage power electronic switch to turn on after the IGBT in the half-voltage power electronic switch is turned on stably, so as to realize soft switching during voltage drop; The fifth unit is used to use an industrial computer to receive the signal transmitted by the digital signal processor through a high-speed communication interface, and calculate and display the voltage drop waveform.

[0076] According to a third aspect of an embodiment of the present invention, an electronic device is provided, including: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0077] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0078] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A simulation method for controllable phase voltage drop time difference, characterized in that: include: The voltage of the input three-phase AC power supply is regulated by a three-phase voltage regulator. The three-phase voltage regulator adopts a sliding contact structure and is driven and controlled by a DC motor to generate a three-phase voltage signal with a set voltage value; Inputting the three-phase voltage signals into the corresponding full-voltage power electronic switch and half-voltage power electronic switch, respectively, wherein the full-voltage power electronic switch and half-voltage power electronic switch of each phase adopt a hybrid switch structure of bidirectional thyristor and IGBT, the bidirectional thyristor is connected in series to the collector terminal of the IGBT, and the emitter terminal of the IGBT is connected in parallel with an RC snubber circuit; The three-phase voltage signals on the power supply side and the load side are collected and processed by a digital signal processor, the zero-crossing point information is extracted to calculate the phase voltage synchronization control signal, the phase voltage synchronization control signal is synchronized with the reference clock signal through a phase-locked loop, and a control pulse signal with phase compensation is generated; According to the control pulse signal, the bidirectional thyristor in the full-voltage power electronic switch is controlled to turn off, and after a preset delay, the IGBT in the full-voltage power electronic switch is controlled to turn off, and at the same time, the IGBT in the half-voltage power electronic switch is controlled to turn on. After the IGBT in the half-voltage power electronic switch is turned on stably, the bidirectional thyristor in the half-voltage power electronic switch is controlled to turn on, so as to realize soft switching during the voltage drop process; An industrial computer is used to receive the signal transmitted by the digital signal processor through a high-speed communication interface, and calculate and display the voltage drop waveform.

2. The method according to claim 1, characterized in that The steps of collecting and processing three-phase voltage signals on the power supply side and the load side by a digital signal processor, extracting zero-crossing point information to calculate a phase voltage synchronization control signal, performing phase-locked loop synchronization processing on the phase voltage synchronization control signal and a reference clock signal to generate a control pulse signal with phase compensation include: The three-phase voltage signals on the power supply side and the load side are conditioned by a differential amplifier circuit, and the output of the differential amplifier circuit is subjected to a second-order Butterworth low-pass filter and then sampled by an analog-to-digital conversion circuit to obtain a digital voltage signal; Performing a three-point linear interpolation operation on the digital voltage signal, calculating a zero-crossing time based on the voltage values ​​of three adjacent sampling points and the corresponding sampling times, and setting a positive hysteresis threshold and a negative hysteresis threshold, determining that a zero-crossing time occurs when the digital voltage signal crosses the positive hysteresis threshold from negative to positive or crosses the negative hysteresis threshold from positive to negative; Calculating a phase voltage synchronization control signal based on a zero-crossing point, generating a reference clock signal through a phase accumulator, inputting the phase voltage synchronization control signal and the reference clock signal into a digital phase-locked loop circuit, the digital phase-locked loop circuit comprising a phase detector, a second-order loop filter, and a digitally controlled oscillator, wherein a proportional coefficient and an integral coefficient of the second-order loop filter are determined according to a system bandwidth and a damping ratio, and the digital phase-locked loop circuit outputs a synchronization phase signal; The synchronous phase signal is compensated according to the transmission delay characteristics of the switching device, including fixed delay compensation and dynamic characteristic compensation, and a control pulse signal is generated by adopting a space vector modulation algorithm according to the compensated phase angle.

