Flexible transformer control method and system considering harmonic suppression and dynamic response

Abstract

This invention proposes a flexible transformer control method and system that balances harmonic suppression and dynamic response. In steady state, the tracking error is written into a periodic memory, and a repetitive control compensation quantity is generated by reading at a delay of N points. This compensation quantity is then superimposed on the fast feedback control quantity for output. Upon detecting a dynamic event, the error address and repetitive control output of the steady-state period prior to the event are locked. The current transient error is replaced by a historical error of the same phase from the steady-state period and written into the memory. The reading delay is switched from N points to 2N points. During dynamic adjustment, the fast feedback control module uses the current error for adjustment, while the repetitive control module continues to output compensation based on the steady-state historical error. After the event ends and the error recovers, the delay is switched back to N points, and the repetitive control output is smoothly transitioned. This method separates dynamic and steady-state errors, preventing transient errors from being periodically stored and repeatedly injected, thus achieving a fast dynamic response while maintaining steady-state harmonic suppression capabilities.

Description

Technical Field

[0001] This invention relates to the field of flexible transformer technology, and in particular to a flexible transformer control method and system that takes into account both harmonic suppression and dynamic response. Background Technology

[0002] Flexible transformers primarily use power frequency transformers as the main power transmission channel, and actively regulate port voltage, current, and power through series converters, parallel converters, controllable windings, magneto-electric coupling regulating units, or combinations thereof. Parallel converters are commonly used for current tracking, active and reactive power regulation, and harmonic or imbalance compensation; series converters are commonly used for output voltage regulation, voltage sag compensation, and voltage waveform management. The quality of port control directly affects the absorption of new energy sources, operation of weak grids, and power supply to sensitive loads. Proportional-integral control, proportional-resonant control, or synchronous rotating coordinate system decoupling control have fast low-frequency dynamic responses, but usually require multiple resonant elements to handle multiple harmonics and periodic disturbances. Repetitive control utilizes the internal mode principle to establish high gain within one fundamental cycle, simultaneously suppressing multiple periodic harmonics, thus making it suitable for steady-state waveform control of port current and voltage in flexible transformers.

[0003] Conventional PI-RC or other fast feedback repetitive control combinations typically feed the same tracking error into both the fast feedback branch and the repetitive control branch. Step changes in reference power, current, or voltage, load switching, and grid voltage dips can generate large-amplitude aperiodic transient errors. After entering the repetitive control module, this error is delayed by one cycle and then acts on the control output again, potentially repeating in subsequent cycles, causing secondary distortion of port current or voltage and extending the settling time. The repetitive control gain exacerbates transient coupling, while reducing the gain weakens steady-state harmonic suppression. Furthermore, existing improved methods typically limit the amplitude of the error entering the repetitive control module. Limiting can reduce transient errors, but the threshold depends on the operating point and steady-state ripple. An excessively large threshold will still pollute the cycle memory, while an excessively small threshold will weaken steady-state compensation. Maintaining the repetitive control output or completely exiting repetitive control during disturbances may result in the loss of existing cycle memory and the generation of a new transient process upon re-entry, making it difficult to achieve a fast dynamic response while maintaining steady-state harmonic suppression capabilities. Summary of the Invention

[0004] In view of this, the present invention proposes a flexible transformer control method and system that takes into account both harmonic suppression and dynamic response. By dividing the tracking error during dynamic events into two independent paths, namely the current error for fast feedback adjustment and the steady-state historical error for repetitive control, and switching the reading delay of the repetitive control module from N points to 2N points after the event is triggered, the transient error is blocked from being periodically stored and repeatedly injected, thereby achieving fast dynamic response while maintaining steady-state harmonic suppression capability.

[0005] The technical solution of this invention is implemented as follows: In a first aspect, this invention provides a flexible transformer control method that balances harmonic suppression and dynamic response, comprising the following sub-steps: Obtain the tracking error between the reference value and the feedback value; In steady-state operation, the tracking error at the current moment is written into the periodic memory, and the historical error delayed by N sampling points relative to the writing moment is read from the periodic memory. The repetitive control compensation amount generated based on the historical error is superimposed with the fast feedback control amount generated based on the tracking error at the current moment to generate a control command output. The N is the number of fundamental periodic sampling points determined by the sampling frequency and the tracked fundamental frequency. When a dynamic trigger event is detected, the error storage address of the most recent steady-state power frequency cycle before the event occurs and the corresponding repetitive control compensation output value are locked and saved. The number of sampling points corresponding to the adjustment process that the fast feedback control quantity undergoes from the start of the response until the tracking error recovers to the preset range is K. d ; During the duration of the dynamic event, the tracking error at the current moment is blocked from being written to the periodic memory. Instead, the historical error at the same interval sampling point in the saved steady-state period is written to the periodic memory to replace the tracking error at the current moment. The generation reference of the repetitive control compensation is switched to read at a delay of 2N points. When the dynamic event ends and the tracking error meets the preset recovery conditions, the 2N-point delay reading is switched back to the N-point delay reading, and the repetitive control compensation amount before and after the switch is smoothly processed.

[0006] Based on the above technical solutions, preferably, the detection and triggering method of the dynamic triggering event includes: Preset entry and exit thresholds, wherein the entry threshold is greater than the exit threshold; When at least one of the following exceeds the corresponding preset threshold: the absolute value of tracking error, the rate of change of tracking error, the increment of reference quantity, the sudden change of feedback quantity, or the voltage drop of the grid, a dynamic triggering event is determined to have occurred. The dynamic event is considered to have ended when the absolute value of the tracking error is lower than the exit threshold and the rate of change of the tracking error is lower than the exit threshold.

[0007] Based on the above technical solutions, preferably, the periodic memory is a circular buffer with a capacity of not less than 2N sampling points, including a current write address pointer, a steady-state periodic error read address pointer, a delayed N-point read address pointer, and a delayed 2N-point read address pointer. Each address pointer advances in steps of the sampling period within the control period. The control period is consistent with the sampling period. The delayed N-point read and delayed 2N-point read are adjusted by the address offset of the circular buffer.

[0008] Based on the above technical solutions, preferably, the repetitive control compensation amount generated based on historical errors includes: Read the delay history error signal from N or 2N sampling points prior to the periodic memory, and filter the delay history error signal through filter Q(z) to obtain the filtered error signal; The current tracking error signal is summed with the historical error signal after Q(z) filtering, and then delayed by N sampling points and stored in the periodic memory; The sum of the error signals after delaying through N sampling points is filtered by filter S(z) to obtain the filtered signal. The filtered signal is subjected to phase compensation processing. The number of lead compensation points in the phase compensation processing is kc. The number of lead compensation points is determined according to the phase lag of the controlled object at the fundamental frequency and its integer multiples of the frequency. It is used to compensate for the phase lag of the controlled object in the main harmonic frequency band to obtain the phase-compensated signal. Multiply the phase-compensated signal by the repetitive control gain K. r This yields the repetitive control compensation amount.

