High-frequency power supply double-group cooperative output control method and system

Abstract

The application provides a high-frequency power supply double-group collaborative output control method and system, and belongs to the technical field of power electronics control, and comprises the following steps: when the instantaneous change rate of secondary current is detected to exceed a preset fluctuation threshold, a double-group collaborative pulse power supply mode is triggered; average output voltage and peak voltage are obtained, and dynamic characteristic parameters are determined; the basic voltage given value is corrected, and the calibration falling back time is obtained; the electric field transient response index is generated, and the peak voltage secondary correction factor is obtained; the composite correction coefficient is formed, and the pulse peak voltage given value is adjusted; the pulse width compression coefficient is calculated, and the current pulse width is real-time reduced; the collaborative control instruction is outputted to drive the base pressure loop to output the dynamic basic voltage, and drive the pulse loop to output the high-voltage pulse. The application solves the problems that the energy mismatch and discharge out of control are caused by the fixed basic voltage, pulse peak value and pulse width and the lack of collaboration when the high-frequency power supply double-group structure is in the load transient fluctuation.

Description

Technical Field

[0001] This invention relates to the field of power electronic control technology, specifically to a high-frequency power supply dual-group coordinated output control method and system. Background Technology

[0002] In industrial applications such as electrostatic precipitators, plasma treatment, and pulsed corona discharge, high-frequency high-voltage power supply systems typically employ a dual-base voltage superimposed pulse architecture. One DC base voltage circuit provides stable basic electric field energy, while the other high-voltage pulse circuit superimposes instantaneous high-peak pulses onto the base voltage to meet the dual requirements of specific processes for electric field strength and transient energy. Due to its high energy utilization efficiency and adjustable discharge characteristics, this architecture is gradually becoming the mainstream solution in the field of high-power pulsed power supplies.

[0003] Existing dual-group coordinated control methods mainly employ fixed timing or preset waveforms, i.e., setting fixed base voltage amplitude, pulse peak value, and pulse width based on offline experience, and outputting in an open loop according to preset parameters during operation. However, in actual operating conditions, load characteristics are not constant. Fluctuations in factors such as flue gas velocity, dust concentration, and gas composition can cause drastic and irregular changes in equivalent impedance, leading to transient distortion of secondary current. Under these conditions, the fixed-parameter control strategy reveals significant shortcomings: on the one hand, when the load suddenly increases, the energy supply of the base voltage loop is insufficient, resulting in a slower current rise rate at the pulse leading edge, failing to form an effective instantaneous power surge and affecting process performance; on the other hand, when the load lightens or flashover warnings occur, excessively high fixed pulse peak values ​​and excessively wide pulse widths can easily cause over-discharge or even breakdown, resulting in energy waste and threatening system safety. Furthermore, there is a lack of information interaction and dynamic coordination between the pulse loop and the base voltage loop based on real-time load conditions. They remain in an independent or simply time-coupled operating state, unable to adaptively adjust parameters according to transient changes in the electric field.

[0004] Therefore, how to sense the transient changes in load in real time and use this as a basis to dynamically coordinate the adjustment of the base voltage amplitude, pulse peak value and pulse width to avoid energy mismatch and discharge runaway caused by fixed parameters is a key technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] This invention provides a method and system for controlling the coordinated output of a dual-group high-frequency power supply, which solves the problem of energy mismatch and uncontrolled discharge caused by the fixed base voltage, pulse peak value and pulse width and lack of coordination in the dual-group high-frequency power supply architecture under load transient fluctuations.

[0006] In view of the above problems, the present invention provides a high-frequency power supply dual-group coordinated output control method and system.

[0007] In a first aspect, the present invention provides a high-frequency power supply dual-group coordinated output control method, comprising: When the instantaneous rate of change of the secondary current exceeds the preset fluctuation threshold, the dual-group cooperative pulse power supply mode is triggered. The average output voltage of the base voltage circuit and the peak voltage of the pulse circuit during the peak holding phase are obtained within the current control cycle, and the dynamic characteristic parameters are determined. The dynamic characteristic parameters include at least the steepness of the secondary current rise edge and the current fall time. The base voltage is corrected based on the steepness of the secondary current rise to obtain the dynamic base voltage. At the same time, the calibration fall-off time is obtained, and the peak compensation coefficient is determined in combination with the current fall-off time. Based on the steepness of the secondary current rise and the current fall time, an electric field transient response index is generated, and the electric field transient response index is input into a preset electric field transient response compensation model to obtain a peak voltage secondary correction factor. A composite correction coefficient is formed by using the peak compensation coefficient and the peak voltage second correction factor. The pulse peak voltage setpoint is then adjusted according to the composite correction coefficient to obtain the dynamic peak voltage. Based on the dynamic base voltage and the dynamic peak voltage, the pulse width compression coefficient is calculated, and the current pulse width is reduced in real time to obtain the limited pulse width; Output a coordinated control command carrying the dynamic base voltage, the dynamic peak voltage, and the limiting pulse width to drive the base voltage circuit to output the dynamic base voltage and drive the pulse circuit to output a high-voltage pulse according to the dynamic peak voltage within the limiting pulse width.

[0008] Secondly, the present invention provides a high-frequency power supply dual-group coordinated output control system, comprising: The transient detection and mode triggering module is used to trigger a dual-group cooperative pulse power supply mode when the instantaneous change rate of the secondary current exceeds a preset fluctuation threshold. The feature parameter extraction module is used to obtain the average output voltage of the base voltage circuit and the peak voltage of the pulse circuit during the peak holding phase within the current control cycle, and to determine the dynamic feature parameters, wherein the dynamic feature parameters include at least the secondary current rise steepness and the current fall time. The base voltage correction module is used to correct the base voltage given value based on the steepness of the secondary current rise edge to obtain the dynamic base voltage, and at the same time obtain the calibration fall-off time, and determine the peak compensation coefficient in combination with the current fall-off time. The transient response evaluation module is used to generate an electric field transient response index based on the rise steepness of the secondary current and the fall time of the current, and input the electric field transient response index into a preset electric field transient response compensation model to obtain a peak voltage secondary correction factor. The peak voltage composite correction module is used to form a composite correction coefficient by using the peak compensation coefficient and the peak voltage second correction factor, and then adjust the pulse peak voltage setpoint according to the composite correction coefficient to obtain the dynamic peak voltage. The pulse width limiting calculation module is used to calculate the pulse width compression coefficient based on the dynamic base voltage and the dynamic peak voltage, and to reduce the current pulse width in real time to obtain the limited pulse width. The collaborative control command output module is used to output a collaborative control command carrying the dynamic base voltage, the dynamic peak voltage and the limiting pulse width, so as to drive the base voltage circuit to output the dynamic base voltage and drive the pulse circuit to output a high voltage pulse according to the dynamic peak voltage within the limiting pulse width.

[0009] One or more technical solutions provided in this invention have at least the following technical effects or advantages: The technical solution of this invention firstly detects the instantaneous change rate of the secondary current in real time and compares it with a statistical threshold, thereby quickly triggering a dual-group coordinated pulse power supply mode when the load undergoes a transient change, thus solving the problem of response lag in the fixed parameter mode.

[0010] Furthermore, the average base voltage and peak pulse voltage are collected simultaneously, and the rise steepness of the secondary current and the fall time of the current are extracted to provide multi-dimensional real-time feedback data for subsequent regulation.

[0011] Furthermore, the dynamic base voltage is obtained by correcting the base voltage setpoint based on the rising edge steepness. At the same time, the peak compensation coefficient is generated by using the deviation between the fall time and the calibrated value, so as to realize the adaptive adjustment of the base voltage amplitude to the load absorption capacity and the initial matching of the pulse peak value to the discharge recovery speed.

[0012] Furthermore, the electric field transient response index is generated by integrating the rise steepness and fall time, and the peak voltage quadratic correction factor is output through a pre-set compensation model to compensate for the insufficient accuracy of single-dimensional correction with nonlinear mapping capability.

[0013] Furthermore, by constructing a fusion weight through dynamic energy ratio, the peak compensation coefficient and the secondary correction factor are integrated into a composite correction coefficient and the pulse peak voltage setpoint is adjusted to achieve dual-dimensional collaborative adaptive adjustment of peak voltage.

[0014] Furthermore, a pulse width compression coefficient is generated based on the dynamic base voltage and dynamic peak voltage, and multi-level corrections are performed by integrating steepness, fall-off time, and energy ratio to reduce the pulse width in real time to match the voltage regulation results.

[0015] Finally, the dynamic base voltage, dynamic peak voltage, and limiting pulse width are encapsulated into coordinated control commands to synchronously drive the base voltage and pulse circuit to achieve real-time linkage output of the three parameters.

[0016] In summary, the technical solution of this invention can automatically complete the entire process from mode triggering, feature extraction, base voltage and peak voltage individual correction to pulse width linkage adjustment when the load experiences transient fluctuations. Ultimately, it drives the coordinated output of the two circuits with a unified command, fundamentally eliminating the total energy mismatch problem caused by the independent operation of base voltage, peak voltage, and pulse width in traditional fixed-parameter control. When load absorption increases or recovery slows, it actively reduces the base voltage, shortens the peak voltage, and simultaneously compresses the pulse width. When the load recovers, it maintains output capability, ensuring the instantaneous power surge effect required by the process while effectively suppressing the risk of over-discharge and flashover, ultimately improving the energy efficiency, reliability, and dynamic quality of the high-frequency high-voltage power supply system. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a high-frequency power supply dual-group coordinated output control method provided by the present invention.

[0018] Figure 2 This is a schematic diagram illustrating the calculation process of the limiting pulse width in a high-frequency power supply dual-group coordinated output control method provided by the present invention.

[0019] Figure 3 This is a schematic diagram of the structure of a high-frequency power supply dual-group cooperative output control system provided by the present invention.

[0020] In the attached diagram, the labels representing each component are as follows: Transient detection and mode triggering module 11, feature parameter extraction module 12, base voltage correction module 13, transient response evaluation module 14, peak voltage composite correction module 15, pulse width limiting calculation module 16, and cooperative control command output module 17. Detailed Implementation

[0021] This invention provides a method and system for controlling the coordinated output of a dual-group high-frequency power supply, which solves the problem of energy mismatch and uncontrolled discharge caused by the fixed base voltage, pulse peak value and pulse width and lack of coordination in the dual-group high-frequency power supply architecture under load transient fluctuations.

