Hybrid transformer-based photovoltaic storage coordination and power quality comprehensive management method and system

Abstract

This invention discloses a method and system for integrated photovoltaic-storage synergy and power quality management based on a hybrid transformer. The system includes a hybrid transformer, a series converter, a parallel converter, a bidirectional DC / DC energy storage converter, a unidirectional DC / DC photovoltaic converter, and a central controller. The method monitors the grid voltage sag and its depth in real time. When the grid is normal, the parallel converter maintains DC bus voltage stability while also managing AC power quality. When there is an asymmetrical voltage sag and system capacity is limited, the series converter injects compensation voltage, and the parallel converter performs capacity concession, switches to DC voltage regulation support, and injects negative sequence compensation current. When the total system capacity is still limited, the energy storage releases active power, and the photovoltaic system exits maximum power point tracking mode to perform constant power load reduction operation. This invention can suppress common DC bus power fluctuations under extreme conditions, alleviate control conflicts, and improve system power supply reliability and power quality.

Description

Technical Field

[0001] This invention relates to the field of power systems and their automation technology, and in particular to a method and system for integrated power quality management based on hybrid transformers and photovoltaic-storage synergy. Background Technology

[0002] With the high proportion of distributed energy (such as photovoltaics) being integrated and the rapid growth of various AC / DC loads (such as electric vehicles, energy storage units, and sensitive manufacturing equipment), AC / DC hybrid microgrids have become an effective way to solve the problem of renewable energy consumption and improve the overall energy efficiency of the system. However, traditional distribution transformers only have basic voltage level transformation and electrical isolation functions and cannot flexibly regulate power flow. They are powerless to deal with the increasingly frequent power quality problems such as voltage sags, voltage rises, and three-phase imbalances in the power grid. According to statistics, voltage sag events account for the vast majority of power quality faults in the distribution network, which can easily lead to the shutdown of sensitive manufacturing equipment on the low-voltage side, loss of internal information, and significant economic losses.

[0003] Existing power quality management solutions typically involve installing separate photovoltaic inverters, active power filters (APFs), and dynamic voltage restorers (DVRs) within the distribution network—a "discrete" approach. This approach not only results in high overall equipment costs and large footprints, but also leads to fragmented control systems that struggle to achieve microgrid-level global coordination. For example, when a deep voltage sag occurs in the distribution network, conventional photovoltaic inverters are highly susceptible to disconnection due to undervoltage protection, causing the microgrid system to lose its critical active power source during the most critical transient periods when power support is most needed.

[0004] To overcome the drawbacks of discrete equipment, unified power quality conditioners (UPQC) or hybrid transformers (HDTs) integrating series and parallel converters have been proposed as key power control hubs in integrated power generation, grid, load, and storage systems. Compared to the expensive and weakly overload-resistant solid-state transformers (SSTs), hybrid transformers based on electromagnetic coupling integration combine the high reliability of traditional transformers with the flexible controllability of power electronic devices. However, when dealing with the extreme conditions of deep asymmetric voltage sags in the distribution network, existing multi-functional hybrid transformers still face serious bottlenecks of "capacity limitation and DC instability." Specifically, at the moment a sag occurs, the series converter needs to rapidly inject a high-amplitude compensation voltage into the grid, which will instantly draw a large amount of active power from the common DC bus. If the system's apparent capacity is limited or the output of photovoltaic and energy storage devices on the DC side is insufficient, it can easily lead to a severe imbalance in instantaneous power between the AC and DC sides. More seriously, the negative sequence component generated by the grid asymmetric voltage drop will interact with the current, generating severe second harmonic (100Hz) instantaneous power fluctuations on the DC bus, which can lead to DC bus voltage instability or even collapse, ultimately causing global instability of the AC / DC microgrid. In addition, when the system capacity is at its limit, most existing solutions use simple step-by-step logic judgment (i.e., black-and-white control command cut-off), lacking a model predictive scheduling (MPC) mechanism for global multi-objective optimization. This results in steady-state power oscillations and device-level scheduling conflicts when the photovoltaic system is forced to exit maximum power point tracking (MPPT).

[0005] Therefore, there is an urgent need for a comprehensive photovoltaic-storage synergy and power quality management method based on hybrid transformers that can deeply integrate energy management and power quality governance, effectively resolve multi-port collaborative control conflicts, eliminate DC bus second harmonic pulsation from a mathematical perspective, and completely cut off mode switching transient coupling distortion under conditions of limited total system capacity and extreme asymmetric sag. Summary of the Invention

[0006] This invention aims to at least partially solve one of the technical problems in related technologies. To this end, one objective of this invention is to propose a method for integrated power quality management based on hybrid transformer-based photovoltaic-storage synergy, which can suppress power fluctuations on the common DC bus under extreme operating conditions, alleviate control conflicts, and improve system power supply reliability and power quality.