[0009] Based on the above technical solutions, preferably, the first counter C is started after the dynamic triggering event occurs. M Timing is initiated, during which the periodic memory reads the delay history error signal from N sampling points prior for control; when the first counter reaches the preset first threshold M, the second counter C is started. L The isolation window is timed, and the duration of the isolation window is L sampling periods. At this time, the period memory reads the delay history error signal before 2N sampling points for control. The preset first threshold M is not less than the number of sampling points K corresponding to the dynamic adjustment process of the fast feedback control quantity. d And not greater than the number of sampling points N of the periodic delay processing minus the number of advance compensation points K of the phase compensation processing. c The difference obtained; The isolation duration L satisfies the following condition: the value obtained by subtracting the difference between the sampling point N of the periodic delay processing and the first threshold M from the isolation duration L is not less than the number of sampling points K corresponding to the dynamic adjustment process of the fast feedback control quantity. d And the number of sampling points N that is not greater than the delay amount of the periodic delay processing; After the isolation window timeout ends, if the tracking error meets the preset recovery conditions, the operation of switching back to delayed N-point reading will be performed.

[0010] Based on the above technical solutions, preferably, after the isolation window timer ends, the following further step is taken: If the tracking error meets the preset recovery conditions, the reading address will be switched from the 2N-point delay path back to the N-point delay path, and the repetitive control compensation amount before and after the switch will be continuously output through weight gradual change or slope limit tracking, and the event flag, the first counter and the second counter will be cleared. If the tracking error does not meet the preset recovery conditions, the isolation window is extended. If the extended isolation window exceeds the preset maximum allowable isolation time, the cycle memory is stopped from being updated, and only the fast feedback control module is kept running.

[0011] Based on the above technical solutions, preferably, when the method is applied to the current control loop of a flexible transformer parallel converter, the reference quantity is the current command value of the parallel converter, and the feedback quantity is the actual output current detection value of the parallel converter. The current command value is generated based on at least one of the active power control outer loop, reactive power control outer loop, DC bus voltage balance loop, harmonic current compensation loop, or unbalanced current compensation loop. When a power command step, grid voltage sag, or load change occurs, the fast feedback control quantity undertakes dynamic current regulation, while the repetitive control compensation quantity continues to perform harmonic compensation based on the steady-state periodic error before the event.

[0012] Based on the above technical solutions, preferably, the method is applied to the voltage control loop of a flexible transformer series converter, the reference quantity is the target injection voltage or target port voltage of the series converter, and the feedback quantity is the actual voltage detection value of the corresponding node; The target injection voltage or target port voltage includes at least one of the following: fundamental amplitude command, fundamental phase command, positive sequence component command, negative sequence component command, specific harmonic compensation command, or voltage sag compensation command. When a sudden change in port voltage command, input voltage disturbance, or load change occurs, the fast feedback control quantity undertakes the dynamic voltage adjustment, while the repetitive control compensation quantity continues to perform voltage harmonic compensation based on the steady-state periodic error before the event.

[0013] Secondly, the present invention also provides a flexible transformer control system that balances harmonic suppression and dynamic response, comprising: The error calculation module is used to calculate the difference between the reference value and the feedback value to obtain the tracking error. The fast feedback control module has its input connected to the output of the error calculation module and is used to generate a fast feedback control quantity based on the tracking error. The repetitive control module has its input connected to the output of the error calculation module and is used to generate repetitive control compensation based on the tracking error. The event detection module is used to monitor changes in the state of reference quantities, feedback quantities, tracking errors, or grid voltage, and outputs dynamic event flags when triggering conditions are met. The superposition output module has its first input terminal connected to the output terminal of the fast feedback control module and its second input terminal connected to the output terminal of the repetitive control module. It is used to superimpose the fast feedback control quantity and the repetitive control compensation quantity to output the final control command.

[0014] Based on the above technical solutions, preferably, it also includes a frequency tracking module, whose output is connected to the control terminal of the repetitive control module, for updating the N value online according to the change of the power grid frequency, and outputting the updated N value to the delay switching module to synchronously update the 2N value, and outputting it to the periodic memory to adjust its delay parameters, and synchronously updating the preset threshold M and isolation duration L according to the updated N value.

[0015] The flexible transformer control method and system of the present invention, which combines harmonic suppression and dynamic response, has the following advantages over the prior art: By dividing the tracking error during a dynamic event into two independent paths—the current error for fast feedback adjustment and the steady-state historical error for repetitive control—the reading delay of the repetitive control module is switched from N points to 2N points after the event is triggered, and the writing of transient error into the periodic memory is blocked. This allows the fast feedback control module to undertake dynamic adjustment, while the repetitive control module continues to output periodic compensation based on the historical error of the steady-state period before the event. This avoids the delayed re-injection of transient error after it is periodically stored, and also preserves the existing harmonic compensation memory of the repetitive control module. Thus, without reducing the steady-state repetitive control gain, both fast dynamic response and steady-state harmonic suppression capabilities are achieved. By switching the periodic delay term from N points to 2N points, the repetitive control module skips the transient error interval within the isolation window. Combined with the counter threshold and isolation duration associated with the fast feedback adjustment time, it ensures that the 2N delay path accurately spans the entire dynamic adjustment process. A circular cache with a capacity of not less than 2N points provides storage support for write isolation, stable periodic error replacement, and N / 2N read address switching. The three work together to ensure that the repetitive control module does not lose steady-state harmonic compensation memory during dynamic events, smoothly completes delay switching and recovery, and avoids secondary disturbances. By applying this method separately or simultaneously to the current loop of parallel converters and the voltage loop of series converters, the fast feedback control module is ensured to undertake dynamic adjustment during power steps, voltage sags, or load abrupt changes, while the repetitive control module continues to compensate for steady-state periodic errors. The method adapts to grid frequency shifts by updating N online through a frequency estimator. Combined with weak grid stabilization measures, the error decoupling triggered by events and the weak grid stabilization measures work in parallel without interference, thus enhancing the adaptability of the method under weak grid conditions. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a schematic diagram of the enhanced fast feedback repetitive control module based on dynamic steady-state error separation according to the present invention. Figure 3 This is a schematic diagram comparing the timing of tracking error invocation before and after a step event between conventional control and the control of this invention; Figure 4 This is the control flowchart of the present invention; Figure 5 This is a schematic diagram comparing the open-loop and disturbance propagation characteristics under the conditions of delay point N and 2N according to the present invention; Figure 6 This is a schematic diagram illustrating the stability verification of the present invention under different delay point numbers and filter parameter deviations. Figure 7 This is a comparison of experimental waveforms between a conventional PI-RC converter and a parallel converter controlled by the present invention under the power step operation condition of the present invention. Figure 8 A comparison of experimental waveforms between conventional PI-RC and parallel converters controlled by this invention under power step-down conditions; Figure 9 This is a comparison diagram of experimental waveforms of parallel converters under the three-phase voltage sag and dynamic reactive power support conditions of the present invention. Detailed Implementation

[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] like Figure 1-9 As shown, in a first aspect, the present invention provides a flexible transformer control method that balances harmonic suppression and dynamic response. The flexible transformer includes a main power transmission channel and at least one controllable adjustment channel, which may be composed of a parallel converter, a series converter, a controllable winding converter, or a combination thereof; the controlled object is the port current, the port voltage, or their coordinate transformation components.