[0022] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0023] Example 1, as Figure 1 As shown, the present invention provides a high-frequency power supply dual-group cooperative output control method, the method comprising: S100: When the instantaneous rate of change of the secondary current exceeds the preset fluctuation threshold, the dual-group cooperative pulse power supply mode is triggered.

[0024] This step is used to quickly identify transient changes in load conditions, such as sudden impedance changes and flashover precursors, to avoid energy mismatch or discharge runaway caused by continuous operation in the conventional fixed parameter mode. By triggering the collaborative control mode in a timely manner, it provides an opportunity to start subsequent dynamic parameter adjustments, ensuring the real-time response and operational safety of the system.

[0025] In this step, the preset fluctuation threshold is determined by calculating the algebraic sum of the mean and standard deviation of the historical values ​​of the instantaneous change rate of multiple secondary currents continuously collected by the electrostatic precipitator under dry flue gas conditions. The dual-group coordinated pulse power supply mode is that the base voltage circuit enters the voltage follower output state, the pulse circuit enters the controlled intermittent chopping output state, and the pulse circuit stops outputting after each pulse cycle until the start of the next pulse cycle.

[0026] Specifically, the preset fluctuation threshold is set as follows: During the stable operation phase of the electrostatic precipitator under dry flue gas conditions, historical values ​​of the instantaneous change rate of the secondary current are continuously collected for several cycles. The arithmetic mean μ and standard deviation σ of this set of data are calculated, and the preset fluctuation threshold is determined according to the formula: Preset fluctuation threshold = μ + k × σ, where k is a positive real number multiplier factor, determined by a combination of the allowable detection sensitivity and false trigger probability under on-site conditions. The larger the k value, the higher the threshold, the lower the false trigger probability but the lower the sensitivity; the smaller the k value, the lower the threshold, the higher the sensitivity but the increased risk of false trigger. In engineering, it is usually selected between 2 and 4 based on on-site commissioning.

[0027] For example, in a power plant, an electrostatic precipitator continuously collects the instantaneous rate of change of secondary current at 1000 sampling points under dry flue gas conditions. The calculated mean μ is 0.15 A / μs, and the standard deviation σ is 0.04 A / μs. Based on the on-site requirements for balancing sensitivity and false trigger rate, k=3 is chosen, and the preset fluctuation threshold is determined to be 0.15 + 3 × 0.04 = 0.27 A / μs. When the instantaneous rate of change of secondary current detected in real time suddenly increases to 0.35 A / μs, exceeding the threshold of 0.27 A / μs, the system determines that a significant transient disturbance has occurred in the load and triggers the dual-group coordinated pulse power supply mode.

[0028] In the dual-group coordinated pulse power supply mode, the base voltage circuit enters the voltage follower output state, that is, its output voltage no longer maintains a fixed value, but is adjusted in real time according to the dynamic base voltage command generated by the subsequent control link; the pulse circuit enters the controlled intermittent chopping output state, and after completing one pulse output according to the dynamic peak voltage and the limited pulse width in each pulse cycle, it is immediately turned off until the beginning of the next pulse cycle when it responds to the trigger signal to perform the next pulse output, thereby completely stopping energy injection during the pulse interval period and avoiding over-discharge or flashover caused by continuous high energy output.

[0029] It should be noted that the above values ​​are for illustrative purposes only and do not constitute a limitation on the present invention.

[0030] In summary, this step, by monitoring the instantaneous rate of change of the secondary current in real time and comparing it with the fluctuation threshold set based on historical statistical data, quickly and reliably triggers the base voltage circuit and pulse circuit to enter the coordinated pulse power supply mode when the load experiences a transient change. This effectively avoids energy mismatch or discharge runaway caused by response lag, ensuring both the timeliness and accuracy of mode switching and reducing the risk of false triggering under normal operating conditions.

[0031] S200: Obtain the average output voltage of the base voltage circuit and the peak voltage of the pulse circuit during the peak hold phase within the current control cycle, and determine the dynamic characteristic parameters, wherein the dynamic characteristic parameters include at least the secondary current rise steepness and the current fall time.

[0032] This step is used to obtain the actual output energy levels of the two sets of loops within the current control cycle and quantify the dynamic response characteristics of the secondary current to pulse excitation. The average base voltage reflects the energy supply state of the base electric field, the peak pulse voltage reflects the instantaneous injection intensity, and the rise steepness and fall time characterize the load's ability to absorb pulse energy and its recovery speed after discharge, respectively. These parameters together constitute the full-dimensional feedback basis required for subsequent dynamic adjustment and are the data foundation for achieving coordinated control of base voltage, peak voltage, and pulse width.

[0033] Step S200 in the method provided by the present invention includes: Within each pulse cycle, the instantaneous value of the secondary current is acquired at a fixed sampling frequency. Based on the change in the secondary current between the pulse start time and the time when it rises to the peak value, and the corresponding time length, the rising edge steepness of the secondary current is determined. At the end of the pulse cycle, the current fall-off time is determined based on the time length between the pulse turn-off time and the time when the secondary current drops to a specified proportion of the current value at the turn-off time. The average value obtained by averaging multiple instantaneous voltage values ​​obtained by sampling the base voltage circuit output voltage at equal intervals multiple times within the current control cycle is used as the average output voltage of the base voltage circuit. The maximum voltage value output by the peak hold circuit in the detection pulse loop within the current pulse period is taken as the peak voltage output by the pulse loop during the peak hold phase.

[0034] In this step, the instantaneous value of the secondary current is first acquired at a fixed sampling frequency within each pulse cycle. The secondary current value corresponding to the pulse trigger start time and the secondary current peak value corresponding to the current rising to the peak value are recorded. The current change and the time length between the two times are calculated. The current change is divided by the time length to obtain the rising edge steepness of the secondary current in the current pulse cycle.

[0035] For example, in one pulse cycle of an electrostatic precipitator, the sampling frequency is 10MHz. At the pulse start time t0 = 0μs, the instantaneous value of the secondary current is 2A; when the current reaches its peak value t1 = 0.8μs, the peak value of the secondary current is 18A. Therefore, the current change is 18A - 2A = 16A, the time duration is 0.8μs, and the rise edge steepness is 16A / 0.8μs = 20A / μs.

[0036] Furthermore, at the end of the pulse cycle, the secondary current begins to decrease after the pulse circuit is turned off. The pulse turn-off time and the secondary current value at this time are recorded, and the time when the secondary current decreases from the turn-off value to a specified proportion is monitored. The time length between the two times is taken as the current fallback time within the current pulse cycle.

[0037] The calculation steps for the specified ratio include: Under the calibration conditions of introducing dry flue gas into the electrostatic precipitator and no flashover of the electric field, the waveform of the secondary current falling segment of multiple pulse cycles was continuously collected. For the secondary current decline segment of each pulse cycle, the current decay rate is calculated point by point starting from the pulse turn-off time. The ratio of the current value when the current decay rate decreases from the maximum value to one-tenth of the maximum value to the current value at the turn-off time is used as a single-cycle proportional sample. Sort the single-cycle proportion samples obtained from all pulse cycles according to their numerical values, remove the extreme samples at both ends, calculate the arithmetic mean of the remaining samples, and use the arithmetic mean as the specified proportion.

[0038] Specifically, under stable calibration conditions where dry flue gas is introduced into the electrostatic precipitator and no flashover occurs in the electric field, waveform data of the secondary current's descent segment are continuously collected for multiple pulse cycles. For example, a power plant is calibrating under dry flue gas conditions, where the flue gas humidity is below 5% and the electric field operates smoothly without flashover. The waveform of the secondary current's descent segment is continuously collected for 1000 pulse cycles at a sampling frequency of 10MHz, with each descent segment recorded from the pulse turn-off moment until the current essentially returns to zero.

[0039] Secondly, for the secondary current decline segment of each pulse cycle, starting from the pulse turn-off moment, the current decay rate between adjacent sampling points is calculated sequentially, i.e., the amount of current reduction per unit time. For example, for the first pulse cycle, the pulse turn-off moment is t=5μs, and the current value is 18A. The sampling interval is 0.1μs. Between t=5μs and t=5.1μs, the current decreases from 18A to 17.2A, and the decay rate = (18A-17.2A) / 0.1μs = 8A / μs. This process is repeated point-by-point from t=5μs to obtain the decay rate sequence for each moment of the current cycle's decline segment.

[0040] Next, in each falling segment, find the maximum value of the decay rate and record the corresponding time; continue to track the change of the decay rate, and when the decay rate drops from the maximum value to one-tenth of the maximum value, record the corresponding current value; calculate the ratio of this current value to the current value at the turn-off time, and use it as a single-cycle proportional sample of the pulse cycle.

[0041] For example, in the first pulse cycle, the calculated maximum decay rate is 8 A / μs, occurring at t=5μs, which is the turn-off time. Continuing to track along the falling segment, when the decay rate drops to 8 A / μs × 0.1 = 0.8 A / μs, the corresponding time is t=9.2μs, at which point the current value is 2.5 A. The current value at the turn-off time is 18 A, so the single-cycle proportion sample of the current cycle = 2.5 A / 18 A ≈ 0.139, or 13.9%.

[0042] Finally, the single-cycle proportion samples obtained from all 1000 pulse cycles are sorted by numerical value, and extreme samples at both ends, such as the first 5% and the last 5%, are removed. The arithmetic mean of the remaining samples is calculated, and the result is the specified proportion.

[0043] For example, sort 1000 proportional samples from smallest to largest, remove the smallest 50 and largest 50 samples, and keep the middle 900 samples. The values ​​of these 900 samples are distributed between 0.08 and 0.15, and their arithmetic mean is 0.102, which means the specified proportion is determined to be 10.2%, and in practical applications, it is rounded to 10%.