[0007] Firstly, this invention proposes a method for integrated power quality management based on hybrid transformers and photovoltaic-storage synergy, applied to a three-port energy routing architecture centered on a common DC bus. This architecture includes a hybrid transformer, a series converter, a parallel converter, a bidirectional DC / DC energy storage converter, a unidirectional DC / DC photovoltaic converter, and a central controller. The method steps are as follows:

[0008] S1. Real-time acquisition of AC and DC side voltage and current to detect voltage dips and voltage drop depth in the power grid;

[0009] S2. When the grid voltage is normal, the parallel converter is controlled to prioritize maintaining the common DC bus voltage, and the remaining capacity is used to perform reactive power compensation and harmonic current suppression on the AC side.

[0010] S3. When an asymmetrical voltage dip is detected in the distribution network and the required compensation capacity of the system reaches the set threshold, the series converter is controlled to inject compensation voltage into the AC side, and the central controller forcibly reduces the reactive power compensation and harmonic suppression commands of the parallel converter on the AC side, so that the available capacity of the parallel converter can be concentrated to maintain the stability of the common DC bus voltage, and negative sequence compensation current is actively injected to suppress the power fluctuation of the common DC bus.

[0011] S4. After the parallel converter reduces the AC side compensation capacity, if the total capacity required to maintain the system power balance is still greater than the set threshold, the central controller will perform global power optimization allocation, control the bidirectional DC / DC energy storage converter to release active power, and control the unidirectional DC / DC photovoltaic converter to exit the maximum power point tracking control logic and switch to constant power load reduction operation.

[0012] Preferably, in step S1:

[0013] To address the asymmetrical voltage sag in the power grid, a method based on a dual-synchronous rotating coordinate system decoupled phase-locked loop (DDSRF-PLL) is employed to filter out negative-sequence and zero-sequence interference and extract the sag depth. The three-phase grid voltages are transformed to a stationary coordinate system using a Clarke transform and then synchronously fed into positive-sequence and negative-sequence decoupled Park transform modules. A cross-decoupling network matrix eliminates the influence of the second harmonic oscillation of the negative-sequence components, yielding pure positive and negative-sequence fundamental voltage components. The self-decoupling formula is as follows:

[0014] Positive sequence voltage decoupling equation:

[0015] ;

[0016] Negative sequence voltage decoupling equation:

[0017] ;

[0018] in, , Representing the original ascending order side respectively axis, Axis voltage components; , Representing the original negative order side respectively axis, Axis voltage components; Indicates the forward order after decoupling Axis voltage components; Indicates the forward order after decoupling Axis voltage components; Represents the negative order after decoupling Axis voltage components; Represents the negative order after decoupling Axis voltage components; This represents the grid phase angle estimated by DDSRF-PLL;

[0019] The decoupled forward sequence Axis voltage components The PI controller is closed-loop to zero to achieve no-delay phase-locked loop under asymmetric transients, and the extracted positive sequence is used. Axis voltage components Calculate the drop depth during the temporary descent.

[0020] Preferably, the method for controlling the injection of negative sequence compensation current into the parallel converter in step S3 to suppress power fluctuations on the common DC bus includes the following steps:

[0021] Constructing a DC component containing instantaneous active power DC component of reactive power and the second harmonic sine and cosine components of instantaneous active power , The fourth-order instantaneous power transfer matrix:

[0022] ;

[0023] in, , , , To detect the extracted grid voltage axis, Positive and negative order components of the axis; , , , For the output current of the parallel converter axis, The positive and negative sequence command components of the shaft; in the control objectives of the parallel converter, the second harmonic component command of the active power is forcibly set. , Substituting the target command into the inverse of the fourth-order instantaneous power transfer matrix, the absolute command of the specific negative-sequence compensation current that needs to be actively injected by the parallel converter is calculated in real time. , This is to counteract the 100Hz power pulsation caused by the asymmetric sag on the DC side.

[0024] Preferably, in step S3, to prevent output distortion caused by sudden changes in the AC side compensation capacity command of the parallel converter, its inner current loop adopts fuzzy proportional repetitive control (FPRC) to cut off dynamic coupling, and the open-loop transfer function of the composite controller is established as follows:

[0025] ;

[0026] in, This is the proportionality coefficient. This is the repeatability control factor. The number of sampling times per power frequency cycle. The transfer function of a robust filter used to improve the high-frequency stability of repetitive control systems. As the controlled object of the system, and These are the gain adjustment amounts for the corresponding dynamic output of the fuzzy logic. For discrete-time systems Complex variables in a field transfer function for A sampling period delay is used to establish the current tracking error. and error change rate The fuzzy logic rules are used as inputs. At the moment when the parallel converter performs AC-side compensation capacity adjustment or a transient change in the power grid, the fuzzy logic adaptively outputs a positive value. and negative By increasing the proportional gain and reducing the repetitive control gain, the proportional control dominates the response to cut off the dynamic coupling between the two channels; after the system tends to steady state, the parameters are adjusted in reverse to suppress steady-state tracking error.