[0020] It should be noted that the main power transmission channel of the flexible transformer undertakes most of the power transmission, while the controllable adjustment channel only undertakes small-capacity voltage or current regulation. Therefore, the control method of this invention does not change the design of the power frequency transformer of the main power transmission channel.

[0021] The control loop includes an error calculation module, a fast feedback control module, a repetitive control module, an event detection module, and a superposition output module. The fast feedback control module can be a PI controller, a PR controller, a state feedback controller, a predictive controller, or other controllers with the required dynamic bandwidth. The repetitive control module is used to generate high gain for the steady-state periodic error at the fundamental period and its integer multiples of the frequency.

[0022] It should be noted that the fast feedback control module has a high control bandwidth and can respond to error changes within several sampling periods, which is used to suppress low-frequency disturbances and provide a fast dynamic response; the repetitive control module is based on the internal model principle and provides theoretically infinite gain at the fundamental frequency and its integer multiples, which is used for steady-state harmonic suppression, but it is slower to respond to aperiodic disturbances; the event identifier is used to monitor reference quantities, feedback quantities and power grid status in real time, and outputs an event flag when the triggering conditions are met.

[0023] The method includes the following sub-steps: Obtain the tracking error between the reference value and the feedback value; In this embodiment, the control loop acquires a reference value and a feedback value in each sampling cycle. The reference value is the command value of the controlled target quantity, generated by the upper-level control strategy; the feedback value is the detected value of the actual output of the controlled object, acquired by a voltage sensor or current sensor and obtained after analog-to-digital conversion. Regarding the reference value... r ( k ) and feedback volume y ( k The difference is calculated to obtain the tracking error at the current time. e ( k ): e ( k )= r ( k )- y ( k ); In one embodiment, when the method is applied to the current control loop of a flexible transformer parallel converter, the reference quantity is the current command value of the parallel converter, and the feedback quantity is the actual output current detection value of the parallel converter. The current command value is generated based on at least one of the active power control outer loop, reactive power control outer loop, DC bus voltage balance loop, harmonic current compensation loop, or unbalanced current compensation loop. When a power command step, grid voltage sag, or load change occurs, the fast feedback control quantity undertakes dynamic current adjustment, and the repetitive control compensation quantity continues to perform harmonic compensation based on the steady-state periodic error before the event.

[0024] In another embodiment, the method is applied to the voltage control loop of a flexible transformer series converter. The reference quantity is the target injection voltage or target port voltage of the series converter, and the feedback quantity is the actual voltage detection value of the corresponding node. The target injection voltage or target port voltage includes at least one of the following: fundamental amplitude command, fundamental phase command, positive sequence component command, negative sequence component command, specific harmonic compensation command, or voltage sag compensation command. When a sudden change in port voltage command, input voltage disturbance, or load change occurs, the fast feedback control quantity undertakes the dynamic voltage adjustment, and the repetitive control compensation quantity continues to perform voltage harmonic compensation based on the steady-state periodic error before the event.

[0025] Parallel current loop embodiment: using a flexible transformer connected in parallel with a converter dq Taking the current control loop as an example, the three-phase output current is collected by a Hall effect current sensor. i a ( k ), i b ( k ), i c ( k After signal conditioning and analog-to-digital conversion, the phase of the grid voltage is obtained using a phase-locked loop. θ ( k The Park transformation converts the three-phase current to synchronous rotation. dq coordinate system, to obtain d Shaft current feedback i d ( k )and q Shaft current feedback i q ( k Simultaneously, the outer loop controlled by the DC bus voltage or the outer loop controlled by the active power is generated. d Shaft current reference i d1 ( k ), generated by the reactive power control outer loop q Shaft current reference i q1 ( k ); d Axis tracking error e d ( k )= i d1 ( k )- i d ( k ), q Axis tracking error e q (k )= i q1 ( k )- i q ( k ).

[0026] Series voltage loop example: Taking the voltage control loop of a flexible transformer series converter as an example, the target voltage command is formed by superimposing the fundamental voltage command, voltage sag compensation component, and harmonic compensation component; the three-phase voltage signal at the series injection point or low-voltage port is collected by the voltage sensor, and the voltage feedback quantity is obtained after conditioning and analog-to-digital conversion; the voltage tracking error is obtained by subtracting the voltage reference quantity from the voltage feedback quantity.

[0027] In steady-state operation, the tracking error at the current moment is written into the periodic memory, and the historical error delayed by N sampling points relative to the writing moment is read from the periodic memory. The repetitive control compensation amount generated based on the historical error is superimposed with the fast feedback control amount generated based on the tracking error at the current moment to generate a control command output. The N is the number of fundamental periodic sampling points determined by the sampling frequency and the tracked fundamental frequency. It should be noted that the above N Sampling frequency f s With the tracked fundamental frequency f The number of sampling points for the fundamental period determined in step 1 is determined as follows: N=round( f s / f 1); Here, `round()` represents rounding to the nearest integer; for example, when the sampling frequency... f s =10kHz, fundamental frequency of the power grid f When 1=50Hz, N =200; When the grid frequency deviates, N can be corrected online by a phase-locked loop, frequency estimator or period identifier.

[0028] Fast feedback control quantity u F ( k ) and repetitive control compensation amount u RC ( k The commands are added together at the superposition points to obtain the overall control command. u ( k The expression is: u ( k )= u F ( k ) +u RC ( k ); General control command u ( k After the limiting, anti-integral saturation and modulation stages, the driving signal of the power device is generated and applied to the controllable adjustment channel of the flexible transformer.

[0029] The periodic memory is a circular buffer with a capacity of not less than 2N sampling points, including a current write address pointer, a steady-state period error read address pointer, a delayed N-point read address pointer, and a delayed 2N-point read address pointer. Each address pointer advances in steps of the sampling period within the control period, which is consistent with the sampling period. The delayed N-point and delayed 2N-point reads are adjusted by the address offset of the circular buffer. During steady-state operation, the current write address pointer increments with each sampling period, and when it reaches the end of the buffer, it loops back to the starting address. The delayed N-point read address pointer always points to the position of the N sampling points before the current write address; the delayed 2N-point read address pointer always points to the position of the 2N sampling points before the current write address. After an event is triggered, the steady-state period error read address pointer points to the starting position of the most recent complete fundamental frequency period before the event.