3. The method according to claim 2, characterized in that The step of compensating according to the transmission delay characteristics of the switching device includes: Inputting the synchronized phase signal into a dead zone compensation unit, calculating the dead zone time according to the turn-off delay time of the switching device, and introducing a transition level within the dead zone time to suppress voltage spikes during the switching process, thereby obtaining a phase angle based on fixed delay compensation; Inputting the phase angle based on the fixed delay compensation into the dynamic characteristic compensation unit, and calculating the compensation amount of the turn-on process and the turn-off process respectively based on the asymmetric characteristics of the rise time and fall time of the switching device to obtain the compensated phase angle; The output voltage distortion rate and switching loss corresponding to the control pulse signal generated based on the compensated phase angle are monitored in real time, and a comprehensive optimization objective function including the output voltage distortion rate and the switching loss is constructed. When the output voltage distortion rate exceeds a preset distortion threshold, the upper bridge arm dead time, the lower bridge arm dead time, the transition level threshold, the turn-on process compensation coefficient, and the turn-off process compensation coefficient are encoded into an optimization parameter vector, and the optimization parameter vector is iteratively optimized using a particle swarm algorithm with an adaptive inertia weight, wherein the adaptive inertia weight decreases with the square of the number of iterations. During the optimization process, an asymmetry constraint is imposed on the dead zone compensation parameter, a variation range constraint is imposed on the dynamic characteristic compensation amount, and a smoothing factor is used to perform a weighted combination of new and old compensation parameters to update the compensated phase angle.

4. The method according to claim 1, wherein The steps of controlling the bidirectional thyristor in the full-voltage power electronic switch to turn off according to the control pulse signal, controlling the IGBT in the full-voltage power electronic switch to turn off after a preset delay, and simultaneously controlling the IGBT in the half-voltage power electronic switch to turn on, and controlling the bidirectional thyristor in the half-voltage power electronic switch to turn on after the IGBT in the half-voltage power electronic switch is stably turned on, to achieve soft switching during voltage drop include: Obtaining the on-state voltage drop, on-state resistance, and through-current of the bidirectional thyristor to calculate the on-state characteristic parameters of the bidirectional thyristor; obtaining the collector-emitter voltage and collector current of the IGBT to calculate the switching loss parameters of the IGBT; and determining a preset delay time of the hybrid switch based on the on-state characteristic parameters and the switching loss parameters; Controlling the hybrid switch to perform a soft switching operation according to the control pulse signal, the soft switching operation comprising: controlling the bidirectional thyristor on the full-voltage side to turn off, controlling the IGBT on the full-voltage side to turn off after a preset delay time, and simultaneously controlling the IGBT on the half-voltage side to turn on, and controlling the bidirectional thyristor on the half-voltage side to turn on after the IGBT on the half-voltage side is stably turned on; During the soft switching process, a voltage stress index and a current stress index of the hybrid switch are calculated in real time, and a switch stress compensation angle is calculated, wherein the voltage stress index is obtained by the root mean square value of the collector-emitter voltage, and the current stress index is obtained by the root mean square value of the collector current; The control pulse signal is subjected to real-time dynamic phase compensation according to the switch stress compensation angle to generate a stress-optimized control pulse signal; the conduction timing of the hybrid switch is adjusted based on the stress-optimized control pulse signal to achieve stress-controlled soft switching control during voltage drop.

5. The method according to claim 4, characterized in that During the soft switching process, the voltage stress index and the current stress index of the hybrid switch are calculated in real time. The steps of calculating the switch stress compensation angle include: A sliding time window is used to sample the collector-emitter voltage and collector current, and a real-time voltage stress index and a real-time current stress index are calculated based on the sampled data. The calculation period of the real-time voltage stress index and the real-time current stress index is an integer multiple of the switching period. Inputting the real-time voltage stress index and the real-time current stress index into an adaptive fuzzy neural network, wherein the adaptive fuzzy neural network includes a stress assessment layer and a compensation decision layer, wherein the stress assessment layer constructs a fuzzy rule base based on expert experience, and the compensation decision layer uses an online learning algorithm to update the network weight; Determining a reference value of a switch stress compensation angle according to the output of the adaptive fuzzy neural network and introducing a dynamic limiting mechanism; The compensation effect of the switch stress compensation angle is monitored in real time, and a comprehensive evaluation index including compensation accuracy and compensation smoothness is constructed. When the comprehensive evaluation index is lower than a preset evaluation threshold, the weight coefficient of the fuzzy rule is updated based on the recursive least squares method.