[0027] Preferably, after adjusting the fuzzy inference output gain, the proportional control coefficient and repetitive control coefficient of the underlying controller are reconstructed in real time using the following update law:

[0028] ;

[0029] in, and These are the initial reference gains for the proportional gain and the repetitive control factor, respectively. and These are the quantization scaling factors corresponding to the defuzzification process. This is the sampling time sequence number.

[0030] Preferably, the method steps for the central controller to perform global power optimization allocation and control the unidirectional DC / DC photovoltaic converter to perform constant power load derating operation in step S4 are as follows:

[0031] Construct a multi-objective cost function in the central controller with the objective of minimizing multi-constraint errors. :

[0032] ;

[0033] in, , , For dynamic weighting coefficients, For the first Predicted value of common DC bus voltage at sampling time This is the reference value for the common DC bus voltage. For the first Predicted state of charge of energy storage at sampling time This is the minimum safe threshold for the state of charge of the energy storage unit. For the power regulation of the photovoltaic system, This refers to the sampling time sequence number; when the total system capacity is limited, the central controller solves the multi-objective cost function through rolling optimization. Calculate and output the optimal photovoltaic power descent command that satisfies DC-side absolute voltage regulation at this time. And energy storage charging and discharging power commands.

[0034] Preferably, in constructing a multi-objective cost function Previously, the central controller first established a controlled autoregressive prediction model for the common DC bus voltage based on the discretized sampling period:

[0035] ;

[0036] in, To control the sampling period of the system, The equivalent capacitance of the common DC bus. , , The first Real-time sampling power of photovoltaic, energy storage, and load at the sampling time. This represents the equivalent loss of the system.

[0037] Secondly, the present invention proposes a photovoltaic-storage synergy and power quality integrated management system based on a hybrid transformer, applying any of the above-mentioned methods for photovoltaic-storage synergy and power quality integrated management based on a hybrid transformer. The system includes: a hybrid transformer, a series converter, a parallel converter, a bidirectional DC / DC energy storage converter, a unidirectional DC / DC photovoltaic converter, and a central controller. The hybrid transformer includes a primary winding, a secondary main winding, and an auxiliary winding. The primary winding is connected to the distribution network, the secondary main winding is connected to a low-voltage AC bus, and the auxiliary winding is connected to a common DC bus through the series converter. The two ends of the parallel converter are respectively connected to the low-voltage AC bus and the common DC bus. The bidirectional DC / DC energy storage converter is connected between the energy storage unit and the common DC bus, and the unidirectional DC / DC photovoltaic converter is connected between the photovoltaic array and the common DC bus. The central controller is connected to the series converter, the parallel converter, the bidirectional DC / DC energy storage converter, and the unidirectional DC / DC photovoltaic converter.

[0038] The beneficial effects of this invention are:

[0039] (1) Based on the common DC bus, a comprehensive governance architecture for photovoltaic-storage synergy and power quality is constructed, which can realize unified scheduling of multiple devices and alleviate the problems of decentralized control and poor coordination among traditional discrete devices;

[0040] (2) Under the condition of unbalanced voltage sag, the coordinated control of series compensation, parallel capacity concession and negative sequence compensation current injection can effectively suppress the second harmonic power pulsation of DC bus and improve the system’s stable operation under extreme conditions.

[0041] (3) By dynamically adjusting the transient switching process through fuzzy proportional repetitive control, the resonance spike and current distortion can be reduced;

[0042] (4) By using model predictive control to optimize the scheduling of energy storage and photovoltaics, the photovoltaics can be smoothly deloaded and the energy storage can be reasonably output when the system capacity is limited, thereby improving the reliability of power supply and the overall power quality. Attached Figure Description

[0043] In the attached diagram:

[0044] Figure 1 The overall architecture and physical topology of the photovoltaic-storage synergy and power quality integrated management system based on hybrid transformers are shown below.