[0030] The repetitive control compensation amount generated based on historical errors includes: reading the delayed historical error signal from N or 2N sampling points prior to the periodic memory; filtering the delayed historical error signal using filter Q(z) to obtain the filtered error signal; summing the current tracking error signal with the historical error signal filtered by Q(z) and delaying it by N sampling points, storing the summation in the periodic memory; filtering the summation of the error signal after the N sampling point delay using filter S(z) to obtain the filtered signal; and performing phase compensation processing on the filtered signal, wherein the number of lead compensation points in the phase compensation processing is... K c The number of advance compensation points is determined based on the phase lag of the controlled object at the fundamental frequency and its integer multiples, and is used to compensate for the phase lag of the controlled object in the main harmonic frequency band to obtain a phase-compensated signal. The phase-compensated signal is multiplied by the repetitive control gain to obtain the repetitive control compensation amount, enabling the repetitive control module to form a high gain at the fundamental frequency and its integer multiples. During steady-state operation, the delay amount of the periodic delay is set to N sampling points, and during the 2N-point delay reading period, the delay amount of the periodic delay is switched to 2N sampling points.

[0031] The discrete transfer function of the insertion-type repetitive control module corresponding to the above signal flow is: ; in, Q ( z) is a robust filter. S ( z () is a phase compensator. K r To control the gain repeatedly, k c This represents the number of phase lead compensation points.

[0032] When switching to a 2N-point delay for reading, the transfer function becomes: ; Specifically, Q ( z The filter is preferably a low-pass or zero-phase filter, which is used for filtering. The filter has a gain close to 1 in the harmonic band formed by the fundamental frequency and its integer multiples, and the gain is attenuated in the high-frequency band above the harmonic band. This is used to suppress high-frequency uncertainties and resonance peaks to obtain the filtering error signal. S ( z () is a phase compensator used to compensate for the amplitude and phase characteristics of the controlled object, wherein, S ( z ) =( z 2 +2 z +1) / (4 z ), number of phase lead compensation points k c Choose 3, and set the repetitive control gain K. r Take 1.

[0033] Single-phase or specific harmonic control: The control of the above-mentioned flexible transformer parallel converter in the dq coordinate system is based on the periodic delay of the fundamental positive sequence component; when it is necessary to control specific harmonics (such as the 5th and 7th harmonics), the coordinates can be transformed to the rotating coordinate system of the corresponding harmonic frequency, and the N value is adjusted accordingly (for example, the control period of the 5th harmonic is N / 5 sampling points). The write isolation and N / 2N switching mechanism of the present invention are also applicable.

[0034] When a dynamic trigger event is detected, the error storage address of the most recent steady-state power frequency cycle before the event occurs and the corresponding repetitive control compensation output value are locked and saved. The number of sampling points corresponding to the adjustment process that the fast feedback control quantity undergoes from the start of the response until the tracking error recovers to the preset range is K. d ; The detection and triggering method for dynamic trigger events includes: a preset entry threshold and an exit threshold, wherein the entry threshold is greater than the exit threshold to form a hysteresis loop and avoid frequent switching caused by noise; when at least one of the following exceeds the corresponding preset entry threshold: the absolute value of the tracking error, the rate of change of the tracking error, the increment of the reference quantity, the sudden change of the feedback quantity, or the voltage drop of the power grid, a dynamic trigger event is determined to have occurred; when the absolute value of the tracking error is lower than the exit threshold and the rate of change of the tracking error is lower than the exit threshold, the dynamic event is determined to have ended.

[0035] To avoid false triggering caused by noise or instantaneous disturbances, the event detection time confirmation mechanism requires that the event flag Flag=1 be set only after the trigger quantity continuously exceeds the entry threshold and reaches the preset confirmation time.

[0036] Freeze operation after event trigger: Simultaneously with Flag=1, the system immediately executes the following steps: save the starting address pointer of the most recent complete steady-state cycle before the event. This involves tracing back at least one fundamental cycle from the current moment, confirming that the absolute value of the error within that cycle is consistently below the exit threshold, and recording the starting address of the cycle as the steady-state cycle base address. The system also saves the current repetitive control compensation output value. u RC ( k This is used as a switching reference during subsequent non-disruptive recovery, and the current write pointer position is saved for address mapping during subsequent error replacement.

[0037] In this embodiment, the entry threshold is typically 5% to 20% of the rated current or rated voltage, and the exit threshold is 2% to 5% of the rated value. The specific values ​​need to be determined comprehensively based on the rated capacity of the controlled object, the steady-state ripple level, the sensor noise, and the control accuracy requirements. Alternatively, the error can be low-pass, moving average, or coordinate transformation can be performed before judgment to ensure that there is no false triggering in steady state and reliable triggering in dynamic events.

[0038] The absolute value criterion for tracking error is mainly used to detect sudden increases in error caused by step changes in the reference quantity or sudden changes in load; the rate of change criterion for tracking error is used to detect rapidly changing disturbances and can be triggered in advance before the error amplitude exceeds the threshold; the increment criterion for the reference quantity can directly respond to changes in power command and realize feedforward triggering; the sudden change criterion for the feedback quantity is used to detect external disturbances such as grid voltage dips or load switching; the grid voltage drop criterion is used to trigger independently when a voltage dip occurs; in practical applications, a logical "OR" combination of the above criteria can be used, and any criterion exceeding the threshold will trigger an event.

[0039] During the duration of the dynamic event, the current tracking error is blocked from being written to the periodic memory. Instead, the current tracking error is written to the periodic memory using the historical error of the same interval sampling points in the saved steady-state period. The generation reference of the repetitive control compensation is switched to read at a delay of 2N points.

[0040] During the period when Flag=1, the write selector does not write the current aperiodic transient error to the event coverage area. Specifically, the system maintains an address mapping table. When the write pointer points to a certain address, the write selector does not write the current error e(k). Instead, it starts from the steady-state periodic base address, reads the corresponding historical error with the same address offset, and writes the historical error to the position indicated by the current write pointer.

[0041] In conventional control, the transient error at the moment of an event is written to the read position before point N, causing the contaminated data to be called in each subsequent cycle; however, in this invention, the transient error at the moment of an event is replaced by the steady-state error, and the read address is switched to a 2N-point delay, so that the repetitive control module completely avoids the contaminated interval.

[0042] Switching the reading delay from N points to 2N points ensures that the repetitive control module reads the error data of the two fundamental frequency cycles before the event within L consecutive sampling points after the event occurs, rather than the data of the one fundamental frequency cycle before and after the event. Since the data of the two fundamental frequency cycles before the event does not contain transient errors, it ensures that the output of the repetitive control module is always based on pure steady-state harmonic components.