[0044] For example, using the aforementioned specified ratio of 10%, continuing the previous example, at the pulse turn-off time t2 = 5μs, the secondary current value is 18A. 10% of the turn-off value is 18A × 10% = 1.8A. The time when the secondary current drops from 18A to 1.8A is t3 = 12μs. Therefore, the current fallback time = t3 - t2 = 12μs - 5μs = 7μs.

[0045] Furthermore, within the current control cycle, the output voltage of the base voltage circuit is sampled multiple times at equal intervals to obtain multiple instantaneous voltage values. All sampled values ​​are then arithmetically averaged, and the resulting value is the average output voltage of the base voltage circuit in the current control cycle.

[0046] For example, the control cycle length is 100μs, and the output voltage of the base voltage circuit is sampled at 1μs intervals to obtain a total of 100 instantaneous voltage values. After summing them up, the average output voltage is divided by 100 to obtain an average output voltage of 50kV.

[0047] Finally, the maximum voltage value captured and held by the peak hold circuit in the pulse loop during the current pulse cycle is the peak voltage output by the pulse loop during the peak hold phase. For example, if the maximum voltage value captured by the peak hold circuit during the current pulse cycle is 80kV, then the peak voltage output by the pulse loop during the peak hold phase is 80kV.

[0048] It should be noted that the above values ​​are for illustrative purposes only and do not constitute a limitation on the present invention.

[0049] In summary, this step can comprehensively and in real-time quantify the energy absorption and recovery characteristics of the load under pulse excitation, providing a precise and synchronous feedback data foundation for subsequent multi-parameter coordinated adjustment based on load transient changes, and effectively solving the problem of adjustment inaccuracy caused by information lag or single dimension in traditional solutions.

[0050] S300: Based on the steepness of the secondary current rise edge, the base voltage setpoint is corrected to obtain the dynamic base voltage. At the same time, the calibration fall-off time is obtained, and the peak compensation coefficient is determined in combination with the current fall-off time.

[0051] This step decouples and corrects the core control parameters of the base voltage circuit and the pulse circuit based on two key characteristics of the load transient response. By adjusting the rise edge steepness of the base voltage, the imbalance between base voltage energy supply and demand caused by load increases or decreases can be compensated. By generating a peak compensation coefficient by the deviation between the fall time and the calibrated value, the pulse peak voltage can be pre-adjusted to match the current discharge recovery speed of the load and avoid over-excitation or under-excitation.

[0052] Step S300 in the method provided by the present invention includes: Under calibration conditions, the calibration rise edge steepness is obtained using the same method as that used to determine the rise edge steepness of the secondary current. Based on the relative increase in the rise edge steepness of the secondary current relative to the calibrated rise edge steepness, a steepness suppression component is determined; The fallback suppression component is determined based on the relative extension of the current fallback time relative to the calibrated fallback time; The ratio of the peak voltage output by the pulse circuit during the peak holding phase to the average output voltage of the base voltage circuit is obtained as the dynamic energy ratio; The pressure drop is determined based on the steepness suppression component, the fallback suppression component, and the dynamic energy ratio, wherein the pressure drop increases with the increase of the steepness suppression component, the fallback suppression component, and the dynamic energy ratio; Subtracting the voltage drop from the given base voltage value yields the dynamic base voltage. Divide the calibrated fallback time by the current fallback time, and use the resulting value as the peak compensation coefficient.

[0053] In this step, under calibration conditions, the same method as in the previous step of determining the rising edge steepness of the secondary current is used. That is, under the stable conditions of dry flue gas and no flashover in the electric field, the secondary current waveform of multiple pulse cycles is collected, the current change and corresponding time length between the pulse start time and the rise to the peak time are recorded, the ratio of the current change to the time length is calculated, and the arithmetic mean of the results of multiple cycles is taken as the calibration rising edge steepness.

[0054] For example, under the calibration conditions of dry flue gas, 1000 pulse cycles were continuously collected. The rise edge steepness of each cycle was calculated in the same way as the previous steps. The results were distributed between 18A / μs and 20A / μs. The arithmetic mean was taken to obtain the calibration rise edge steepness as 19A / μs.

[0055] Furthermore, the measured rise time steepness of the secondary current is compared with the calibrated rise time steepness to determine the steepness suppression component. When the measured rise time steepness is greater than or equal to the calibrated rise time steepness, it indicates that the load absorption capacity is maintained or enhanced, which is a normal or excellent operating condition. At this time, the steepness suppression component is zero, and the base voltage drop caused by the rise time change is not triggered. When the measured rise time steepness is less than the calibrated rise time steepness, it indicates that the load absorption capacity is weakened. The relative decrease of the calibrated value with respect to the measured value is calculated, that is, the steepness suppression component = (calibrated value - measured value) / calibrated value, which is used as the steepness suppression component.

[0056] For example, the calibrated rise edge steepness is 19 A / μs. If the current measured rise edge steepness is 20 A / μs, which is greater than the calibrated value of 19 A / μs, the load absorption capacity is enhanced, and the steepness suppression component is zero. If, under another operating condition, the current measured rise edge steepness is 15 A / μs, which is less than the calibrated value of 19 A / μs, the load absorption capacity is weakened, then the steepness suppression component = (19-15) / 19 ≈ 0.211.

[0057] Furthermore, the measured current fallback time is compared with the pre-stored calibrated fallback time, and the relative extension of the current value relative to the calibrated value is calculated, i.e., relative extension = (current value - calibrated value) / calibrated value, which is used as the fallback suppression component. If the calculation result is negative, the fallback suppression component is set to zero.

[0058] The calibration fallback time is calculated by taking the arithmetic average of the current fallback time of multiple pulse cycles under the calibration conditions where dry flue gas is introduced into the electrostatic precipitator and no flashover occurs in the electric field. This arithmetic average is then pre-calculated and stored in the controller's fixed parameters.

[0059] For example, the measured current fall-off time is compared with the pre-stored calibration fall-off time. The calibration fall-off time is 7 μs. With the current load undergoing a transient change, the measured current fall-off time is extended to 10.5 μs. Therefore, the relative extension is (10.5 - 7) / 7 = 0.5, and the fall-off suppression component is determined to be 0.5, meaning the load discharge recovery speed is 50% slower than the calibration condition.

[0060] Furthermore, the original values ​​of the peak voltage of the pulse circuit and the average output voltage of the base voltage circuit, acquired and latched at the beginning of the current control cycle, are obtained, and their ratio is calculated as the dynamic energy ratio. The dynamic energy ratio remains unchanged within this control cycle and is not affected by the voltage correction results being calculated in this cycle.

[0061] For example, in the current control cycle, the peak voltage of the pulse circuit obtained in the previous steps is 80kV, and the average output voltage of the base voltage circuit is 50kV. Then the dynamic energy ratio = 80kV / 50kV = 1.6.

[0062] Furthermore, the steepness suppression component, the fallback suppression component, and the dynamic energy ratio are used as input variables, and the voltage drop is calculated according to a preset formula. The voltage drop increases with the increase of the three variables. The specific calculation formula is: Voltage drop = (steepness suppression component + fallback suppression component) × dynamic energy ratio × unit voltage value, where the unit voltage value is a preset reference voltage dimension parameter.

[0063] Specifically, the unit voltage value is determined based on the rated output voltage level of the base voltage circuit and the load characteristics. The setting rule is: unit voltage value = base voltage setpoint × K, where K is a dimension conversion coefficient, and its value ranges from 0.01% to 5% of the base voltage setpoint. The selection of the K value is related to the sensitivity of the electric field to voltage fluctuations. For loads sensitive to voltage fluctuations, such as those prone to flashover, a smaller K value is used to achieve fine adjustment, while for loads insensitive to voltage fluctuations, a larger K value is used to accelerate the adjustment response speed.

[0064] For example, if the base voltage is 50kV, and the electric field is sensitive to voltage fluctuations, taking K=0.02%, then the unit voltage value is 50kV × 0.02% = 10V; if the electric field is not sensitive to voltage fluctuations, taking K=20%, then the unit voltage value is 50kV × 20% = 10kV. Thus, the unit voltage value is automatically matched to the specific electric field level and process characteristics, ensuring that the voltage drop has reproducible engineering significance in systems of different voltage levels.

[0065] For example, the current measured rise time steepness is 20A / μs, which is greater than the calibration value. The steepness suppression component is zero, the fallback suppression component is 0.5, the dynamic energy ratio is 1.6, and the unit voltage value is 10kV. Then the voltage drop = (0 + 0.5) × 1.6 × 10kV = 0.5 × 1.6 × 10kV = 8kV.

[0066] Furthermore, the calculated voltage drop is subtracted from the base voltage setpoint, and the result is used as the dynamic base voltage, which is then used as the output command for the base voltage circuit within the current control cycle. For example, if the base voltage setpoint is 50kV and the voltage drop is 8kV, then the dynamic base voltage = 50kV - 8kV = 42kV.

[0067] Finally, the calibrated fallback time is divided by the currently measured current fallback time, and the resulting ratio is used as the peak compensation coefficient. For example, if the calibrated fallback time is 7 μs and the current current fallback time is 10.5 μs, then the peak compensation coefficient = 7 μs / 10.5 μs ≈ 0.667. A peak compensation coefficient less than 1 indicates that the load recovery speed is slower than the calibrated condition, requiring a corresponding reduction in the pulse peak voltage.

[0068] It should be noted that the above values ​​are for illustrative purposes only and do not constitute a limitation on the present invention.

[0069] In summary, this step achieves synchronous adaptive adjustment of the base voltage amplitude and pulse peak voltage in response to load transient changes: when the load absorption capacity increases or the discharge recovery speed slows down, the base voltage can be actively reduced to suppress over-energy injection. At the same time, a peak compensation coefficient is generated by the fallback time ratio to match the pulse peak voltage with the actual load recovery rate, effectively avoiding the energy overshoot or under-excitation problem caused by response lag under traditional fixed parameter control.

[0070] S400: Based on the rising edge steepness of the secondary current and the falling time of the current, generate an electric field transient response index, and input the electric field transient response index into a preset electric field transient response compensation model to obtain a peak voltage secondary correction factor.