[0045] Figure 2 This is a flowchart of a method for integrated power quality management based on photovoltaic-storage synergy using hybrid transformers;

[0046] Figure 3 A waveform comparison of DC bus voltages;

[0047] Figure 4 This is a waveform diagram showing the voltage drop and compensation of the power grid. Detailed Implementation

[0048] Example 1:

[0049] Reference Figure 1 A photovoltaic-storage synergy and power quality management system based on a hybrid transformer includes: a hybrid transformer, a series converter, a parallel converter, a bidirectional DC / DC energy storage converter, a unidirectional DC / DC photovoltaic converter, and a central controller. The hybrid transformer includes a primary winding, a secondary main winding, and an auxiliary winding. The primary winding is connected to the distribution network, the secondary main winding is connected to a low-voltage AC bus, and the auxiliary winding is connected to a common DC bus through the series converter. The two ends of the parallel converter are connected to the low-voltage AC bus and the common DC bus, respectively. The bidirectional DC / DC energy storage converter is connected between the energy storage unit and the common DC bus, and the unidirectional DC / DC photovoltaic converter is connected between the photovoltaic array and the common DC bus. The central controller is connected to the series converter, the parallel converter, the bidirectional DC / DC energy storage converter, and the unidirectional DC / DC photovoltaic converter, respectively.

[0050] In this embodiment, the central controller is configured to: when an asymmetrical voltage sag is detected in the distribution network and the system capacity reaches a set threshold, issue a compensation voltage control command to the series converter and a capacity concession control command to the parallel converter to reduce its AC side reactive power compensation and harmonic mitigation commands, so that the parallel converter can switch to providing absolute voltage stabilization support for the common DC bus, and at the same time issue a negative sequence compensation current command to the parallel converter.

[0051] Example 2:

[0052] Reference Figure 2 The integrated power quality management method based on hybrid transformer-based photovoltaic-storage synergy constructs a three-port (grid-photovoltaic-storage) energy routing architecture with a common DC bus as the core, through the system built in Example 1. The central controller performs global energy millisecond-level scheduling based on the priority of "grid security > power quality > energy management". The method steps are as follows:

[0053] S1. Real-time acquisition of AC and DC side voltage and current to detect voltage dips and voltage drop depth in the power grid;

[0054] In this embodiment, accurately and without delay extracting the positive-sequence fundamental voltage amplitude during asymmetric sags is a prerequisite for initiating sag compensation. Traditional phase-locked loops (PLLs) are affected by negative-sequence components during asymmetric sags, resulting in second-harmonic oscillations. This embodiment employs a dual-synchronous rotating coordinate system decoupled phase-locked loop (DDSRF-PLL) technology to eliminate interference at its mathematical origin.

[0055] First, the grid voltage components in the stationary coordinate system are synchronously fed into the positive-sequence and negative-sequence decoupled Park transform modules. To filter out the second harmonic (100Hz) oscillations generated by the opposite-sequence components in the dq coordinate system to the greatest extent possible, this embodiment constructs a cross-decoupling network matrix, whose self-decoupling mathematical formula is as follows:

[0056] Positive sequence voltage decoupling equation:

[0057] ;

[0058] Negative sequence voltage decoupling equation:

[0059] ;

[0060] in, , Representing the original ascending order side respectively q-axis voltage components; , Representing the original negative order side respectively q-axis voltage components; Indicates the forward order after decoupling Axis voltage components; Indicates the forward order after decoupling Axis voltage components; Represents the negative order after decoupling Axis voltage components; Represents the negative order after decoupling Axis voltage components; This represents the grid phase angle estimated by DDSRF-PLL;

[0061] Through the aforementioned low-pass feedback cross-cancellation, the pure positive-sequence voltage fundamental component is extracted. and Then, the decoupled forward sequence will be... Axis voltage components By closing the loop with a PI controller to zero, phase-locked loop acquisition of the grid phase is achieved under asymmetrical sag. At the same time, the decoupling yields the positive order. shaft voltage It is directly used as the positive sequence fundamental amplitude value for the central controller to accurately calculate the grid sag depth and trigger the sag compensation action of the series converter at a speed of milliseconds.

[0062] S2. When the grid voltage is normal, the parallel converter is controlled to prioritize maintaining the common DC bus voltage, and the remaining capacity is used to perform reactive power compensation and harmonic current suppression on the AC side.

[0063] S3. When an asymmetrical voltage dip is detected in the distribution network and the required compensation capacity of the system reaches the set threshold, the series converter is controlled to inject compensation voltage into the AC side, and the central controller forcibly reduces the reactive power compensation and harmonic suppression commands of the parallel converter on the AC side, so that the available capacity of the parallel converter can be concentrated to maintain the stability of the common DC bus voltage, and negative sequence compensation current is actively injected to suppress the power fluctuation of the common DC bus.

[0064] In this embodiment: Under power quality priority mode, asymmetrical voltage drops in the power grid will generate severe negative sequence voltage. The interaction between the negative sequence voltage and the converter AC current will inevitably generate a 100Hz double-frequency active power pulsation on the common DC bus, which can easily cause DC-side instability and collapse.