[0043] After the dynamic triggering event occurs, the first counter C is started. M Timing is initiated, during which the periodic memory reads the delay history error signal from N sampling points prior for control; when the first counter reaches the preset first threshold M, the second counter C is started. L The isolation window is timed, and its duration is L sampling periods. During this time, the period memory reads the delay history error signal from 2N sampling points prior to the isolation window for control. The preset first threshold M is not less than the number of sampling points K corresponding to the dynamic adjustment process of the fast feedback control quantity. d And not greater than the number of sampling points N of the periodic delay processing minus the number of advance compensation points K of the phase compensation processing. c The obtained difference; the isolation duration L satisfies: the value obtained by subtracting the difference between the sampling point N of the periodic delay processing and the first threshold M from the isolation duration L is not less than the number of sampling points corresponding to the dynamic adjustment process of the fast feedback control quantity. K d The sampling point N is not greater than the delay amount of the periodic delay processing; after the isolation window timeout ends, if the tracking error meets the preset recovery conditions, the operation of switching back to reading at delay N points is executed.

[0044] Let the number of sampling points corresponding to the dynamic adjustment process of the fast feedback control quantity be . K d , K d =int( τ F / T s ),in τ F To quickly provide feedback on the adjustment time of the control module, T s The sampling period.

[0045] After the dynamic trigger event occurs, the first counter C is started immediately. M Counting starts from zero; in C M Reaching the preset first threshold M Previously, the system only performed write isolation and error replacement, but did not perform address switching, and still maintained delayed N-point reads; during this stage, it waited for the fast feedback control module to complete the main dynamic adjustment process to avoid the switching time being too early and affecting the dynamic response of fast feedback.

[0046] When C M = M When the second counter is started, C L The isolation window is timed; simultaneously, the delay switch performs an operation to switch the read address from point N to 2N points. The isolation window continues. L One sampling period.

[0047] The preset first threshold M The value of is not less than K d and not greater than N - k c This inequality ensures that: M ≥ K d The isolation window should not be activated earlier than the moment when the fast feedback control module completes its main dynamic adjustments. M ≤ N - k c Ensure sufficient margin is maintained within the effective range of the N-point delay path.

[0048] The isolation duration L satisfy: K d ≤ L -( N - M )≤ N This inequality ensures that: L -( N - M )≥ Kd The effective delay of the 2N delay path is greater than the fast feedback adjustment time, ensuring that the switching recovery point falls after the fast feedback adjustment is completed. L -( N -M )≤ N The isolation window span should not exceed one fundamental frequency cycle to avoid introducing errors from other cycles.

[0049] The parameter design first ensures the stability of the fast feedback closed loop without the introduction of a repetitive control module, and ensures that its bandwidth covers the specified dynamic response range; then, based on the equivalent controlled object... G 0( z ) design Q ( z ), S ( z ), k c and K r This ensures that the convergence condition for repetitive control is met within the frequency range of interest: ; In the formula, ω is the discrete angular frequency. Since N or 2N only changes the periodic delay term, when the fast feedback closed loop is stable and the convergence condition is met, the delay switching does not change the basic boundary of this condition; it mainly changes the frequency selectivity and error call time during the event. In actual design, digital delay, modulation saturation, sampling noise, and device voltage and current constraints should also be checked.

[0050] For changes in filter inductors, capacitors, transformer leakage reactance, or grid impedance, stability margin can be improved through high-frequency attenuation of Q(z), robust compensation of S(z), passive damping, active damping, grid voltage feedforward with low-pass or resonant filtering, and online impedance identification. Event-triggered error decoupling does not change the action path of these stability measures and can be used in combination with them.

[0051] When the dynamic event ends and the tracking error meets the preset recovery conditions, the 2N-point delay reading is switched back to the N-point delay reading, and the repetitive control compensation amount before and after the switch is smoothly processed.

[0052] After the isolation window timeout ends, the process further includes: if the tracking error meets the preset recovery conditions, the read address is switched from the 2N-point delay path back to the N-point delay path, and the repetitive control compensation amount before and after the switch is continuously output through weighted gradual change or slope limiting tracking, and the event flag, the first counter, and the second counter are cleared; if the tracking error does not meet the preset recovery conditions, the isolation window is extended; if the extended isolation window exceeds the preset maximum allowable isolation time, the periodic memory is stopped from being updated, and only the fast feedback control module is kept running.

[0053] During the switching process, the specific implementation of the weight gradient method is as follows: during a preset recovery transition time... T transWithin this framework, the weight of the output from the path with a delay of N points is gradually increased while the weight of the output from the path with a delay of 2N points is simultaneously decreased, as expressed in the following formula: ; In the formula, u RC,N ( k ) represents the compensation amount calculated for a path with a delay of N points. u RC,2N ( k ) represents the compensation amount calculated for a path with a delay of 2N points. α ( k ) represents the weighting coefficient, which gradually increases from 0 to 1 during the recovery transition time. α ( k It can change linearly: α ( k )= t / T trans ,in t This represents the elapsed time since the start of the switch; it can also be varied using an S-curve to achieve a smoother transition. The recovery transition time is typically taken as... T trans = (0.5~2) fundamental frequency cycles, the specific value needs to balance smoothness and switching speed.

[0054] The slope-limited tracking method is an alternative to the weighted gradual change method: the difference in repeated control compensation Δ before and after the switch. u = u RC,N ( k )− u RC,2N ( k Slope limiting processing is performed to ensure that the output value of the repetitive control compensation gradually transitions at a set rate of change within the recovery transition time, avoiding output disturbances caused by step changes. The mathematical representation of slope limiting tracking is: ; In the formula, sat( x ,Δ U max ) indicates that x is restricted to [-Δ U max ,Δ U max Within the range, Δ U max This represents the maximum allowable change per sampling period.

[0055] After the switch is completed, the event flag (Flag) and the first counter (C) are cleared. M Second counter C L .

[0056] If the tracking error does not meet the preset recovery conditions, the isolation window is extended, and the process returns to continue determining whether the tracking error meets the recovery conditions; if the extended isolation window exceeds the preset maximum allowable isolation time... L max If the error occurs, timeout protection is triggered, stopping the updating of the cycle memory and retaining only the fast feedback control module to maintain safe system operation. Stopping the updating of the cycle memory means that the repetitive control compensation is frozen to the current value or gradually decays to zero, preventing the repetitive control module from outputting outdated data and causing a decline in control quality if it cannot recover for a long time. In timeout protection mode, the system continuously monitors and tracks the error. Once the error decreases below the recovery condition, the repetitive control module can be reactivated according to the normal recovery procedure.