[0071] This step combines the secondary current rise steepness and current fall time obtained in the previous step to generate an electric field transient response index. This index is then input into a pre-built electric field transient response compensation model deployed in the controller. The model outputs a peak voltage secondary correction factor, used for fine-tuning the pulse peak voltage. Step S400 in the method provided by the present invention includes: Under the calibration conditions, obtain the calibration rise steepness and calibration fall time; The relative increment of steepness is determined based on the relative increase of the secondary current rising edge steepness relative to the calibrated rising edge steepness; The relative increment of the fall-off time is determined based on the relative extension of the current fall-off time relative to the calibrated fall-off time; The ratio of the peak voltage output by the pulse circuit during the peak holding phase to the average output voltage of the base voltage circuit is obtained as the dynamic energy ratio; Based on the relative steepness increment and the dynamic energy ratio, a steepness contribution term is determined, wherein the steepness contribution term is proportional to both the relative steepness increment and the dynamic energy ratio; Based on the relative increment of fallback time and the dynamic energy ratio, a fallback contribution term is determined, wherein the fallback contribution term is proportional to both the relative increment of fallback time and the dynamic energy ratio; The steepness contribution term and the fallback contribution term are combined to obtain the electric field transient response index.

[0072] In this step, the calibration rise edge steepness and calibration fall time obtained under the calibration conditions are first retrieved from the controller's preset parameters as the benchmark for calculating the relative change. For example, the calibration rise edge steepness is 19A / μs, and the calibration fall time is 7μs.

[0073] Further, the measured rising edge steepness of the secondary current obtained in the previous steps is compared with the calibrated rising edge steepness, and the relative steepness increment is calculated as (measured value - calibrated value) / calibrated value. For example, if the measured rising edge steepness is 20 A / μs and the calibrated value is 19 A / μs, the relative steepness increment is approximately (20-19) / 19 = 0.053. When the calculation result is negative, the relative steepness increment is zero.

[0074] Furthermore, the measured current fallback time obtained in the previous steps is compared with the calibrated fallback time, and the relative increase in fallback time is calculated as (measured value - calibrated value) / calibrated value. For example, if the measured fallback time is 10.5 μs and the calibrated value is 7 μs, the relative increase in fallback time is (10.5 - 7) / 7 = 0.5. When the calculation result is negative, the relative increase in fallback time is zero.

[0075] Furthermore, at the start of the current control cycle, the original value of the peak voltage of the pulse circuit and the original value of the average output voltage of the base voltage circuit are acquired and latched. The ratio between the two is calculated as the dynamic energy ratio. For example, if the latched pulse peak voltage is 80kV and the average output voltage of the base voltage is 50kV, the dynamic energy ratio = 80kV / 50kV = 1.6.

[0076] Furthermore, multiplying the relative increment of steepness by the dynamic energy ratio yields the steepness contribution term, which is calculated as: Steepness Contribution Term = Relative Increment of Steepness × Dynamic Energy Ratio. This term reflects the degree to which the change in the steepness of the rising edge contributes to the transient response of the electric field. For example, the steepness contribution term = 0.053 × 1.6 ≈ 0.085.

[0077] Furthermore, multiplying the relative increment of the fallback time by the dynamic energy ratio yields the fallback contribution term, which equals the relative increment of the fallback time multiplied by the dynamic energy ratio. This term reflects the degree to which the change in the current fallback time contributes to the transient response of the electric field. For example, the fallback contribution term = 0.5 × 1.6 = 0.8.

[0078] Finally, the steepness contribution term and the fallback contribution term are added together to obtain the electric field transient response index. The electric field transient response index comprehensively quantifies the degree to which the transient response of the load under pulse excitation deviates from the calibrated operating condition. For example, the electric field transient response index = 0.085 + 0.8 = 0.885. This indicates that the current load transient response is generally worse than the calibrated operating condition, and a secondary correction of the pulse peak voltage is required.

[0079] The generated electric field transient response exponent is input into the pre-built electric field transient response compensation model deployed in the controller. The model outputs a peak voltage quadratic correction factor. The training steps of the electric field transient response compensation model include: Multiple training samples are extracted from historical operating data. For each training sample, the corresponding electric field transient response index is obtained, and the corresponding ideal peak voltage quadratic correction factor is determined. The ideal peak voltage quadratic correction factor is calculated by back-calculating the pulse peak voltage given value and the required dynamic peak voltage when the electric field is in the ideal response state. The electric field transient response index is used as the sample input feature, and the ideal peak voltage quadratic correction factor is used as the sample output label to form a training dataset; A transient response compensation model for the initial electric field is constructed based on machine learning. The model parameters are optimized by taking the electric field transient response exponent in the training dataset as the model input, the corresponding ideal peak voltage quadratic correction factor as the supervision signal, and minimizing the error between the peak voltage quadratic correction factor output by the model and the supervision signal. Training stops when the verification error converges to the target range, resulting in a trained electric field transient response compensation model.

[0080] Specifically, multiple training samples are first extracted from the historical operating data of the electrostatic precipitator, with each sample covering a complete load transient response event cycle. For each sample, the electric field transient response index is calculated as described in the previous steps, and the corresponding ideal peak voltage quadratic correction factor is determined.

[0081] Among them, the ideal peak voltage second correction factor is the ratio obtained by back-calculating the given value of the pulse peak voltage when the electric field is in an ideal response state, and then obtaining the dynamic peak voltage required to achieve the best discharge effect.

[0082] Specifically, within the calibration period, a candidate peak voltage sequence is generated with the given pulse peak voltage as the upper limit and 50% of the given pulse peak voltage as the lower limit, according to a preset step size. The pulse peak voltage is adjusted point by point to each candidate value and run for several pulse cycles, recording the dust removal efficiency, flashover rate, and single-pulse injection energy. The discharge effect corresponding to each candidate value is quantified using the composite evaluation function F=α×η+β×(1-R)+γ×(1-E / E0), where η is the dust removal efficiency, R is the flashover rate, E is the single-pulse injection energy, E0 is the rated injection energy, and α, β, and γ are weighting coefficients, with β taking the maximum weight to prioritize low flashover rate. After eliminating candidate values ​​with flashover rates exceeding the safety limit, the peak voltage corresponding to the largest F value is selected as the optimal dynamic peak voltage. The optimal dynamic peak voltage is divided by the given pulse peak voltage to obtain the ideal peak voltage secondary correction factor.

[0083] It should be noted that the aforementioned use of composite evaluation functions is a mature comprehensive evaluation technique in the existing technology, widely used in industrial process control, multi-objective decision optimization, and other fields. Specifically, the online calculation of dust removal efficiency η is based on real-time monitoring data from inlet and outlet dust concentration sensors, which is a conventional measurement technique in this field; the statistical calculation of flashover rate R is based on the detection of sudden changes in secondary voltage or secondary current within the pulse period, which is a conventional flashover discrimination method in the field of electrostatic precipitator control; the single-pulse injected energy E is obtained by integrating the instantaneous product of pulse voltage and pulse current over the pulse width time, which is a conventional energy calculation method in the field of pulse power supplies. The methods for obtaining the above sub-indicators are all existing technologies well-known to those skilled in the art, and will not be elaborated further here.

[0084] For example, in a certain historical operation, the transient response index of the electric field was 0.885. Under the current operating conditions, the pulse peak voltage setpoint is 80kV, but on-site commissioning found that adjusting the peak voltage to 60kV resulted in the best discharge effect and no flashover risk. Therefore, the ideal peak voltage second correction factor = 60kV / 80kV = 0.75.

[0085] Secondly, the transient response exponent of the electric field for each training sample is used as the input feature, and the corresponding ideal peak voltage quadratic correction factor is used as the output label, which are combined to form the training dataset. For example, the input feature of a sample is 0.885, and the output label is 0.75. A total of 5,000 similar valid samples are extracted from the historical data to form a training dataset containing 5,000 records.

[0086] Next, an initial electric field transient response compensation model is constructed based on a machine learning regression algorithm. The model structure can be a regression model suitable for nonlinear mappings, such as support vector regression, neural networks, or random forests. For example, a three-layer BP neural network is chosen as the initial model, with one neuron in the input layer to receive the electric field transient response exponent, eight neurons in the hidden layer, and one neuron in the output layer to output the peak voltage quadratic correction factor. The sigmoid function is selected as the activation function.

[0087] Then, using the electric field transient response exponent in the training dataset as the model input and the corresponding ideal peak voltage quadratic correction factor as the supervision signal, the error between the peak voltage quadratic correction factor output by the model and the supervision signal is calculated. With minimizing the error as the optimization objective, the model parameters are iteratively updated through algorithms such as backpropagation or gradient descent.

[0088] For example, if 5000 samples are input into the neural network in batches, and the model outputs 0.82 after inputting 0.885 to a certain sample, the mean square error between the output and the supervision signal of 0.75 is (0.82-0.75). 2=0.0049. This error is backpropagated using the Adam optimizer, adjusting the weights and biases of neurons in each layer to gradually bring the model output closer to 0.75.

[0089] Finally, during the training process, the model error is evaluated periodically using a validation set. When the validation error converges to the preset target range, training is stopped, the model parameters are saved, and the trained electric field transient response compensation model is obtained and deployed to the controller for online use.

[0090] For example, after 200 iterations, the root mean square error on the validation set stabilizes below 0.015, meeting the engineering accuracy requirements. Training is then stopped, and the model parameters are permanently written into the controller. During online operation, when the calculated electric field transient response exponent is 0.885 and the model output peak voltage quadratic correction factor is 0.76, the peak voltage quadratic correction factor can be used for subsequent composite correction calculations.

[0091] To prevent the load characteristics from slowly drifting due to factors such as equipment aging and seasonal changes in operating conditions, which would gradually reduce the model's mapping accuracy, an online incremental update mechanism is added to the model: during periods when the electric field is in an ideal response state and the control effect is manually confirmed to be excellent, the current electric field transient response index and the corresponding ideal correction factor are added as new samples to the training set, and the model is fine-tuned online or periodically retrained to maintain the model's long-term adaptability to slowly changing load characteristics.

[0092] It should be noted that the above values ​​are for illustrative purposes only and do not constitute a limitation on the present invention.