[0065] To address this pain point, this embodiment implements negative-sequence compensation current injection control for the parallel converter. The specific mechanism is as follows: Based on instantaneous power theory, a DC component containing instantaneous active power is constructed. DC component of reactive power and the second harmonic sine and cosine components of instantaneous active power The fourth-order instantaneous power transfer matrix:

[0066] ;

[0067] in, , , , The extracted grid voltage is detected in step S1. axis, Positive and negative order components of the axis; , , , The d-axis represents the output current of the parallel converter. Axis positive and negative sequence command components; Since the core control objective is to eliminate 100Hz pulsation on the DC side, the control objective of the parallel converter forcibly sets the second harmonic component command of the active power. , Simultaneously, the required steady-state DC active power is set as follows: The target command is substituted into the inverse of the fourth-order instantaneous power transfer matrix for solution, and the absolute command of the specific negative sequence compensation current that needs to be actively injected by the parallel converter is calculated in real time. , The converter outputs the negative sequence current as instructed, which can perfectly offset the power fluctuations caused by the asymmetrical sag on the AC side, fundamentally ensuring the absolute voltage regulation of the common DC bus during the AC side compensation capacity period.

[0068] At the instant of switching between the AC side compensation capacity and the command change, in conventional proportional repetitive control (PRC), due to the inherent one power frequency cycle (20ms) data update delay in the repetitive control internal model, a strong dynamic conflict and error accumulation will occur between the high dynamic response of proportional control and the lag effect of repetitive control, which is very likely to excite the resonance spike of the AC side LCL filter.

[0069] Therefore, this embodiment introduces fuzzy proportional repetitive control (FPRC) in the inner current loop, and its composite open-loop transfer function can be expressed as:

[0070] ;

[0071] in, This is the proportionality coefficient. This is the repeatability control factor. The number of sampling times per power frequency cycle. This robust filter, used to improve the high-frequency stability of repetitive control systems, is specifically designed as a low-pass filter with a cutoff frequency lower than the system's Nyquist frequency, or a positive real constant slightly less than 1. It attenuates the system's high-frequency gain and shrinks the closed-loop poles of the control system to within the unit circle. As the controlled object of the system, and These are the gain adjustment amounts for the corresponding dynamic output of the fuzzy logic. For discrete-time systems Complex variables in a field transfer function for A sampling period delay is used to establish the current tracking error. and error change rate The fuzzy logic rules are input at the moment when the parallel converter performs AC-side compensation capacity or a transient change in the power grid (i.e., and When the value is large, the fuzzy logic adaptively outputs a positive direction. and negative By increasing the proportional gain and reducing the repetitive control gain, the proportional control dominates the response to cut off the dynamic coupling between the two channels, effectively eliminating transient resonance spikes and current distortion from a physical mechanism perspective; as the system approaches steady state (i.e., Approaching zero), fuzzy inference output is negative. and positive It automatically increases the weight of repetitive control and effectively eliminates periodic steady-state tracking errors by utilizing the high-gain characteristics of repetitive control.

[0072] After adjusting the fuzzy inference output gain, the proportional control coefficient and repetitive control coefficient of the underlying controller are reconstructed in real time using the following update law:

[0073] ;

[0074] in, and These are the initial reference gains for the proportional gain and the repetitive control factor, respectively. and These are the quantization scaling factors corresponding to the defuzzification process. This is the sampling time sequence number.

[0075] S4. After the parallel converter reduces the AC side compensation capacity, if the total capacity required to maintain the system power balance is still greater than the set threshold, the central controller will perform global power optimization allocation, control the bidirectional DC / DC energy storage converter to release active power, and control the unidirectional DC / DC photovoltaic converter to exit the maximum power point tracking control logic and switch to constant power load reduction operation.

[0076] In this embodiment: when the series converter compensates for a severe load drop, even if the parallel converter completely yields its reactive power capacity, the total system capacity may still exceed the limit. The simple "device removal" in traditional logic control can lead to system power oscillations. This embodiment abandons traditional load reduction logic and directly deploys multi-constraint model predictive control (MPC) in the central controller for global optimization.