[0057] In another embodiment, in a flexible transformer with both series and parallel converters, the series voltage loop and the parallel current loop can be configured using the method of the present invention respectively. During normal steady-state operation, both loops perform steady-state harmonic suppression with an N-point delay. When the system-level power command changes, the two loops can share the global event flag generated by the power command change. The parallel current loop first undertakes dynamic power regulation, while the series voltage loop maintains port voltage stability. When the input voltage is disturbed, the series voltage loop prioritizes establishing a compensation voltage, and the event flag is triggered by the grid voltage drop. The parallel current loop provides support based on the DC bus energy and device capacity. Both loops independently perform write isolation and delay switching operations without interfering with each other. To avoid the two regulation channels reaching the limit simultaneously, dynamic regulation responsibilities can be allocated according to port voltage safety, parallel current capacity, DC bus energy, and power quality targets, and independent limiting is implemented for repetitive control outputs.

[0058] The invention is described below with reference to the accompanying drawings and experiments. Three embodiments are provided to illustrate controller and memory management, dynamic and robustness verification, and series voltage loop and engineering deployment, respectively; the technical features of each embodiment can be combined where there is no conflict.

[0059] Example 1: Parallel Current Loop, Event Triggering, and Periodic Storage Management This embodiment uses a PI controller and an insertion-type repetitive control module in the dq current loop of a flexible transformer parallel converter. The sampling frequency is... f s The fundamental frequency is f 1. Number of sampling points for the fundamental frequency period N =round( f s / f 1) The length of the circular cache is not less than 2. N。

[0060] During steady-state operation, the tracking error at the current moment is recorded in each sampling period. e (k Write to the circular cache, and the repetition control module reads from the cache with a delay relative to the write time. N The historical error of each sampling point, after Q ( z ) Filtering, S ( z Phase compensation, multiplied by gain K r The repetitive control compensation quantity is generated by superimposing the delayed feedback quantity, and then superimposed on the PI controller output to generate the control command.

[0061] When the absolute value of the tracking error is | e ( k If the power reference increment exceeds the corresponding entry threshold, a dynamic trigger event is determined to have occurred, and the event flag is set. Flag =1, and save the starting address of the most recent steady-state cycle before the event, the current repetitive control compensation output value, and the current write pointer position. After the event flag is set, the current error is still sent to the PI controller for dynamic adjustment, but the value written to the loop buffer is no longer the current error, but the steady-state cycle error read from the saved steady-state cycle starting address with the same address offset.

[0062] After the event flag is set, the first counter C M Start timing from zero; in C M Reaching the preset threshold M Previously, the system only performed write isolation and error replacement, without performing delayed switching, and the repeat control module still used... N The historical error is read with a delay, and the PI controller is waited for to complete the main dynamic adjustment.

[0063] C M Start C after reaching M L Within L sampling points, the delay is switched from N to 2N; Figure 3 This indicates that a regular N-point delay will call data containing step error in the first and subsequent cycles after the event; the 2N path crosses the corresponding transient error interval; Figure 4 The event flags, two counters, delay selection, and recovery process are given.

[0064] If both |e(k)| and |e(k)-e(k-1)| are below the exit threshold at the end of the isolation period, the read address will be gradually restored to an N-point delayed address with a weighted transition of one to several sampling points. That is, within the preset recovery transition time, the proportion of the output of the delayed N-point path will be gradually increased according to the weight coefficient, while the proportion of the output of the delayed 2N-point path will be gradually decreased until it is completely switched to the delayed N-point path output, and the Flag and C will be cleared. M and C LIf the error is still large, then extend L; if the maximum isolation time is exceeded, then stop updating the repeating memory and only retain PI control and device protection.

[0065] Example 2: Power Step, Voltage Sag and Robustness Verification This embodiment experimentally verifies the practical effect of the invention in the current loop of a parallel converter. The experimental platform changes the d-axis current reference value of the parallel converter, causing the corresponding output active power to jump from 0.466kW to 1.400kW, and then back from 1.400kW to 0.466kW. Three-phase port voltage, parallel converter output current, d-axis tracking error, and d-axis repetitive control output are collected. The rated capacity and power values ​​of the device are for verification purposes only and do not limit the application capacity of the invention.

[0066] During power step increases, conventional PI-RC systems write a large step tracking error into the repetitive control module, resulting in visible current distortion in subsequent cycles, lasting approximately 85ms. With this invention, the transient error at the event moment is replaced by the stable cycle error before the event, with a 2N delay. The main current adjustment process takes approximately 4ms, and no significant repetitive distortion occurs in subsequent cycles. (See...) Figure 7 .

[0067] When the power step decreases, the tracking error direction reverses; conventional PI-RC will still repeat the error in subsequent cycles; with the present invention, the repetitive control module continues to process the steady-state periodic error, the PI controller completes rapid current reduction, and the output current transitions smoothly, see... Figure 8 The results show that this method is effective for both forward and reverse reference mutations.

[0068] A programmable AC power supply generates a 30% three-phase voltage sag, altering the reactive power reference of the parallel converter during the sag. Both the voltage sag and forced reactive power injection simultaneously produce a significant aperiodic current tracking error. Conventional PI-RC systems would delay and repeat this error, causing current distortion over multiple cycles.

[0069] With this invention, the voltage sag and tracking error jointly trigger the event flag; the PI current controller completes the dynamic reactive current establishment, and the repetitive control module does not write the non-periodic error at the sag initiation time. Figure 9 The experimental waveforms shown indicate that the converter can maintain a relatively smooth three-phase output current and perform reactive power support during disturbances.

[0070] The stability trajectories were plotted for the N and 2N delay states, respectively, by verifying the LC filter parameters of the parallel converter with a relative nominal value variation of ±20%. Figure 6 The trajectories shown are all within the specified stable region, indicating that the delayed switching did not disrupt stability under these implementation parameters. The parameter deviation range and stable region should be re-verified in conjunction with the specific flexible transformer capacity, filter, and sampling delay.

[0071] In weak grid environments, grid voltage feedforward low-pass or resonant filtering, virtual impedance, active damping or online impedance identification can be added, and the high-frequency gain of repetitive control can be reduced. The event triggering and error separation of this invention are located in the error call path, which can reduce the coupling between PI and repetitive control in dynamic events while retaining the above-mentioned stability measures.

[0072] Example 3: Coordination and Engineering Deployment of Series Voltage Loop and Dual Control Loop The controlled object in Example 1 is replaced with the injection voltage or low-voltage port voltage of a flexible transformer series converter. The voltage reference is formed by superimposing the rated voltage, voltage regulation command, voltage phase command, and power quality compensation, and the feedback quantity is the sampled voltage at the series injection point or port.