[0093] In summary, this step transforms the multi-dimensional characteristics of the load transient response into a refined compensation amount for the pulse peak voltage, effectively compensating for the insufficient accuracy of correction based solely on the fall-off time dimension. This allows the peak voltage adjustment to more accurately match the actual load condition, improving the discharge control quality and safety under complex operating conditions.

[0094] S500: Using the peak compensation coefficient and the peak voltage secondary correction factor, a composite correction coefficient is formed, and then the pulse peak voltage setpoint is adjusted according to the composite correction coefficient to obtain the dynamic peak voltage.

[0095] This step integrates two different dimensions of peak voltage correction: the peak compensation coefficient provides a basic correction from the perspective of load discharge recovery speed, while the peak voltage secondary correction factor provides refined compensation from the perspective of comprehensive evaluation of electric field transient response. By fusing them into a unified composite correction coefficient, multi-level adaptive adjustment of the pulse peak voltage setpoint is achieved, ensuring that the dynamic peak voltage can both match the load recovery characteristics and meet the comprehensive requirements of transient response.

[0096] Step S500 in the method provided by the present invention includes: Obtain the peak compensation coefficient and the peak voltage quadratic correction factor; The ratio of the peak voltage output by the pulse circuit during the peak holding phase to the average output voltage of the base voltage circuit is obtained as the dynamic energy ratio; Based on the dynamic energy ratio, a first fusion weight and a second fusion weight are determined, wherein the first fusion weight is positively correlated with the dynamic energy ratio, the second fusion weight is negatively correlated with the dynamic energy ratio, and the sum of the first fusion weight and the second fusion weight is 1. The peak voltage secondary correction factor is weighted using the first fusion weight, and the peak compensation coefficient is weighted using the second fusion weight. The weighted results are then added together to obtain the composite correction coefficient. The dynamic peak voltage is obtained by correcting the given value of the pulse peak voltage based on the composite correction coefficient, including: When the composite correction coefficient is less than 1, the pulse peak voltage setpoint is multiplied by the composite correction coefficient, and the result is used as the dynamic peak voltage. When the composite correction coefficient is greater than or equal to 1, the given value of the pulse peak voltage is directly used as the dynamic peak voltage.

[0097] In this step, the peak compensation coefficient and peak voltage quadratic correction factor for the current control cycle are first obtained from the previous steps and used as two inputs for the fusion calculation. For example, the peak compensation coefficient calculated in the previous steps is 0.667, and the peak voltage quadratic correction factor output by the electric field transient response compensation model is 0.76.

[0098] Secondly, the original values ​​of the peak voltage of the pulse circuit and the average output voltage of the base voltage circuit, acquired and latched at the beginning of the current control cycle, are obtained, and their ratio is calculated as the dynamic energy ratio. For example, if the latched peak voltage of the pulse is 80kV and the average output voltage of the base voltage is 50kV, the dynamic energy ratio = 80kV / 50kV = 1.6.

[0099] Next, the first fusion weight and the second fusion weight are calculated based on the dynamic energy ratio. The first fusion weight is positively correlated with the dynamic energy ratio, and the second fusion weight is negatively correlated with the dynamic energy ratio; their sum is always 1. Specifically, the preset fusion weight calculation rules are: first fusion weight = dynamic energy ratio / (dynamic energy ratio + 1); second fusion weight = 1 - first fusion weight.

[0100] For example, the first fusion weight = dynamic energy ratio / (dynamic energy ratio + 1) = 1.6 / 2.6 ≈ 0.615; the second fusion weight = 1 - the first fusion weight = 1 - 0.615 = 0.385. The larger the dynamic energy ratio, the higher the energy proportion of the pulse circuit relative to the base voltage circuit, and the higher the confidence level of the secondary correction factor output by the model should be.

[0101] Then, the peak voltage secondary correction factor is weighted by the first fusion weight, and the peak compensation coefficient is weighted by the second fusion weight. The weighted results are added together to obtain the composite correction coefficient. Specifically, the composite correction coefficient = first fusion weight × peak voltage secondary correction factor + second fusion weight × peak compensation coefficient. For example, the composite correction coefficient = 0.615 × 0.76 + 0.385 × 0.667 ≈ 0.467 + 0.257 = 0.724.

[0102] Finally, the composite correction coefficient is compared with the pulse peak voltage setpoint to obtain the dynamic peak voltage. When the composite correction coefficient is less than 1, the pulse peak voltage setpoint is multiplied by the composite correction coefficient to obtain the dynamic peak voltage; when the composite correction coefficient is greater than or equal to 1, the pulse peak voltage setpoint is kept unchanged.

[0103] For example, if the given pulse peak voltage is 80kV and the composite correction factor is 0.724 < 1, then the dynamic peak voltage = 80kV × 0.724 ≈ 57.9kV. The dynamic peak voltage being lower than the given value indicates that the pulse peak voltage needs to be reduced under the current load conditions to match the load recovery characteristics. If, under another operating condition, the composite correction factor is 1.05 ≥ 1, then the dynamic peak voltage can be directly taken as 80kV without reduction.

[0104] It should be noted that the above values ​​are for illustrative purposes only and do not constitute a limitation on the present invention.

[0105] In summary, this step achieves a two-dimensional coordinated response of the pulse peak voltage to the combined state of load discharge recovery characteristics and electric field transient response: when the composite correction coefficient is less than 1, the peak voltage is actively reduced to avoid overexcitation; when the correction requirement disappears, the given value is maintained to ensure output capability, thereby reducing flashover risk and energy waste while ensuring discharge effect.

[0106] S600: Calculate the pulse width compression coefficient based on the dynamic base voltage and the dynamic peak voltage, and reduce the current pulse width in real time to obtain the limited pulse width.

[0107] like Figure 2As shown, this step, based on the parameters obtained above and with the base voltage and peak voltage already dynamically adjusted, further modifies the pulse width. When the base voltage decreases or the peak voltage is lowered, the energy injection time required by the load in a single pulse cycle is correspondingly shortened, avoiding energy redundancy or excessive discharge caused by the pulse width not being reduced synchronously due to the voltage parameters being lowered, thus achieving full-dimensional coordinated control of the base voltage, peak voltage, and pulse width.

[0108] Step S600 in the method provided by the present invention includes: Based on the dynamic base voltage and the dynamic peak voltage, a base compression coefficient is determined, wherein the base compression coefficient increases with the increase of the dynamic base voltage and decreases with the increase of the dynamic peak voltage; Based on the steepness of the secondary current rise edge and the calibrated rise edge, a steepness correction factor is generated, and the basic compression factor is multiplied by the steepness correction factor to obtain the steepness correction compression factor; Based on the current fall-off time and the calibrated fall-off time, a fall-off correction factor is generated, and combined with the steepness correction compression coefficient, a composite compression coefficient is obtained. Obtain the dynamic energy ratio. If the dynamic energy ratio is greater than 1, divide the composite compression coefficient by the dynamic energy ratio to obtain the energy-corrected compression coefficient. When the dynamic energy ratio is not greater than 1, the composite compression coefficient is directly used as the energy-corrected compression coefficient. The smaller of the energy correction compression factor and the value 1 is selected as the final pulse width compression factor, and the current pulse width is multiplied by the final pulse width compression factor to obtain the amplitude-limited pulse width.

[0109] In this step, the base compression factor is first calculated based on the dynamic base voltage and dynamic peak voltage obtained in the previous step. The base compression factor increases with the increase of the dynamic base voltage and decreases with the increase of the dynamic peak voltage. The preset formula for calculating the base compression factor is: Base compression factor = Dynamic base voltage / Dynamic peak voltage.

[0110] For example, the dynamic base voltage is 41.15kV and the dynamic peak voltage is 57.9kV. Therefore, the base compression factor = 41.15kV / 57.9kV ≈ 0.711. A higher dynamic base voltage allows for more energy injection into the load, thus allowing for a wider pulse width; conversely, a higher dynamic peak voltage results in greater energy per pulse, requiring a corresponding reduction in pulse width.

[0111] It should be noted that if the dynamic peak voltage is affected by some extreme situation, such as sensor failure or an initial power-on transient value of 0 or close to 0, it will lead to division by zero anomaly or an extremely large coefficient. Therefore, a limiting protection is added after the basic compression coefficient calculation step: when the dynamic peak voltage is lower than the preset minimum threshold, the basic compression coefficient is forcibly set to 1 to avoid division by zero anomaly.

[0112] Secondly, based on the current secondary current rising edge steepness and the calibrated rising edge steepness obtained in the previous steps, a steepness correction factor is generated. The steepness correction factor = calibrated value / measured value. The basic compression factor is multiplied by the steepness correction factor to obtain the steepness correction compression factor.

[0113] For example, the current measured rise edge steepness is 20 A / μs, and the calibrated rise edge steepness is 19 A / μs. The steepness correction factor = 19 / 20 = 0.95. The steepness correction compression factor = base compression factor × steepness correction factor = 0.711 × 0.95 ≈ 0.675. A steeper rise edge indicates enhanced load absorption capacity, requiring further pulse width compression to avoid over-energy injection.

[0114] Next, based on the current current fall-off time and the calibrated fall-off time obtained in the previous steps, a fall-off correction factor is generated, where the fall-off correction factor = calibrated value / measured value. The fall-off correction factor is combined with the steepness correction compression coefficient to obtain the composite compression coefficient, where the composite compression coefficient = steepness correction compression coefficient × fall-off correction factor.

[0115] For example, the current measured fallback time is 10.5 μs, and the calibrated fallback time is 7 μs. The fallback correction factor = 7 / 10.5 ≈ 0.667. The composite compression coefficient = 0.675 × 0.667 ≈ 0.450. A prolonged fallback time indicates a slower load discharge recovery rate, requiring further pulse width compression to prevent the next energy injection before the load has fully recovered.

[0116] Then, the dynamic energy ratio is obtained, and the composite compression coefficient is differentially corrected based on whether its value is greater than 1. When the dynamic energy ratio is greater than 1, the composite compression coefficient is divided by the dynamic energy ratio to obtain the energy-corrected compression coefficient; when the dynamic energy ratio is not greater than 1, the composite compression coefficient is directly used as the energy-corrected compression coefficient.