[0077] Before constructing the multi-objective cost function, the central controller first establishes a controlled autoregressive prediction model for the DC bus voltage based on the discretized sampling period:

[0078] ;

[0079] in, For the first The predicted value of the common DC bus voltage at the sampling time. To control the sampling period of the system, Equivalent capacitance of the common DC bus , , The first Real-time sampling power of photovoltaic, energy storage, and load at the sampling time. This represents the equivalent system loss. Subsequently, the central controller constructs a multi-objective cost function aimed at minimizing the multi-constraint error. Its mathematical equation is:

[0080] ;

[0081] in, This is the DC bus voltage stabilization weighting coefficient. State of charge of energy storage battery ( Safety threshold weighting coefficient This is the photovoltaic power smoothness penalty coefficient. This is the reference value for the common DC bus voltage. For the first Predicted state of charge of energy storage at the sampling time. This is the minimum safe threshold for the state of charge of the energy storage unit. For the power regulation of the photovoltaic system, This refers to the sampling time sequence number; under extreme and constrained conditions, the MPC algorithm maintains... At reference value As the highest priority, through the objective cost function By performing rolling optimization to find the minimum value, the absolute value of the optimal photovoltaic power descent required to maintain the power balance of the system at this time can be directly calculated. And the optimal discharge command for the energy storage system. The unidirectional DC / DC photovoltaic converter receives this absolute command. Subsequently, it directly jumps out of the original Maximum Power Point Tracking (MPPT) operating mode and seamlessly switches to the constant power output mode, directly outputting the target power calculated by MPC. This method uses rigorous multi-objective matrix optimization to replace the traditional tentative step-size perturbation, effectively eliminating steady-state power oscillations during sudden changes in illumination or power limit command switching, and realizing precise dynamic control and optimal power decoupling of the source-grid-load-storage microgrid under extreme operating conditions.

[0082] like Figure 3 As shown, Figure 3 The diagram shows a comparison of DC bus voltage response waveforms under asymmetrical voltage sags provided in this embodiment. During an asymmetrical voltage sag in the distribution network (as shown in the diagram, from 0.15s to 0.4s), if a traditional control strategy is used (shown by the dashed line in the diagram), a severe 100Hz (second harmonic) instantaneous active power pulsation will occur on the common DC bus due to the interaction between the negative sequence component generated by the asymmetrical sag and the current. This poses a risk of DC-side instability or even collapse. However, by employing the comprehensive mitigation strategy of this invention (shown by the solid line in the diagram), a specific negative sequence compensation current absolute command is actively injected into the parallel converter. Within the same voltage sag range, the DC bus voltage is consistently maintained near the target reference value, effectively eliminating the 100Hz voltage pulsation. This demonstrates that this invention can fundamentally offset the power pulsation caused by asymmetrical sags on the DC side, ensuring the stable operation of the common DC bus during the AC-side compensation capacity execution of the parallel converter.

[0083] like Figure 4 As shown, Figure 4This diagram illustrates the grid voltage dip and compensation waveforms provided in this embodiment. The diagram contains three sets of dynamic waveforms from top to bottom, fully demonstrating the system's dynamic compensation process in response to unbalanced voltage sags: First, as shown in the grid-side voltage waveform (first set of dynamic waveforms), an unbalanced voltage sag occurs in the distribution network between 0.1s and 0.3s, with a severe drop in the voltage amplitude of phase A. Second, as shown in the series converter compensation voltage waveform (second set of dynamic waveforms), after detecting the aforementioned unbalanced voltage sag in the distribution network and the system capacity reaching a set threshold, the series converter responds rapidly, accurately injecting a high-amplitude compensation voltage into the AC side within the fault interval of 0.1s to 0.3s. Finally, as shown in the low-voltage AC bus voltage waveform (third set of dynamic waveforms), through the effective management of the series converter, the voltage on the load side (i.e., the low-voltage AC bus) maintains a standard three-phase sinusoidal waveform throughout the entire fault period (0.1s to 0.3s), and the three-phase voltage is completely restored to balance. This verifies that the present invention can effectively manage voltage faults on the grid side through series compensation under extreme operating conditions, thereby ensuring reliable power supply to sensitive equipment on the low-voltage side and the overall power quality of the system.

[0084] In summary, this invention achieves:

[0085] (1) Global Energy Dispatch and Capacity Concession Mechanism: This invention constructs a global energy dispatch and capacity forced concession mechanism based on "grid security > power quality > energy management". Under extreme conditions such as deep voltage sag in the distribution network, the parallel converters are forced to relinquish their available capacity and instead fully support the common DC bus voltage; at the same time, the bidirectional DC / DC energy storage converter is controlled to release its maximum active power. If the total system capacity is still limited, a model predictive control (MPC) with a multi-objective cost function including DC bus voltage deviation and energy storage state constraints is constructed for global optimization, and the absolute power command is directly output to control the photovoltaic system to exit MPPT and perform constant power smooth load reduction. This MPC-based multi-objective optimization algorithm realizes the smooth curtailment of photovoltaic power and the optimal power allocation of energy storage under capacity constraints, fundamentally solving the problem of multi-machine control conflict.