[0073] When a large voltage error is caused by an input voltage sag, a sudden change in port voltage command, or load switching, the event identifier triggers dynamic steady-state separation. The fast feedback voltage controller quickly establishes the required injected voltage, and the repetitive control module uses stable periodic errors to maintain harmonic compensation and restores the N-point delay after the port voltage stabilizes. This avoids the transient voltage error being repeatedly injected into the series voltage after one cycle.

[0074] In a flexible transformer with both series and parallel converters, the present invention is configured for the series-side voltage loop and the parallel-side current loop. Under normal conditions, both loops perform steady-state harmonic suppression with an N-point delay; when the system-level power command changes, they share an event flag, and the parallel current loop first undertakes dynamic power adjustment; when the input voltage is disturbed, the series voltage loop prioritizes establishing a compensation voltage, while the parallel current loop provides support based on the DC bus energy and device capacity.

[0075] The control software can be deployed as an independent functional module in the flexible transformer controller, or it can be integrated into the existing PI-RC control program through firmware upgrades. The main additions include event recognition, at least a two-cycle error buffer, address switching, and recovery logic, without requiring changes to the main power devices. Applicable scenarios include active distribution networks, microgrids, new energy and energy storage integration, industrial power quality management, electric vehicle charging facilities, rail transit, data centers, hospitals, and other sensitive load power supply.

[0076] With the increase in distributed renewable energy, DC loads, and high-reliability power consumption scenarios, flexible transformers need to simultaneously undertake voltage regulation, power regulation, and harmonic compensation. This invention achieves dynamic performance upgrades using control software and limited storage resources. It has an engineering foundation for reuse across device capacities, converter topologies, and application scenarios. It can be used for building new flexible transformers and is also suitable for the control retrofitting of existing power electronic transformers and power quality devices.

[0077] This method only adds event judgment, dual-cycle buffering and address switching, and can be implemented by DSP, FPGA or existing controller firmware without changing the main power device. It is suitable for newly built flexible transformers, as well as as a software upgrade of existing PI-RC devices, and can be ported to dynamic voltage restorers, unified power quality regulators, energy storage and new energy grid-connected converters, targeting distribution networks, industrial production lines, data centers, charging facilities, rail transit and new energy power stations.

[0078] Other implementation methods The fast feedback control module can employ PI, PR, state feedback, or predictive control; the repetitive control module can employ an insertion type, parallel type, odd harmonic, or multi-rate structure. As long as non-periodic errors are prevented from being written to the periodic memory during dynamic events, and the stable periodic error is used to replace and switch the N / 2N read address, the invention can be implemented accordingly.

[0079] The event criteria, M, and L can be fixed or adjusted online based on the error peak value, frequency, and operating status. When consecutive events occur, the most recently confirmed stable cycle is cached and the isolation window is recalculated.

[0080] Delayed switching can be used for only one phase, one coordinate axis, or a selected sequence component; event criteria, M, and L can be set fixedly or adjusted online based on error peak value, controlled object bandwidth, frequency deviation, and operating status. When multiple dynamic events occur consecutively, the isolation window can be recalculated and the most recent stable cycle can be retained in a cache.

[0081] Secondly, the present invention also provides a flexible transformer control system that balances harmonic suppression and dynamic response, characterized in that it includes: The error calculation module is used to calculate the difference between the reference value and the feedback value to obtain the tracking error. The fast feedback control module has its input connected to the output of the error calculation module and is used to generate a fast feedback control quantity based on the tracking error. The repetitive control module has its input connected to the output of the error calculation module and is used to generate repetitive control compensation based on the tracking error. The event detection module is used to monitor changes in the state of reference quantities, feedback quantities, tracking errors, or grid voltage, and outputs dynamic event flags when triggering conditions are met. The superposition output module has its first input terminal connected to the output terminal of the fast feedback control module and its second input terminal connected to the output terminal of the repetitive control module. It is used to superimpose the fast feedback control quantity and the repetitive control compensation quantity to output the final control command.

[0082] In a preferred embodiment, the system further includes a frequency tracking module. The input of the frequency tracking module is connected to a grid voltage sampling signal, used to obtain the real-time grid frequency via a phase-locked loop or a frequency estimator. The output of the frequency tracking module is connected to the control terminal of the repetitive control module, used to update the N value online according to grid frequency changes, output the updated N value to a delay switching module to synchronously update the 2N value, and output it to a periodic memory to adjust its delay parameters. The preset threshold M and isolation duration L are synchronously updated according to the updated N value. To prevent control output jitter caused by frequent changes in the N value during frequency tracking, the frequency tracking module has a dead-time comparison unit that freezes the N value update when the frequency change is less than a set dead time.

[0083] It should be noted that this system corresponds to the aforementioned flexible transformer control method that balances harmonic suppression and dynamic response. All implementation methods in the above method embodiments are applicable to the embodiments of this system and can achieve the same technical effect.

[0084] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0085] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system and modules described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0086] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0087] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0088] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0089] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0090] Furthermore, it should be noted that in the system and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent solutions of the present invention. Moreover, the steps performing the above series of processes can naturally be executed in the order described, but are not necessarily required to be executed in chronological order; some steps can be executed in parallel or independently of each other. Those skilled in the art will understand that all or any step or component of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or network of computing devices, in hardware, firmware, software, or a combination thereof. This is something that those skilled in the art can achieve by using their basic programming skills after reading the description of the present invention.

[0091] Therefore, the object of the present invention can also be achieved by running a program or a set of programs on any computing system. The computing system can be a known general-purpose system. Therefore, the object of the present invention can also be achieved simply by providing a program product containing program code implementing the method or apparatus. That is, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any known storage medium or any storage medium developed in the future. It should also be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent to the present invention. Furthermore, the steps performing the above series of processes can naturally be performed in the order described, but are not necessarily required to be performed in chronological order. Some steps can be performed in parallel or independently of each other.

[0092] 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.

Claims

1. A flexible transformer control method that balances harmonic suppression and dynamic response, characterized in that, Includes the following sub-steps: Obtain the tracking error between the reference value and the feedback value; In steady-state operation, the tracking error at the current moment is written into the periodic memory, and the historical error delayed by N sampling points relative to the writing moment is read from the periodic memory. The repetitive control compensation amount generated based on the historical error is superimposed with the fast feedback control amount generated based on the tracking error at the current moment to generate a control command output. The N is the number of fundamental periodic sampling points determined by the sampling frequency and the tracked fundamental frequency. When a dynamic trigger event is detected, the error storage address of the most recent steady-state power frequency cycle before the event occurs and the corresponding repetitive control compensation output value are locked and saved. The number of sampling points corresponding to the adjustment process that the fast feedback control quantity undergoes from the start of the response until the tracking error recovers to the preset range is: K d ; During the duration of the dynamic event, the tracking error at the current moment is blocked from being written to the periodic memory. Instead, the historical error at the same interval sampling point in the saved steady-state period is written to the periodic memory to replace the tracking error at the current moment. The generation reference of the repetitive control compensation is switched to read at a delay of 2N points. When the dynamic event ends and the tracking error meets the preset recovery conditions, the 2N-point delayed reading is switched back to the N-point delayed reading, and the repetitive control compensation amount before and after the switch is smoothly processed.