[0117] For example, if the dynamic energy ratio is 1.6 > 1, then the energy correction compression factor = composite compression factor / dynamic energy ratio = 0.450 / 1.6 ≈ 0.281. A dynamic energy ratio greater than 1 indicates that the pulse energy ratio is too high. Dividing by the dynamic energy ratio can further reduce the pulse width to balance the total energy injection.

[0118] In another possible embodiment, the dynamic base voltage is 55kV, the dynamic peak voltage is 50kV, and the dynamic energy ratio = 50kV / 55kV ≈ 0.91 ≤ 1. In this case, the composite compression coefficient, calculated from the first three steps, is 0.85. Since the dynamic energy ratio is not greater than 1, it indicates that the pulse energy proportion is not too high, and no further correction is needed; the energy correction compression coefficient is directly taken as 0.85. Under this condition, the base voltage energy supply is relatively sufficient, and the pulse width reduction is relatively mild, matching the current energy structure.

[0119] Finally, the smaller of the energy correction compression factor and the value 1 is selected as the final pulse width compression factor to ensure that the pulse width only decreases and does not increase. The current pulse width is multiplied by the final pulse width compression factor to obtain the amplitude-limited pulse width.

[0120] The current pulse width is the amplitude-limited pulse width output and actually executed by the coordinated control instruction in the previous control cycle; that is, the current pulse width equals the amplitude-limited pulse width effective in the previous control cycle. The current pulse width is latched into a buffer register at the end of each control cycle, serving as the reference input for pulse width compression calculation in the next control cycle.

[0121] Specifically, during the initial control cycle of the first power-on or the first switch from the conventional fixed parameter mode to the dual-group coordinated pulse power supply mode, the current pulse width is taken as the system's preset rated pulse width as the initial reference value. The rated pulse width is a fixed pulse width value preset according to the electrostatic precipitator design parameters and process requirements, and is stored in the controller's non-volatile memory.

[0122] For example, the energy correction compression factor is 0.281, which is the smaller of 1, resulting in a final pulse width compression factor of 0.281. The current pulse width is taken as the preset rated pulse width of 10μs as the initial reference, so the limited pulse width = 10μs × 0.281 ≈ 2.81μs. After pulse width reduction, it is output in coordination with the reduced dynamic base voltage of 41.15kV and dynamic peak voltage of 57.9kV to ensure that the load receives a matching energy injection under conditions of slower recovery and reduced energy demand, thus avoiding over-discharge.

[0123] For example, in the next cycle, the width of the limiting pulse output and effective in the previous control cycle is 2.81 μs, which is latched and used as the current pulse width for the next control cycle. Then, the final pulse width compression coefficient calculated in the next control cycle is 0.281, and the limiting pulse width = 2.81 μs × 0.281 ≈ 0.79 μs, realizing the iterative reduction of the pulse width cycle by cycle.

[0124] It should be noted that the above values ​​are for illustrative purposes only and do not constitute a limitation on the present invention.

[0125] In summary, this step achieves automatic tracking of pulse width adjustment results for base voltage and peak voltage: when the base voltage decreases or the peak voltage is reduced, the pulse width is synchronously shortened to avoid energy redundancy; when the load absorption is enhanced or the recovery is slowed down, the pulse width is further compressed to prevent over-discharge, ensuring that the energy injected by a single pulse is always precisely matched with the real-time state of the load, thus eliminating the potential energy runaway caused by the voltage parameters being adjusted but the pulse width not being synchronized in traditional fixed pulse width control from a time perspective.

[0126] S700: Outputs a coordinated control command carrying the dynamic base voltage, the dynamic peak voltage, and the limiting pulse width to drive the base voltage circuit to output the dynamic base voltage and drive the pulse circuit to output a high-voltage pulse according to the dynamic peak voltage within the limiting pulse width.

[0127] In this step, the dynamic base voltage, dynamic peak voltage, and limiting pulse width obtained in the previous step are encapsulated to generate a coordinated control command. The coordinated control command carries three parameters, allowing the base voltage circuit and the pulse circuit to execute synchronously within the current control cycle.

[0128] For example, in the previous example, the dynamic base voltage was 41.15kV, the dynamic peak voltage was 57.9kV, and the limiting pulse width was 2.81μs. These three are encapsulated into a single coordinated control command frame, which includes a command header, parameter segment, and checksum, and is sent to the drive unit through the controller communication bus.

[0129] Next, the base voltage circuit drive unit parses the coordinated control command, extracts the dynamic base voltage value, and adjusts the modulation ratio or phase shift angle of the base voltage inverter so that the output voltage of the base voltage circuit follows the dynamic base voltage value in real time.

[0130] For example, when the base voltage circuit drive unit receives a command value of 41.15kV for the dynamic base voltage, the current base voltage setpoint is adjusted from 50kV to 41.15kV. Through PID closed-loop regulation, the voltage across the base voltage output capacitor is stabilized at 41.15kV, providing the base electric field with an energy base that matches the current load state.

[0131] Finally, the pulse loop drive unit parses the coordinated control command and extracts the dynamic peak voltage and the limiting pulse width. When the pulse trigger moment arrives, the pulse loop outputs a high-voltage pulse according to the dynamic peak voltage. To prevent the IGBT from hard-shutdown due to sudden pulse width changes, an adjustment rate limit is applied to the target limiting pulse width: the maximum pulse width adjustment step is defined as 20% of the preset rated pulse width, meaning the pulse width reduction within each control cycle does not exceed 2μs, or 20% of the rated pulse width, ensuring a smooth transition in pulse width changes. This completes the pulse output.

[0132] Specifically, the current pulse width is compared with the target limiting pulse width: if the absolute value of the difference between the current pulse width and the target limiting pulse width does not exceed the maximum adjustment step size, the target limiting pulse width is used as the actual limiting pulse width output in this cycle; if it exceeds the maximum adjustment step size, the value of the current pulse width after adjusting it one maximum adjustment step size in the target direction is used as the actual limiting pulse width output in this cycle.

[0133] For example, the pulse loop drive unit receives a dynamic peak voltage of 57.9kV, and the target limiting pulse width is calculated to be 2.81μs. The current pulse width is 10μs, and the reduction is 10μs - 2.81μs = 7.19μs, exceeding the maximum adjustment step size by 2μs. Therefore, the actual output limiting pulse width for this control cycle is 10μs - 2μs = 8μs. At the beginning of the pulse cycle, the pulse switch is turned on, the primary energy stored in the pulse transformer is released, and the secondary output peak voltage of 57.9kV is superimposed on the base voltage of 41.15kV. After 8μs, the controller turns off the pulse switch, and the pulse output for this cycle terminates.

[0134] Upon entering the next control cycle, data is reacquired and recalculated. Assuming the operating conditions remain unchanged, the final pulse width compression coefficient is still 0.281. The current pulse width has now been updated to 8 μs. Therefore, the target limiting pulse width = 8 μs × 0.281 ≈ 2.25 μs, with a reduction of 8 μs - 2.25 μs = 5.75 μs, still exceeding the maximum adjustment step size by 2 μs. Thus, the actual output limiting pulse width for this cycle = 8 μs - 2 μs = 6 μs, and the pulse switch is turned off after 6 μs. Similarly, in the fifth control cycle, the current pulse width is updated to 2 μs, the target limiting pulse width = 2 μs × 0.281 ≈ 0.56 μs, with a reduction of 2 μs - 0.56 μs = 1.44 μs, not exceeding the maximum adjustment step size by 2 μs, and the actual output target limiting pulse width is 0.56 μs.

[0135] Through a smooth transition of approximately 4 to 5 control cycles, the pulse width gradually decreases from 10 μs to near the safety lower limit. Compared to the peak voltage of 80 kV and pulse width of 10 μs in the conventional mode, the peak voltage of the pulse during the transition period is reduced by 27.6%. The pulse width gradually decreases with the cycle, and the total injected energy decreases steadily, gradually matching the actual state of slower load recovery and reduced energy demand. It also effectively avoids the risk of IGBT hard turn-off voltage spikes caused by single-step sudden changes.

[0136] It should be noted that the above values ​​are for illustrative purposes only and do not constitute a limitation on the present invention.

[0137] In summary, this step achieves real-time linkage of base voltage amplitude, pulse peak value, and pulse width across all dimensions: the base voltage circuit provides a dynamic base electric field that matches the load's energy requirements, and the pulse circuit completes a single energy injection with a peak voltage that adapts to the load's recovery characteristics within a precisely limited pulse width. This eliminates parameter mismatch between the two circuits in both time and amplitude dimensions, ensuring that the total injected energy is highly consistent with the real-time state of the load, and maximizing system energy efficiency and operational safety while ensuring discharge performance.

[0138] In summary, this invention fundamentally eliminates the total energy mismatch problem caused by the independent operation of base voltage, peak voltage, and pulse width in traditional fixed-parameter control. It actively reduces the base voltage, shortens the peak voltage, and simultaneously compresses the pulse width when load absorption increases or recovery slows down, while maintaining output capability when the load recovers. This ensures the instantaneous power surge required by the process while effectively suppressing the risks of over-discharge and flashover, ultimately improving the energy efficiency, reliability, and dynamic performance of the high-frequency, high-voltage power supply system.

[0139] Example 2, as Figure 3 As shown, the present invention provides a high-frequency power supply dual-group cooperative output control system, the system comprising: The transient detection and mode triggering module 11 is used to trigger the dual-group cooperative pulse power supply mode when the instantaneous change rate of the secondary current exceeds the preset fluctuation threshold.

[0140] The preset fluctuation threshold is determined by calculating the algebraic sum of the mean and standard deviation of the historical values ​​of the instantaneous change rate of multiple secondary currents continuously collected by the electrostatic precipitator under dry flue gas conditions. The dual-group coordinated pulse power supply mode is that the base voltage circuit enters the voltage follower output state, the pulse circuit enters the controlled intermittent chopping output state, and the pulse circuit stops outputting after each pulse cycle until the start of the next pulse cycle.

[0141] The feature parameter extraction module 12 is used to obtain the average output voltage of the base voltage circuit and the peak voltage of the pulse circuit during the peak holding phase in the current control cycle, and to determine the dynamic feature parameters, wherein the dynamic feature parameters include at least the steepness of the secondary current rise edge and the current fall time.