[0086] (2) Precise Detection and DC Fluctuation Suppression of Asymmetric Sags: In the condition monitoring step, this invention employs a method based on a dual synchronous rotating coordinate system decoupled phase-locked loop (DDSRF-PLL) to filter out negative-sequence and zero-sequence interference during asymmetric sags to the greatest extent possible from a mathematical perspective, achieving rapid phase-locking of the positive-sequence fundamental wave and precise extraction of sag depth. More importantly, to address DC bus instability caused by grid asymmetric sags, this invention introduces a negative-sequence current injection mechanism in the control of the parallel converter. By constructing a fourth-order instantaneous power transfer matrix, a specific negative-sequence compensation current command is calculated and forced to be injected into the parallel converter in real time, thereby mathematically offsetting the 100Hz second harmonic power pulsation caused by asymmetric sags on the DC side, ensuring the absolute stability of the DC bus.

[0087] (3) Transient distortion suppression and dynamic coupling cutoff: To address the dynamic lag and control coupling issues arising from conventional PRC control during mode switching or transient abrupt changes, this invention introduces a fuzzy proportional repetitive control (FPRC) strategy into the underlying current inner loop of the parallel converter. This strategy establishes fuzzy logic rules with current tracking error and its rate of change as input. At the moment of transient abrupt change, the fuzzy logic adaptively outputs a positive proportional gain increment and a negative repetitive control gain increment, with proportional control dominating the response to cut off the dynamic coupling between the two channels and eliminate transient resonance spikes; after the system tends to steady state, the parameters are adjusted in reverse, with repetitive control dominating to eliminate steady-state tracking errors.

Claims

1. A method for integrated power quality management based on photovoltaic-storage synergy using hybrid transformers, characterized in that: This method is applied to a three-port energy routing architecture centered on a common DC bus. The architecture includes a hybrid transformer, series converter, parallel converter, bidirectional DC / DC energy storage converter, unidirectional DC / DC photovoltaic converter, and a central controller. The steps are as follows: S1. Real-time acquisition of AC and DC side voltage and current to detect voltage dips and voltage drop depth in the power grid; S2. When the grid voltage is normal, the parallel converter is controlled to prioritize maintaining the common DC bus voltage, and the remaining capacity is used to perform reactive power compensation and harmonic current suppression on the AC side. S3. When an asymmetrical voltage dip is detected in the distribution network and the required compensation capacity of the system reaches the set threshold, the series converter is controlled to inject compensation voltage into the AC side, and the central controller forcibly reduces the reactive power compensation and harmonic suppression commands of the parallel converter on the AC side, so that the available capacity of the parallel converter can be concentrated to maintain the stability of the common DC bus voltage, and negative sequence compensation current is actively injected to suppress the power fluctuation of the common DC bus. S4. After the parallel converter reduces the AC side compensation capacity, if the total capacity required to maintain the system power balance is still greater than the set threshold, the central controller will perform global power optimization allocation, control the bidirectional DC / DC energy storage converter to release active power, and control the unidirectional DC / DC photovoltaic converter to exit the maximum power point tracking control logic and switch to constant power load reduction operation.

2. The method for integrated power quality management based on photovoltaic-storage synergy and power quality control using a hybrid transformer according to claim 1, characterized in that, In step S1: To address the asymmetrical voltage sag in the power grid, a method based on a dual-synchronous rotating coordinate system decoupled phase-locked loop (DDSRF-PLL) is employed to filter out negative-sequence and zero-sequence interference and extract the sag depth. The three-phase grid voltages are transformed to a stationary coordinate system using a Clarke transform and then synchronously fed into positive-sequence and negative-sequence decoupled Park transform modules. A cross-decoupling network matrix eliminates the influence of the second harmonic oscillation of the negative-sequence components, yielding pure positive and negative-sequence fundamental voltage components. The self-decoupling formula is as follows: Positive sequence voltage decoupling equation: ； Negative sequence voltage decoupling equation: ； in, , Representing the original ascending order side respectively axis, Axis voltage components; , Representing the original negative order side respectively axis, Axis voltage components; Indicates the forward order after decoupling Axis voltage components; Indicates the forward order after decoupling Axis voltage components; Represents the negative order after decoupling Axis voltage components; Represents the negative order after decoupling Axis voltage components; This represents the grid phase angle estimated by DDSRF-PLL; The decoupled forward sequence Axis voltage components The PI controller is closed-loop to zero to achieve no-delay phase-locked loop under asymmetric transients, and the extracted positive sequence is used. Axis voltage components Calculate the drop depth during the temporary descent.