2. The flexible transformer control method that balances harmonic suppression and dynamic response as described in claim 1, characterized in that, The detection and triggering methods for the dynamic triggering event include: Preset entry and exit thresholds, wherein the entry threshold is greater than the exit threshold; When at least one of the following exceeds the corresponding preset threshold: the absolute value of tracking error, the rate of change of tracking error, the increment of reference quantity, the sudden change of feedback quantity, or the voltage drop of the grid, a dynamic triggering event is determined to have occurred. The dynamic event is considered to have ended when the absolute value of the tracking error is lower than the exit threshold and the rate of change of the tracking error is lower than the exit threshold.

3. The flexible transformer control method that balances harmonic suppression and dynamic response as described in claim 1, characterized in that, The cycle memory is a circular buffer with a capacity of not less than 2N sampling points, including a current write address pointer, a steady-state cycle error read address pointer, a delayed N-point read address pointer, and a delayed 2N-point read address pointer. Each address pointer advances in steps of the sampling period within the control period. The control period is consistent with the sampling period. The delayed N-point read and delayed 2N-point read are adjusted by the address offset of the circular buffer.

4. The flexible transformer control method that balances harmonic suppression and dynamic response as described in claim 1, characterized in that, The repetitive control compensation amount generated based on historical errors includes: Read the delay history error signal from N or 2N sampling points prior to the periodic memory, and filter the delay history error signal through filter Q(z) to obtain the filtered error signal; The current tracking error signal is summed with the historical error signal after Q(z) filtering, and then delayed by N sampling points and stored in the periodic memory; The sum of the error signals after delaying through N sampling points is filtered by filter S(z) to obtain the filtered signal. The filtered signal is then subjected to phase compensation processing, and the number of lead compensation points in the phase compensation processing is: K c The number of advance compensation points is determined based on the phase lag of the controlled object at the fundamental frequency and its integer multiples, and is used to compensate for the phase lag of the controlled object in the main harmonic frequency band to obtain the phase-compensated signal. Multiply the phase-compensated signal by the repetitive control gain K. r This yields the repetitive control compensation amount.

5. The flexible transformer control method that balances harmonic suppression and dynamic response as described in claim 4, characterized in that, After the dynamic triggering event occurs, the first counter C is started. M Timing is initiated, during which the periodic memory reads the delay history error signal from N sampling points prior for control; when the first counter reaches the preset first threshold M, the second counter C is started. L The isolation window is timed, and the duration of the isolation window is L sampling periods. At this time, the period memory reads the delay history error signal before 2N sampling points for control. The preset first threshold M is not less than the number of sampling points K corresponding to the dynamic adjustment process of the fast feedback control quantity. d And not greater than the number of sampling points N of the periodic delay processing minus the number of advance compensation points K of the phase compensation processing. c The difference obtained; The isolation duration L satisfies the following condition: the value obtained by subtracting the difference between the sampling point N of the periodic delay processing and the first threshold M from the isolation duration L is not less than the number of sampling points K corresponding to the dynamic adjustment process of the fast feedback control quantity. d And the number of sampling points N that is not greater than the delay amount of the periodic delay processing; After the isolation window timeout ends, if the tracking error meets the preset recovery conditions, the operation of switching back to delayed N-point reading will be performed.

6. The flexible transformer control method that balances harmonic suppression and dynamic response as described in claim 5, characterized in that, After the isolation window timer expires, the following also applies: If the tracking error meets the preset recovery conditions, the reading address will be switched from the 2N-point delay path back to the N-point delay path, and the repetitive control compensation amount before and after the switch will be continuously output through weight gradual change or slope limit tracking, and the event flag, the first counter and the second counter will be cleared. If the tracking error does not meet the preset recovery conditions, the isolation window is extended. If the extended isolation window exceeds the preset maximum allowable isolation time, the cycle memory is stopped from being updated, and only the fast feedback control module is kept running.

7. The flexible transformer control method that balances harmonic suppression and dynamic response as described in claim 1, characterized in that, When the method is applied to the current control loop of a flexible transformer parallel converter, the reference quantity is the current command value of the parallel converter, and the feedback quantity is the actual output current detection value of the parallel converter. The current command value is generated based on at least one of the active power control outer loop, reactive power control outer loop, DC bus voltage balance loop, harmonic current compensation loop, or unbalanced current compensation loop. When a power command step, grid voltage sag, or load change occurs, the fast feedback control quantity undertakes dynamic current regulation, while the repetitive control compensation quantity continues to perform harmonic compensation based on the steady-state periodic error before the event.

8. The flexible transformer control method that combines harmonic suppression and dynamic response as described in claim 1, characterized in that, The method is applied to the voltage control loop of a flexible transformer series converter, where the reference quantity is the target injection voltage or target port voltage of the series converter, and the feedback quantity is the actual voltage detection value of the corresponding node. The target injection voltage or target port voltage includes at least one of the following: fundamental amplitude command, fundamental phase command, positive sequence component command, negative sequence component command, specific harmonic compensation command, or voltage sag compensation command. When a sudden change in port voltage command, input voltage disturbance, or load change occurs, the fast feedback control quantity undertakes the dynamic voltage adjustment, while the repetitive control compensation quantity continues to perform voltage harmonic compensation based on the steady-state periodic error before the event.

9. A flexible transformer control system that balances harmonic suppression and dynamic response, characterized in that, include: The error calculation module is used to calculate the difference between the reference value and the feedback value to obtain the tracking error. The fast feedback control module has its input connected to the output of the error calculation module and is used to generate a fast feedback control quantity based on the tracking error. The repetitive control module has its input connected to the output of the error calculation module and is used to generate repetitive control compensation based on the tracking error. The event detection module is used to monitor changes in the state of reference quantities, feedback quantities, tracking errors, or grid voltage, and outputs dynamic event flags when triggering conditions are met. The superposition output module has its first input terminal connected to the output terminal of the fast feedback control module and its second input terminal connected to the output terminal of the repetitive control module. It is used to superimpose the fast feedback control quantity and the repetitive control compensation quantity to output the final control command.

10. The flexible transformer control system that combines harmonic suppression and dynamic response as described in claim 9, characterized in that, It also includes a frequency tracking module, whose output is connected to the control terminal of the repetitive control module. It is used to update the N value online according to the change of the power grid frequency, and output the updated N value to the delay switching module to update the 2N value synchronously, and output it to the period memory to adjust its delay parameters. It also updates the preset threshold M and isolation time L synchronously according to the updated N value.

Images