[0142] This includes obtaining the average output voltage of the base voltage circuit and the peak voltage of the pulse circuit during the peak hold phase within the current control cycle, and determining dynamic characteristic parameters, including: Within each pulse cycle, the instantaneous value of the secondary current is acquired at a fixed sampling frequency. Based on the change in the secondary current between the pulse start time and the time when it rises to the peak value, and the corresponding time length, the rising edge steepness of the secondary current is determined. At the end of the pulse cycle, the current fall-off time is determined based on the time length between the pulse turn-off time and the time when the secondary current drops to a specified proportion of the current value at the turn-off time. The average value obtained by averaging multiple instantaneous voltage values ​​obtained by sampling the base voltage circuit output voltage at equal intervals multiple times within the current control cycle is used as the average output voltage of the base voltage circuit. The maximum voltage value output by the peak hold circuit in the detection pulse loop within the current pulse period is taken as the peak voltage output by the pulse loop during the peak hold phase.

[0143] The calculation steps for the specified ratio include: Under the calibration conditions of introducing dry flue gas into the electrostatic precipitator and no flashover of the electric field, the waveform of the secondary current falling segment of multiple pulse cycles was continuously collected. For the secondary current decline segment of each pulse cycle, the current decay rate is calculated point by point starting from the pulse turn-off time. The ratio of the current value when the current decay rate decreases from the maximum value to one-tenth of the maximum value to the current value at the turn-off time is used as a single-cycle proportional sample. Sort the single-cycle proportion samples obtained from all pulse cycles according to their numerical values, remove the extreme samples at both ends, calculate the arithmetic mean of the remaining samples, and use the arithmetic mean as the specified proportion.

[0144] The base voltage correction module 13 is used to correct the base voltage given value based on the steepness of the secondary current rise edge to obtain the dynamic base voltage, and at the same time obtain the calibration fall-off time, and determine the peak compensation coefficient in combination with the current fall-off time.

[0145] Specifically, the dynamic base voltage is obtained by correcting the base voltage setpoint based on the steepness of the secondary current rise, and the calibrated fall-off time is acquired. The peak compensation coefficient is then determined by combining the current fall-off time with the calibrated fall-off time. Under calibration conditions, the calibration rise edge steepness is obtained using the same method as that used to determine the rise edge steepness of the secondary current. Based on the relative increase in the rise edge steepness of the secondary current relative to the calibrated rise edge steepness, a steepness suppression component is determined; The fallback suppression component is determined based on the relative extension of the current fallback time relative to the calibrated fallback time; The ratio of the peak voltage output by the pulse circuit during the peak holding phase to the average output voltage of the base voltage circuit is obtained as the dynamic energy ratio; The pressure drop is determined based on the steepness suppression component, the fallback suppression component, and the dynamic energy ratio, wherein the pressure drop increases with the increase of the steepness suppression component, the fallback suppression component, and the dynamic energy ratio; Subtracting the voltage drop from the given base voltage value yields the dynamic base voltage. Divide the calibrated fallback time by the current fallback time, and use the resulting value as the peak compensation coefficient.

[0146] The transient response evaluation module 14 is used to generate an electric field transient response index based on the steepness of the secondary current rise edge and the current fall time, and input the electric field transient response index into a preset electric field transient response compensation model to obtain a peak voltage secondary correction factor.

[0147] The electric field transient response index is generated based on the rise steepness of the secondary current and the fall time of the current, including: Under the calibration conditions, obtain the calibration rise steepness and calibration fall time; The relative increment of steepness is determined based on the relative increase of the secondary current rising edge steepness relative to the calibrated rising edge steepness; The relative increment of the fall-off time is determined based on the relative extension of the current fall-off time relative to the calibrated fall-off time; The ratio of the peak voltage output by the pulse circuit during the peak holding phase to the average output voltage of the base voltage circuit is obtained as the dynamic energy ratio; Based on the relative steepness increment and the dynamic energy ratio, a steepness contribution term is determined, wherein the steepness contribution term is proportional to both the relative steepness increment and the dynamic energy ratio; Based on the relative increment of fallback time and the dynamic energy ratio, a fallback contribution term is determined, wherein the fallback contribution term is proportional to both the relative increment of fallback time and the dynamic energy ratio; The steepness contribution term and the fallback contribution term are combined to obtain the electric field transient response index.

[0148] The training steps for the electric field transient response compensation model include: Multiple training samples are extracted from historical operating data. For each training sample, the corresponding electric field transient response index is obtained, and the corresponding ideal peak voltage secondary correction factor is determined. The ideal peak voltage secondary correction factor is calculated by back-calculating the pulse peak voltage given value and the required dynamic peak voltage when the electric field is in the ideal response state. The electric field transient response index is used as the sample input feature, and the ideal peak voltage quadratic correction factor is used as the sample output label to form a training dataset; A transient response compensation model for the initial electric field is constructed based on machine learning. The model parameters are optimized by taking the electric field transient response exponent in the training dataset as the model input, the corresponding ideal peak voltage quadratic correction factor as the supervision signal, and minimizing the error between the peak voltage quadratic correction factor output by the model and the supervision signal. Training stops when the verification error converges to the target range, resulting in a trained electric field transient response compensation model.

[0149] The peak voltage composite correction module 15 is used to form a composite correction coefficient by using the peak compensation coefficient and the peak voltage secondary correction factor, and then adjust the pulse peak voltage setpoint according to the composite correction coefficient to obtain the dynamic peak voltage.

[0150] Specifically, a composite correction coefficient is formed by using the peak compensation coefficient and the peak voltage quadratic correction factor, and then the pulse peak voltage setpoint is adjusted according to the composite correction coefficient to obtain the dynamic peak voltage, including: Obtain the peak compensation coefficient and the peak voltage quadratic correction factor; The ratio of the peak voltage output by the pulse circuit during the peak holding phase to the average output voltage of the base voltage circuit is obtained as the dynamic energy ratio; Based on the dynamic energy ratio, a first fusion weight and a second fusion weight are determined, wherein the first fusion weight is positively correlated with the dynamic energy ratio, the second fusion weight is negatively correlated with the dynamic energy ratio, and the sum of the first fusion weight and the second fusion weight is 1. The peak voltage secondary correction factor is weighted using the first fusion weight, and the peak compensation coefficient is weighted using the second fusion weight. The weighted results are then added together to obtain the composite correction coefficient. The dynamic peak voltage is obtained by correcting the given value of the pulse peak voltage based on the composite correction coefficient, including: When the composite correction coefficient is less than 1, the pulse peak voltage setpoint is multiplied by the composite correction coefficient, and the result is used as the dynamic peak voltage. When the composite correction coefficient is greater than or equal to 1, the given value of the pulse peak voltage is directly used as the dynamic peak voltage.

[0151] The pulse width limiting calculation module 16 is used to calculate the pulse width compression coefficient based on the dynamic base voltage and the dynamic peak voltage, and to reduce the current pulse width in real time to obtain the limited pulse width.

[0152] Specifically, based on the dynamic base voltage and the dynamic peak voltage, a pulse width compression coefficient is calculated, and the current pulse width is reduced in real time to obtain a limited pulse width, including: Based on the dynamic base voltage and the dynamic peak voltage, a base compression coefficient is determined, wherein the base compression coefficient increases with the increase of the dynamic base voltage and decreases with the increase of the dynamic peak voltage; Based on the steepness of the secondary current rise edge and the calibrated rise edge, a steepness correction factor is generated, and the basic compression factor is multiplied by the steepness correction factor to obtain the steepness correction compression factor; Based on the current fall-off time and the calibrated fall-off time, a fall-off correction factor is generated, and combined with the steepness correction compression coefficient, a composite compression coefficient is obtained. Obtain the dynamic energy ratio. If the dynamic energy ratio is greater than 1, divide the composite compression coefficient by the dynamic energy ratio to obtain the energy-corrected compression coefficient. When the dynamic energy ratio is not greater than 1, the composite compression coefficient is directly used as the energy-corrected compression coefficient. The smaller of the energy correction compression factor and the value 1 is selected as the final pulse width compression factor, and the current pulse width is multiplied by the final pulse width compression factor to obtain the amplitude-limited pulse width.

[0153] The collaborative control command output module 17 is used to output a collaborative control command carrying the dynamic base voltage, the dynamic peak voltage and the limiting pulse width, so as to drive the base voltage circuit to output the dynamic base voltage and drive the pulse circuit to output a high voltage pulse according to the dynamic peak voltage within the limiting pulse width.

[0154] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0155] This specification and accompanying drawings are merely illustrative examples of the invention and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its scope. Therefore, if such modifications and modifications fall within the scope of the invention and its equivalents, the invention is intended to include these modifications and modifications.

Claims

1. A high-frequency power supply dual-group coordinated output control method, characterized in that, The method includes: When the instantaneous rate of change of the secondary current exceeds the preset fluctuation threshold, the dual-group cooperative pulse power supply mode is triggered. The average output voltage of the base voltage circuit and the peak voltage of the pulse circuit during the peak holding phase are obtained within the current control cycle, and the dynamic characteristic parameters are determined. The dynamic characteristic parameters include at least the steepness of the secondary current rise edge and the current fall time. The base voltage is corrected based on the steepness of the secondary current rise to obtain the dynamic base voltage. At the same time, the calibration fall-off time is obtained, and the peak compensation coefficient is determined in combination with the current fall-off time. Based on the steepness of the secondary current rise and the current fall time, an electric field transient response index is generated, and the electric field transient response index is input into a preset electric field transient response compensation model to obtain a peak voltage secondary correction factor. A composite correction coefficient is formed by using the peak compensation coefficient and the peak voltage second correction factor. The pulse peak voltage setpoint is then adjusted according to the composite correction coefficient to obtain the dynamic peak voltage. Based on the dynamic base voltage and the dynamic peak voltage, the pulse width compression coefficient is calculated, and the current pulse width is reduced in real time to obtain the limited pulse width; Output a coordinated control command carrying the dynamic base voltage, the dynamic peak voltage, and the limiting pulse width to drive the base voltage circuit to output the dynamic base voltage and drive the pulse circuit to output a high-voltage pulse according to the dynamic peak voltage within the limiting pulse width.