3. The method for integrated power quality management based on photovoltaic-storage synergy and power quality control using a hybrid transformer according to claim 1, characterized in that, The steps of controlling the injection of negative sequence compensation current into the parallel converter in step S3 to suppress power fluctuations on the common DC bus are as follows: Constructing a DC component containing instantaneous active power DC component of reactive power and the second harmonic sine and cosine components of instantaneous active power , The fourth-order instantaneous power transfer matrix: ； in, , , , To detect the extracted grid voltage axis, Positive and negative order components of the axis; , , , For the output current of the parallel converter axis, The positive and negative sequence command components of the shaft; in the control objectives of the parallel converter, the second harmonic component command of the active power is forcibly set. , Substituting the target command into the inverse of the fourth-order instantaneous power transfer matrix, the absolute command of the specific negative-sequence compensation current that needs to be actively injected by the parallel converter is calculated in real time. , This is to counteract the 100Hz power pulsation caused by the asymmetric sag on the DC side.

4. The method for integrated power quality management based on photovoltaic-storage synergy and power quality control using a hybrid transformer according to claim 1, characterized in that, In step S3, to prevent output distortion caused by sudden changes in the AC side compensation capacity command of the parallel converter, its inner current loop uses fuzzy proportional repetitive control (FPRC) to cut off dynamic coupling, and the open-loop transfer function of the composite controller is established as follows: ； in, This is the proportionality coefficient. This is the repeatability control factor. The number of sampling times per power frequency cycle. The transfer function of a robust filter used to improve the high-frequency stability of repetitive control systems. As the controlled object of the system, and These are the gain adjustment amounts for the corresponding dynamic output of the fuzzy logic. For discrete-time systems Complex variables in a field transfer function For a delay of N sampling periods, establish a current tracking error... and error change rate The fuzzy logic rules are used as inputs. At the moment when the parallel converter performs AC-side compensation capacity adjustment or a transient change in the power grid, the fuzzy logic adaptively outputs a positive value. and negative By increasing the proportional gain and reducing the repetitive control gain, the proportional control dominates the response to cut off the dynamic coupling between the two channels; after the system tends to steady state, the parameters are adjusted in reverse to suppress steady-state tracking error.

5. The method for integrated power quality management based on photovoltaic-storage synergy and power quality control using a hybrid transformer according to claim 4, characterized in that, After adjusting the fuzzy inference output gain, the proportional control coefficient and repetitive control coefficient of the underlying controller are reconstructed in real time using the following update law: ； in, and These are the initial reference gains for the proportional gain and the repetitive control factor, respectively. and These are the quantization scaling factors corresponding to the defuzzification process. This is the sampling time sequence number.

6. The method for integrated power quality management based on photovoltaic-storage synergy and power quality control using a hybrid transformer according to claim 1, characterized in that, The steps for the central controller to perform global power optimization allocation and control the unidirectional DC / DC photovoltaic converter to perform constant power load derating operation in step S4 are as follows: Construct a multi-objective cost function in the central controller with the objective of minimizing multi-constraint errors. : ； in, , , For dynamic weighting coefficients, For the first The predicted value of the common DC bus voltage at the sampling time. This is the reference value for the common DC bus voltage. For the first Predicted state of charge of energy storage at the sampling time. This is the minimum safe threshold for the state of charge of the energy storage unit. For the power regulation of the photovoltaic system, This refers to the sampling time sequence number; when the total system capacity is limited, the central controller solves the multi-objective cost function through rolling optimization. Calculate and output the optimal photovoltaic power descent command that satisfies DC-side absolute voltage regulation at this time. And energy storage charging and discharging power commands.

7. The method for integrated power quality management based on photovoltaic-storage synergy and power quality control using a hybrid transformer according to claim 6, characterized in that, Constructing a multi-objective cost function Previously, the central controller first established a controlled autoregressive prediction model for the common DC bus voltage based on the discretized sampling period: ； in, To control the sampling period of the system, The equivalent capacitance of the common DC bus. , , The first Real-time sampling power of photovoltaic, energy storage, and load at the sampling time. This represents the equivalent loss of the system.

8. A photovoltaic-storage synergy and integrated power quality management system based on a hybrid transformer, characterized in that: The system, employing the photovoltaic-storage synergy and power quality comprehensive management method based on a hybrid transformer as described in any one of claims 1-7, comprises: a hybrid transformer, a series converter, a parallel converter, a bidirectional DC / DC energy storage converter, a unidirectional DC / DC photovoltaic converter, and a central controller. The hybrid transformer includes a primary winding, a secondary main winding, and an auxiliary winding. The primary winding is connected to the distribution network, the secondary main winding is connected to a low-voltage AC bus, and the auxiliary winding is connected to a common DC bus via the series converter. The two ends of the parallel converter are respectively connected to the low-voltage AC bus and the common DC bus. The bidirectional DC / DC energy storage converter is connected between the energy storage unit and the common DC bus, and the unidirectional DC / DC photovoltaic converter is connected between the photovoltaic array and the common DC bus. The central controller is connected to the series converter, the parallel converter, the bidirectional DC / DC energy storage converter, and the unidirectional DC / DC photovoltaic converter.

Images