6. The method according to claim 5, characterized in that The step of inputting the real-time voltage stress index and the real-time current stress index into the adaptive fuzzy neural network comprises: Inputting the real-time voltage stress index and the real-time current stress index into the stress assessment layer of the adaptive fuzzy neural network, the stress assessment layer maps the real-time voltage stress index and the real-time current stress index into multiple language variables, and constructing a fuzzy rule base based on expert experience, including a basic rule group and a dynamic rule group. The basic rule group includes multiple expert experience rules, and the dynamic rule group adaptively generates rules according to the operating status; Calculate the rule credibility based on historical operation data. The rule credibility is determined by the ratio of the number of successful rule applications to the total number of rule applications. The rule priority is calculated based on the rule credibility and the rule execution effect evaluation value. The rule priority is updated using exponential smoothing. Inputting the rule activation strength of the stress assessment layer into the compensation decision layer, the compensation decision layer adopts a variable structure neuron design, and updates the network weights through an online learning algorithm. The learning rate of the online learning algorithm decays exponentially with the number of iterations and a minimum learning rate is set; The fuzzy rule base is dynamically optimized, including: calculating rule strength, where the rule strength is determined based on the control error change rate and updated through a forgetting factor mechanism; merging rules when rule similarity exceeds a preset similarity threshold, increasing rule accuracy when the control error exceeds a preset error threshold, and deleting rules when the rule is not activated within a preset period and the rule strength is lower than a preset strength threshold.

7. The method according to claim 5, characterized in that The dynamic limiting mechanism includes: collecting the switch voltage and switch current, calculating the root mean square value of the switch voltage change rate and the switch current change rate within the sliding sampling window to obtain the voltage stress characteristic and the current stress characteristic, calculating the thermal stress characteristic based on the switch junction temperature and the switch power loss, and constructing the voltage stress characteristic, the current stress characteristic, and the thermal stress characteristic into a stress characteristic vector; calculating a state transition relationship and a control action relationship based on a historical data sequence of the stress characteristic vector, predicting a stress characteristic vector of the next cycle based on the state transition relationship and the control action relationship, and calculating a stress overlimit probability based on the predicted stress characteristic vector; A clipping boundary curve is constructed based on a hyperbolic tangent function, wherein the amplitude coefficient of the clipping boundary curve is adaptively adjusted according to the stress over-limit probability, the slope coefficient of the clipping boundary curve is proportional to the second-order derivative of the compensation angle, and the current compensation angle is compared with the clipping boundary curve to obtain an initial clipping result; Constructing a stress balance evaluation function, wherein the stress balance evaluation function calculates an adjustment coefficient based on a balance degree of the three-dimensional stress characteristics, and correcting the initial clipping result according to the adjustment coefficient to obtain a clipping compensation angle; The limiting compensation angle is input into a dynamic buffer, the smoothing coefficient of the dynamic buffer is in an exponential decay relationship with the compensation angle change rate, and the original compensation angle and the buffer compensation angle are weightedly combined according to the smoothing coefficient to obtain the final compensation angle.

8. A simulation system with controllable phase voltage drop time difference, used to implement the method according to any one of claims 1 to 7, characterized in that: include: The first unit is used to regulate the voltage of the input three-phase AC power supply through a three-phase voltage regulator. The three-phase voltage regulator adopts a sliding contact structure and is driven and controlled by a DC motor to generate a three-phase voltage signal with a set voltage value; The second unit is used to input the three-phase voltage signals into the corresponding full-voltage power electronic switch and half-voltage power electronic switch, respectively. The full-voltage power electronic switch and half-voltage power electronic switch of each phase adopt a hybrid switch structure of bidirectional thyristor and IGBT, wherein the bidirectional thyristor is connected in series to the collector terminal of the IGBT, and the emitter terminal of the IGBT is connected in parallel with an RC snubber circuit; The third unit is used to collect and process the three-phase voltage signals on the power supply side and the load side through a digital signal processor, extract zero-crossing point information to calculate the phase voltage synchronization control signal, perform phase-locked loop synchronization processing on the phase voltage synchronization control signal and the reference clock signal, and generate a control pulse signal with phase compensation; The fourth unit is used for the digital signal processor to control the conduction timing of the bidirectional thyristor and the IGBT according to the control pulse signal, control the bidirectional thyristor in the full-voltage power electronic switch to turn off, control the IGBT in the full-voltage power electronic switch to turn off after a preset delay, and simultaneously control the IGBT in the half-voltage power electronic switch to turn on, and control the bidirectional thyristor in the half-voltage power electronic switch to turn on after the IGBT in the half-voltage power electronic switch is turned on stably, so as to realize soft switching during voltage drop; The fifth unit is used to use an industrial computer to receive the signal transmitted by the digital signal processor through a high-speed communication interface, and calculate and display the voltage drop waveform.

9. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images