2. The high-frequency power supply dual-group coordinated output control method according to claim 1, characterized in that, The preset fluctuation threshold is determined by calculating the algebraic sum of the mean and standard deviation of the historical values ​​of the instantaneous change rate of multiple secondary currents continuously collected by the electrostatic precipitator under dry flue gas conditions. The dual-group coordinated pulse power supply mode is that the base voltage circuit enters the voltage follower output state, the pulse circuit enters the controlled intermittent chopping output state, and the pulse circuit stops outputting after each pulse cycle until the start of the next pulse cycle.

3. The high-frequency power supply dual-group coordinated output control method according to claim 1, characterized in that, Obtain the average output voltage of the base voltage circuit and the peak voltage of the pulse circuit during the peak hold phase within the current control cycle, and determine the dynamic characteristic parameters, including: Within each pulse cycle, the instantaneous value of the secondary current is acquired at a fixed sampling frequency. Based on the change in the secondary current between the pulse start time and the time when it rises to the peak value, and the corresponding time length, the rising edge steepness of the secondary current is determined. At the end of the pulse cycle, the current fall-off time is determined based on the time length between the pulse turn-off time and the time when the secondary current drops to a specified proportion of the current value at the turn-off time. The average value obtained by averaging multiple instantaneous voltage values ​​obtained by sampling the base voltage circuit output voltage at equal intervals multiple times within the current control cycle is used as the average output voltage of the base voltage circuit. The maximum voltage value output by the peak hold circuit in the detection pulse loop within the current pulse period is taken as the peak voltage output by the pulse loop during the peak hold phase.

4. The high-frequency power supply dual-group coordinated output control method according to claim 3, characterized in that, The calculation steps for the specified ratio include: Under the calibration conditions of introducing dry flue gas into the electrostatic precipitator and no flashover of the electric field, the waveform of the secondary current falling segment of multiple pulse cycles was continuously collected. For the secondary current decline segment of each pulse cycle, the current decay rate is calculated point by point starting from the pulse turn-off time. The ratio of the current value when the current decay rate decreases from the maximum value to one-tenth of the maximum value to the current value at the turn-off time is used as a single-cycle proportional sample. Sort the single-cycle proportion samples obtained from all pulse cycles according to their numerical values, remove the extreme samples at both ends, calculate the arithmetic mean of the remaining samples, and use the arithmetic mean as the specified proportion.

5. The high-frequency power supply dual-group coordinated output control method according to claim 1, characterized in that, The base voltage is corrected based on the steepness of the secondary current rise to obtain the dynamic base voltage. Simultaneously, the calibrated fallback time is acquired, and the peak compensation coefficient is determined by combining this fallback time with the current rise time. This includes: Under calibration conditions, the calibration rise edge steepness is obtained using the same method as that used to determine the rise edge steepness of the secondary current. Based on the relative increase in the rise edge steepness of the secondary current relative to the calibrated rise edge steepness, a steepness suppression component is determined; The fallback suppression component is determined based on the relative extension of the current fallback time relative to the calibrated fallback time; The ratio of the peak voltage output by the pulse circuit during the peak holding phase to the average output voltage of the base voltage circuit is obtained as the dynamic energy ratio; The pressure drop is determined based on the steepness suppression component, the fallback suppression component, and the dynamic energy ratio, wherein the pressure drop increases with the increase of the steepness suppression component, the fallback suppression component, and the dynamic energy ratio; Subtracting the voltage drop from the given base voltage value yields the dynamic base voltage. Divide the calibrated fallback time by the current fallback time, and use the resulting value as the peak compensation coefficient.

6. The high-frequency power supply dual-group coordinated output control method according to claim 1, characterized in that, Based on the rising edge steepness of the secondary current and the falling time of the current, an electric field transient response index is generated, including: Under the calibration conditions, obtain the calibration rise steepness and calibration fall time; The relative increment of steepness is determined based on the relative increase of the secondary current rising edge steepness relative to the calibrated rising edge steepness; The relative increment of the fall-off time is determined based on the relative extension of the current fall-off time relative to the calibrated fall-off time; The ratio of the peak voltage output by the pulse circuit during the peak holding phase to the average output voltage of the base voltage circuit is obtained as the dynamic energy ratio; Based on the relative steepness increment and the dynamic energy ratio, a steepness contribution term is determined, wherein the steepness contribution term is proportional to both the relative steepness increment and the dynamic energy ratio; Based on the relative increment of fallback time and the dynamic energy ratio, a fallback contribution term is determined, wherein the fallback contribution term is proportional to both the relative increment of fallback time and the dynamic energy ratio; The steepness contribution term and the fallback contribution term are combined to obtain the electric field transient response index.

7. The high-frequency power supply dual-group coordinated output control method according to claim 1, characterized in that, The training steps for the electric field transient response compensation model include: Multiple training samples are extracted from historical operating data. For each training sample, the corresponding electric field transient response index is obtained, and the corresponding ideal peak voltage secondary correction factor is determined. The ideal peak voltage secondary correction factor is calculated by back-calculating the pulse peak voltage given value and the required dynamic peak voltage when the electric field is in the ideal response state. The electric field transient response index is used as the sample input feature, and the ideal peak voltage quadratic correction factor is used as the sample output label to form a training dataset; A transient response compensation model for the initial electric field is constructed based on machine learning. The model parameters are optimized by taking the electric field transient response exponent in the training dataset as the model input, the corresponding ideal peak voltage quadratic correction factor as the supervision signal, and minimizing the error between the peak voltage quadratic correction factor output by the model and the supervision signal. Training stops when the verification error converges to the target range, resulting in a trained electric field transient response compensation model.

8. The high-frequency power supply dual-group coordinated output control method according to claim 1, characterized in that, Using the peak compensation coefficient and the peak voltage quadratic correction factor, a composite correction coefficient is formed. Then, the pulse peak voltage setpoint is adjusted according to the composite correction coefficient to obtain the dynamic peak voltage, including: Obtain the peak compensation coefficient and the peak voltage quadratic correction factor; The ratio of the peak voltage output by the pulse circuit during the peak holding phase to the average output voltage of the base voltage circuit is obtained as the dynamic energy ratio; Based on the dynamic energy ratio, a first fusion weight and a second fusion weight are determined, wherein the first fusion weight is positively correlated with the dynamic energy ratio, the second fusion weight is negatively correlated with the dynamic energy ratio, and the sum of the first fusion weight and the second fusion weight is 1. The peak voltage secondary correction factor is weighted using the first fusion weight, and the peak compensation coefficient is weighted using the second fusion weight. The weighted results are then added together to obtain the composite correction coefficient. The dynamic peak voltage is obtained by correcting the given value of the pulse peak voltage based on the composite correction coefficient, including: When the composite correction coefficient is less than 1, the pulse peak voltage setpoint is multiplied by the composite correction coefficient, and the result is used as the dynamic peak voltage. When the composite correction coefficient is greater than or equal to 1, the given value of the pulse peak voltage is directly used as the dynamic peak voltage.

9. The high-frequency power supply dual-group coordinated output control method according to claim 1, characterized in that, Based on the dynamic base voltage and the dynamic peak voltage, the pulse width compression coefficient is calculated, and the current pulse width is reduced in real time to obtain the limited pulse width, including: Based on the dynamic base voltage and the dynamic peak voltage, a base compression coefficient is determined, wherein the base compression coefficient increases with the increase of the dynamic base voltage and decreases with the increase of the dynamic peak voltage; Based on the steepness of the secondary current rise edge and the calibrated rise edge, a steepness correction factor is generated, and the basic compression factor is multiplied by the steepness correction factor to obtain the steepness correction compression factor; Based on the current fall-off time and the calibrated fall-off time, a fall-off correction factor is generated, and combined with the steepness correction compression coefficient, a composite compression coefficient is obtained. Obtain the dynamic energy ratio. If the dynamic energy ratio is greater than 1, divide the composite compression coefficient by the dynamic energy ratio to obtain the energy-corrected compression coefficient. When the dynamic energy ratio is not greater than 1, the composite compression coefficient is directly used as the energy-corrected compression coefficient. The smaller of the energy correction compression factor and the value 1 is selected as the final pulse width compression factor, and the current pulse width is multiplied by the final pulse width compression factor to obtain the amplitude-limited pulse width.

10. A high-frequency power supply dual-group coordinated output control system, characterized in that, For implementing the high-frequency power supply dual-group coordinated output control method according to any one of claims 1-9, the system comprises: The transient detection and mode triggering module is used to trigger a dual-group cooperative pulse power supply mode when the instantaneous change rate of the secondary current exceeds a preset fluctuation threshold. The feature parameter extraction module is used to obtain the average output voltage of the base voltage circuit and the peak voltage of the pulse circuit during the peak holding phase within the current control cycle, and to determine the dynamic feature parameters, wherein the dynamic feature parameters include at least the secondary current rise steepness and the current fall time. The base voltage correction module is used to correct the base voltage given value based on the steepness of the secondary current rise edge to obtain the dynamic base voltage, and at the same time obtain the calibration fall-off time, and determine the peak compensation coefficient in combination with the current fall-off time. The transient response evaluation module is used to generate an electric field transient response index based on the rise steepness of the secondary current and the fall time of the current, and input the electric field transient response index into a preset electric field transient response compensation model to obtain a peak voltage secondary correction factor. The peak voltage composite correction module is used to form a composite correction coefficient by using the peak compensation coefficient and the peak voltage second correction factor, and then adjust the pulse peak voltage setpoint according to the composite correction coefficient to obtain the dynamic peak voltage. The pulse width limiting calculation module is used to calculate the pulse width compression coefficient based on the dynamic base voltage and the dynamic peak voltage, and to reduce the current pulse width in real time to obtain the limited pulse width. The collaborative control command output module is used to output a collaborative control command carrying the dynamic base voltage, the dynamic peak voltage and the limiting pulse width, so as to drive the base voltage circuit to output the dynamic base voltage and drive the pulse circuit to output a high voltage pulse according to the dynamic peak voltage within the limiting pulse width.

Images